User`s Manual

User`s Manual

User’s Manual

Motion Analysis Corporation: Software License Agreement Terms and Conditions

Definitions

The following terms are defined for the purpose of this Agreement as follows:

(a) “Designated System” means the specified computer system described on the facing of this agreement, which Licensee has purchased from the Licensor.

(b) “Licensed Program” means the software program described on the facing page of this Agreement in object code form only, any updates subsequently provided by License, all permitted copies made by Licensee, and all basic or related materials pertinent to such programs.

License

Under a license granted under this Agreement, License is authorized, on a non-exclusive basis, to use the Licensed Program on the Designated System. License shall refrain from taking any action, such as reverse assembly or reverse compilation, to derive a source code equivalent of the Licensed Program. A license shall be valid until terminated under this Agreement. The license fee is part of the purchase price of the Designated System.

Title

The original, and any copies of the Licensed Program, in whole or in part, which are made by Licensee, are the property of

Licensor.

Copies

With each license, Licensee may make one (1) copy of the Licensed Program in object code form only for use by Licensee with the Designated System for backup or archive purposes. Licensee agrees to maintain records of each copy of the Licensed

Program and the serial number of each computer system with which the Licensed Program is incorporated or used.

Copyright Protection

The Licensed Program is copyrighted by Licensor. Copies may be made only as permitted by this Agreement. Licensee agrees to reproduce and apply the copyright notice and proprietary notice of Licensor to all copies, in whole or in part, in any form, of

Licensed Program made hereunder.

Warranty Exclusion

Any maintenance obligations of Licensor shall be subject to a separate maintenance or update agreement between and licensee.

LICENSOR MAKES NO WARRANTIES, EITHER EXPRESS OR IMPLIED ON ANY LICENSED PROGRAMS, AND

EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR

PURPOSE.

Patent and Copyright Indemnification

Licensor shall indemnify Licensee against liability for patent and copyright infringement upon the terms and conditions applicable to Licensee’s purchase of the Designated System.

Termination

This agreement and any licenses granted hereunder may be terminated by Licensor upon written notice if Licensee fails to comply with any of the terms and conditions of this agreement. Upon termination of any license, Licensee will, at Licensor’s option, either return the Licensed Program to Licensor, or destroy the original and all copies and parts thereof.

General

This agreement is not assignable. The rights under this Agreement or any license granted hereunder may not be assigned, sublicensed or otherwise transferred without the prior written consent of Licensor. Except as otherwise specified herein, this is the entire agreement between the parties relating to the subject matter hereof and may only be modified in writing signed by each party. This agreement is governed by the laws of the State of California.

For further information regarding this EVaRT 5.0.3 User’s Manual or other products, please contact:

Motion Analysis Corporation

3617 Westwind Boulevard

Santa Rosa, CA 95403 USA tel: 707.579.6500

fax: 707.526.0629

www.motionanalysis.com

Copyright

©

2007

Chapter 1

Introduction

Topic

Overview

System Requirements

Installing the Software and Licenses

Software Packages within EVaRT

Software Packages Used with EVaRT

For More Information

Overview

Figure 1-1. EVaRT User Interface

Page

1-1

1-3

1-4

1-7

1-15

1-20

This instructional User’s Manual provides a complete description of the

EVaRT

software and its capabilities, along with many step-by-step procedures critical to a successful motion capture project. Motion capture theory is separated from the body of this manual in the form of appendices so that the tutorial approach does not become cluttered.

1-1

Chapter 1: Introduction EVaRT 5.0 User’s Manual

EVaRT

is a complete package, capable of meeting the most demanding requirements of the motion capture industry. Output is generated in realtime making

EVaRT

a suitable engine for a number of widely used 3D animation packages as well as custom applications created using the supplied Software Developers Kit (SDK). Being a real-time application, the results of a motion capture session can be viewed instantly while simultaneously saved in several file formats. In addition, you can graphically edit data with a complete suite of tools without resorting to other off-the-shelf software packages.

EVaRT

handles image data from systems comprised of up to 32 cameras.

System setup and calibration is fast and simple with immediate feedback and a high degree of accuracy and precision. Motion capture sessions are managed using new directory and file access tools and the motion data generated is of the highest quality. Post Processing data is accomplished graphically using intuitive controls integrated with mouse and keyboard functions for fast and easy editing. Model Edit features give you access to the properties of the current set of named markers, virtual markers, linkages, and skeletal segments.

EVaRT

combines three major functions into a single software package:

1.

2.

3.

Calibration of your capture volume

Tracking and identifying marker locations in your calibrated 3D space

Post processing tools for tracking, editing, and preparing the data for other packages

The advanced calibration procedures calibrate the 3D volume with ease and accuracy.

Options for using the software include:

5.

6.

7.

1.

2.

3.

4.

Color video capture-synchronized with the

EVaDV

software—either on your capture computer or on one or more auxiliary computers

Synchronized analog channels—capable of collecting 32 and 64 channels of analog data at any frequency between 60 and 5000 Hz

Genlock to your studio camera with the Eagle and Hawk cameras

Si—Solver for generating constant bone-length skeletons for high quality animation

OrthoTrak—clinical gait evaluation module

SIMM—Software Interaction Muscle Modeling

KinTrak—research module for defining research projects with kinematic and kinetic data and a trial and subject database

1-2

EVaRT 5.0 User’s Manual Chapter 1: Introduction

System Requirements

Table 1-1. Required Minimum and Recommended Specifications

Required Minimum Specifications

Dual 2.0 GHz CPU (up to 12 cameras)

512 MB RAM

Windows 2000

TM

or

Windows XP

TM

operating system

OpenGL video card with 64 MB RAM capable of (1280x1024) resolution

17 inch or larger monitor capable of

(1280x1024) resolution

20 GB hard drive (IDE or SCSI)

1.44 MB or larger floppy disk drive

Internal CD-RW drive

1-port, 1000 Mbps Network Interface Card

(NIC)

104 key keyboard

Three-button mouse. The program requires a middle mouse button for zooming and selecting in several of the graphical panes.

Recommended Specifications

Dual 2.8 GHz Xeon CPU

1.0 GB RAM

Windows 2000

TM

or

Windows XP

TM

operating system

OpenGL video card with 128 MB RAM capable of (1280x1024) resolution

19 inch or larger monitor capable of

(1280x1024) resolution

80 GB hard drive (IDE or SCSI)

1.44 MB or larger floppy disk drive

Internal CD-RW drive

1-port, 1000 Mbps Network Interface Card

(NIC), quantity of 2

104 key keyboard

Three-button mouse. The program requires a middle mouse button for zooming and selecting in several of the graphical panes.

USB 2.0

IEEE 1394 (standard Firewire)

Hardware

Using Eagle and

Hawk Digital

Cameras

Using Falcon

Cameras

Middle Mouse

Button

EVaRT

will perform best with a dual processor host computer with an

OpenGL graphics card. This is used for all types of cameras.

The Eagle and Hawk digital camera motion capture system includes a set of Eagle and Hawk digital cameras with ring lights, LAN/power cables, and an EagleHub system.

If you ar using the Falcon, Cohu, or Pulnix cameras, real-time capture requires a

Midas

computer with a fast CPU, fast memory and upgraded video capture cards. Also, the

Midas

operating system must be upgraded to

Windows NT

TM

to allow the large memory space necessary for realtime capture.

For

EVaRT

operations, the middle button is key for zooming and translating through the 3D and XYZ Graphs display. You will need to verify that the middle mouse button is set to the middle button function.

1-3

Chapter 1: Introduction EVaRT 5.0 User’s Manual

Software

Two new software components are required:

The

EVaRT

program, running under the host

Windows 2000 Professional

TM

, or

Windows XP Professional

TM operating system.

Only if using Falcon cameras—

RT Midas

software with extended capability to handle real-time capture.

Installing the Software and Licenses

To install

EVaRT

for the first time, simply insert the installation CD-ROM into your computer and select the

Setup EVaRT50 No Samples.exe

or

Setup EVaRT50 With Samples.exe

file

Note:

To run

EVaRT,

you will need both a license file and a dongle from Motion

Analysis Corporation. The license file you receive is keyed to your Motion Analysis dongle number printed on the dongle.

Figure 1-2. Parallel and USB Port Dongles, and Flash Drive

Parallel Port Dongle

USB Port Dongle

Motion Analysis Flash Drive

Installation Using the Flash Drive

For new users,

EVaRT

licensing is now provided and setup using Flash

Drives. For installation, please follow these steps:

Note:

Please make sure to remove all Dongles from your computer prior to running the

EVaRT

setup file on the CD. Failure to do so may result in damage to your dongle.

1.

2.

3.

4.

5.

6.

Install the new version of

EVaRT

from the CD-ROM.

Install the Sentinel Drivers after the

EVaRT

installation is finished.

Insert the Dongle into the USB or Parallel port

Insert the Motion Analysis USB Flash drive into a USB port. Doubleclick on the program: Install.Mac.License and follow the instructions.

Press

Y

or

Enter

and the license will be installed onto your hard drive, in the

C:\Program Files\Motion Analysis directory

.

On the Task bar, left-click the Green arrow icon and select

Stop USB mass Storage Device

.

Unplug the USB Flash Drive and store it in a safe place.

If you need any further information, please consult the

readme.txt

file located on the Motion Analysis USB Flash Drive.

1-4

EVaRT 5.0 User’s Manual Chapter 1: Introduction

If You Only Have

EVaRT Installed

If you have

EVaRT

software already installed, you will need to add a new line to your Motion Analysis license file (provided by Motion Analysis

Customer Support).

1.

2.

3.

4.

Launch

Notepad

,

Wordpad

, or your favorite ACSII text editor.

Navigate to

C:\Program Files\Motion Analysis

.

Open the

mac_lic.dat

file.

Add the new line beginning with [EVaRT

v50]

from your new license to the bottom of your current license as shown below.

Figure 1-3. Sample Motion Analysis License File

Motion Analysis License File

Customer: MAC Customer

Platform: NT

SystemID: 19c

Created: 9/15/2005 1:42:26 PM

Sales Order#: 05-xxx

Entered By: Support

[EVa RealTime v5.0] aed50167

[Analog Input]

[OrthoTrak] b9806c31 b2df5e69

[Animation Plugins] b1a50160

[Director/Sequencer] e1a04e65

[RT2 Animation Plugins] e3f05340

[Analog Input] b9806c31

[Calcium 4] e7ed5923

[Skeleton Builder 4] a3f44279

[Reference Video 3.0] eb92592f

[Talon Streaming 4] ecb36136

[Talon Viewer 4]

[BioFeedTrak]

86fb0714 ac943872

[Motion Composer] c7f00e25

This license has no expiration.

873b2d56 d1567841

8964274a

805b5c49

85745819 a069081b d1567841 c363151f

99780c5b cf636a13 d65b4b14 f43d037e

92026c54 c534083f

1-5

Chapter 1: Introduction EVaRT 5.0 User’s Manual

If You Install a

Dongle

Operating Systems in Different

Languages

Computers with a new dongle installed need to load the dongle drivers so that the application will detect the dongle.

You can choose to install the dongle drivers when initially installing the

EVaRT

software or you can run the drivers independently by running the program in the

Sentinel Drivers

directory under the

EVaRT

folder.

If the operating system you are using on your tracking or post processing computer is a non-English version, some characters may not be recognized and you may experience installation problems.

If you are experiencing this, you will need to go into the computer and set it to allow for English Unicode characters. In

Windows XP,

you can do this by going to

Start > Control Panel > Regional and Language Settings

. This brings up a window that has 3 tabs, and the second one is the

Language Tab. Under the Text Services and Input Languages tab, you need to click on the

Details

button and add the setting for English (United

States). This will add the necessary text characters to the computer.

Alternatively, you can install an English language OS on your computer.

1-6

EVaRT 5.0 User’s Manual Chapter 1: Introduction

Software Packages within EVaRT

The following are software products offered by Motion Analysis that are integrated within the

EVaRT

user interface. These files will require a license file and an OCX file to use.

Calcium

Calcium is the graphical user interface to the Solver engine. Solver is the powerful numerical tool for calculating skeleton motion from marker data. The Calcium interface in

EVaRT

is what allows you to correlate the positions of a marker pose to the initial pose of a skeleton. The skeleton is usually created in an outside animation package, such as Maya, 3D Studio

Max or Kaydara and then exported to an HTR file by a Motion Analysis file IO plugin for that package.

Figure 1-4. Calcium Interface in EVaRT

Solver Interface (Si) is the same software as Calcium, only that it’s interface is separate from

EVaRT

.

1-7

Chapter 1: Introduction EVaRT 5.0 User’s Manual

Skeleton Builder

(SkB)

A skeleton, in animation parlance, is a hierarchically connected set of bones with translation and rotation data. Each bone has a parent and potentially any number of children. One special bone has no parent and is usually referred to as the "root" of the skeleton. Skeleton Builder, as the name implies, is a tool that allows you to construct a skeleton by creating bones and arranging them in a hierarchy. Each bone is defined by the motion of three markers used to construct its rotation data. The Skeleton

Builder interface is incorporated with the

EVaRT

interface.

Skeleton Builder bone definitions are stored in the

EVaRT

project file.

Any time you wish to save the definitions you have created simply save out a project file. Various project files are stored in the sample directory which contain the example skeleton at various stages of construction.

Figure 1-5. SkB Skeleton

1-8

EVaRT 5.0 User’s Manual Chapter 1: Introduction

Motion

Composer

Motion Composer is a suite of tools for collating, integrating, and presenting interactive motion capture data. Motion Composer is a collection of authoring tools, data structures, and visualization panes. These are integrated into

EVaRT

to help achieve a seamless workflow for the user to package and present a motion capture session. Some of the key features to be found in Motion Composer are described in the following sections.

Integrated

Authoring

Motion Composer is designed as a plugin into

EVaRT

. This integration allows new and existing users a seamless pathway from data collection to collation and presentation with a minimal learning curve. For current users, this integration leverages their existing knowledge of

EVaRT

. Presentation output can be as simple as redirecting an

EVaRT

project to presentation format.

Interactive Player

Motion Composer ships with Motion View, a freely distributable interactive player that enables customers to distribute their presentations quickly and easily. When it’s time to send research data to a colleague or take it on the road, authors can simply pack-and-go, turning their presentation into a single packaged ZIP file for quick burning to CD or emailing to a friend.

With the interactive player inside, presentations are ready for launch on any Windows 2000/XP operating system.

Rich Media Support

Motion Composer supports the import of a wide variety of information formats, allowing users to import not just motion capture data but data from third party applications, such as EMG analysis graphs, additional color video footage, and still images. It also allows users to import user generated data such as clinical notes, Microsoft Word documents, and embedded HTML link. Below is list of the file formats Motion Composer supports:

Text (TXT, RTF, HTML, XLS)

Color Video (MPG, AVI)

Images (JPG, GIF, BMP)

Motion Analysis OrthoTrak XLS files (cycles, moment, forces, angles, powers)

Project files (PRJ)

Tracks (TRC/TRB)

Analog data (ANB)

User defined data types/views

1-9

Chapter 1: Introduction EVaRT 5.0 User’s Manual

Presentation Tools

The Motion Composer authoring interface provides a simple hierarchical interface to easily manage a user’s disparate files, offering views of referenced files, data structures (e.g. moments, powers, forces), or relationship structures (e.g. Subject/Condition/Trial/Cycle). The Motion Composer authoring interface also makes it simple to create presentations. As data is added, the author simply creates a view of the data they wish to convey and stores this view as a slide. Presentation viewing becomes as simple as a slideshow.

Figure 1-6. Motion Composer Interface

1-10

EVaRT 5.0 User’s Manual

BioFeedTrak

Chapter 1: Introduction

BioFeedTrak

is a general condition, evaluation, and response program for designing and implementing biofeedback programs that can enable clinicians and patients to receive instantaneous audio feedback to kinematic and kinetic variables.

BioFeedTrak

is able to give real-time feedback in the form of sounds based on kinematic and kinetic variables that fall within certain bounds during the pre-defined performance of any type of physical task.

Kinematic variables include position, velocity and acceleration of individual markers placed on key anatomical points of interest. Included angle between two segments (defined by three or four markers) as well as the angle of inclination of a segment (defined by two markers), with respect to a global reference coordinate system, can be used to provide feedback.

Kinetic data include the following:

• horizontal and vertical forces

• moment about the vertical axis

• the coordinates of the center of pressure with respect to the global reference coordinate system

The program works in conjunction with the Motion Analysis Eagle/Hawk

Digital System or the Falcon Analog System. For a typical application procedure, the user will do the following:

1.

2.

3.

4.

Choose and set the variables to be monitored

Determine the starting and ending parameters for each variable to be assessed

Choose the volume and frequency of the audio feedback

Start the Real-Time system

1-11

Chapter 1: Introduction EVaRT 5.0 User’s Manual

With this, the patient and clinician are able to work side by side to retrain areas of the body that need further optimization.

Figure 1-7. BioFeedTrak User Interface with EVaRT

1-12

BioFeedTrak allows users to define a set of variable-condition-response relationships with motion capture data. Each variable-condition-response is referred to as an event. For example, "If the elbow angle of the subject is less than 45 degrees play a tone" where "elbow angle" is the variable,

"less than 45 degrees" is the condition, "play a tone" is the response.

EVaRT 5.0 User’s Manual Chapter 1: Introduction

Digital Video

Option (EVaDV

Software)

The color Digital Video option allows you to record a time-matched Reference Video along with your motion capture trial on a separate computer.

With this option, you will record a time-matched color video AVI file with the same trial name in your motion capture folder. A separate computer is used in order to not burden your

EVaRT

Host computer, which is an issue if your computer is too slow for the number of markers being tracked. For single person captures, you may connect the DV Camera directly to the

EVaRT

Host computer. In this case, the

EVaDV

software is not needed. It is built into the

EVaRT

software. You can run

EVaDV

on one or more computers and then capture multiple AVI files (multiple views). They will all have the same AVI file name. You may experience a small delay in frames from the

EVaRT

software and the

EVaDV

software when capturing. The Color Video display has a pop-up menu with one item, Adjust Frame Offset. This allows for time-matching data streams.

1-13

Chapter 1: Introduction EVaRT 5.0 User’s Manual

QuickDB

QuickDB is an integrated database tool for EVaRT that allows the user to easily track all EVaRT project information. Microsoft Access databases are used to tabulate information about your projects (you don't have to own a copy of Access to take advantage of this tool). A master database keeps a list of all the individual projects you make. Each individual project (called a "session list") keeps track of all the data associated with a specific project directory (project files, VC files, tracks files, etc...). It is fast and easy to create a session list of any data you already have,

QuickDB will scan your project directory for the data you have already collected and will create the session database for you.

QuickDB is all of the following:

Is very handy for keeping track of all your projects, you can scan for and then load project data with ease.

Will record your trials as you collect data.

Contains tracking information about a trial's Post Process status.

Allow's multiple user access to shared session databases.

Makes it easy to share databases with other users.

Allows the creation of capture lists ahead of time so that capture names can be loaded from QuickDB while recording a session.

Is an SQL database that can be used to generate reports on project status.

Figure 1-8. QuickDB Interface

1-14

EVaRT 5.0 User’s Manual Chapter 1: Introduction

Software Packages Used with EVaRT

The following are software products offered by Motion Analysis that are used in conjunction with

EVaRT

. These files will require a license file and a separate installation package.

Animation

Plugins

The MAC Animation Plugins, also known as the File IO plugins, are used to read and write Motion Analysis motion capture files (TRC or TRB).

These plugins are available in all the major animation software packages including:

Maya

3DS MAX

Softimage XSI

Lightwave

Alias Motion Builder.

The Motion Analysis HTR file format is used for skeleton motion capture data. The Animation Plugins can take an existing character skeleton and export it to an HTR file (usually for use in Calcium) and it can take an

HTR file created by Calcium and apply it to the character in the animation software.

The Motion Analysis TRC file format is used for tracked (markers) motion capture data. This data is generally only imported into the animation software. It is used for bringing in full body marker data, face data and prop data.

Figure 1-9. Animation Plugins Interface

1-15

Chapter 1: Introduction EVaRT 5.0 User’s Manual

Talon Plugins

The MAC Talon Plugins, also known as the streaming plugins, are used to stream data from a live, realtime connection to

EVaRT

into an animation software package. Both skeleton and marker data can be streamed. This function is available for Maya, Mocap, and 3DSMAX. Additional animation packages are currently under development.

This same interface is used by outside developers to stream motion capture data into their own custom environments. This programming interface is called the SDK and is available upon request.

Figure 1-10. Talon Plugins Maya Interface

1-16

EVaRT 5.0 User’s Manual Chapter 1: Introduction

Talon Viewer

Note:

Talon Viewer is no longer supported by Motion Analysis Corp.

OrthoTrak

OrthoTrak

is a completely integrated full-body gait analysis package designed for use in clinical and research studies of human locomotion. The system provides state of the art software designed to be used by clinicians in orthopaedics, neurology, and physical therapy, or for any person interested in assessing locomotor abilities of humans.

The system provides:

• quantification of 3D body, segmental, and joint motions

• analysis of the forces occurring in locomotion

• records of neuromuscular function through electromyography

Intended Use

OrthoTrak

is designed primarily for analyzing a walking motion over level ground but can also be used for walking on treadmills, up and down stairs, or other activities. Gait data are presented in graphs which describe the kinematic (angles), kinetics (moments), and muscle activity (EMG).

Data can be exported as industry-standard XLS data sets which can easily be imported into

Microsoft Excel

TM

or other graphics and analysis packages.

Figure 1-11. OrthoTrak Interface

1-17

Chapter 1: Introduction

KinTrak

Director/

Sequencer

SIMM

EVaRT 5.0 User’s Manual

KinTrak

is a movement analysis system that enables you to import and analyze three-dimensional kinetic (force), kinematic (video), and channel

(e.g. EMG) data from a biomechanical perspective.

Existing biomechanical software tends to be geared toward solving specific problems, and is not usually adaptable in new situations. This has resulted in programs that are too restrictive in data control, collection and analysis.

KinTrak

is designed to overcome these problems by allowing users to tailor the program to suit their specifications and the requirements of the project or study they are undertaking.

KinTrak

is intended to be used by researchers and clinicians in the fields of biomechanics and human movement. In addition, coaches and assistants of sports teams may find it useful for their purposes. Possible uses of

KinTrak

include studies for the purpose of prevention of athletic injuries, analysis of gait characteristics, assessment of athletic footwear, and development of prosthetic appliances.

Director/Sequencer

allows the animator to perform non-linear editing of motion data on screen. Just as a sound engineer uses a multi-track digital workstation to edit and combine many tracks of sound, the

Director/Sequencer

gives the animator the ability to do the same with motion captured data.

The animator starts with a collection of files containing motion data for an object or a hierarchical skeleton. Each file that is loaded is displayed as a move in

Director/Sequencer

. Each move is represented as a time line in an

Editor

area and a 3D character in a

Viewer

area.

By simply pointing and clicking with the mouse, the animator can slide a move backward or forward in time in the Editor area. In the Viewer area, both the position and orientation of each figure can be modified. Two moves can be joined together so that there exists a user defined region where one is blended into the other. Finally, any part of a figure’s motion can be hand edited to give the animator complete control of the finished animation.

Each move is stored as a Hierarchal Translations and Rotations (

*.htr

or

HTR) file which contains motion data for a hierarchical skeleton. After all moves have been choreographed, they are once again saved as HTR files which can then be used as input to one of the popular animation software packages to generate the final animated scene.

SIMM

(Software for Interactive Musculoskeletal Modeling) is a software system that enables you to create and analyze graphics-based models of the musculoskeletal system. In

SIMM

, a musculoskeletal model consists of a set of bones that are connected by joints. Muscle-tendon actuators and ligaments span the joints. The muscles and ligaments develop force, thus generating moments about the joints.

SIMM

allows you to analyze and test a musculoskeletal model by calculating the moment arms and lengths of the muscles and ligaments. Given muscle activations, the forces and joint moments (muscle force multiplied by moment arm) that each muscle generates can be computed for any

1-18

EVaRT 5.0 User’s Manual

Applications

Chapter 1: Introduction

body position. By manipulating a model on the computer graphics system, you can quickly explore the effects of changing musculoskeletal geometry and other model parameters.

Since the software can be used to study many different musculoskeletal structures, it can enhance the productivity of investigators working on diverse problems in biomechanics.

SIMM

provides a framework that organizes the parameters of a model and allows people to work together on a modeling project. The moving, three-dimensional images of anatomical structures that you can create are extremely valuable when developing a model and when communicating the results of an analysis.

SIMM

has a wide variety of applications. A few examples include the following.

Biomechanics researchers are using

SIMM

to create models of the human elbow, wrist, jaw, and other anatomical structures. These models can be altered according to particular surgical procedures to study how the surgical alterations affect muscle function. SIMM can also be used to analyze and display the mechanics of injuries.

Neuroscientists are using

SIMM

to study how the central nervous system controls movement. For example, muscle activation patterns determined from electromyographic recordings can be used to estimate muscle forces and joint moments generated during a task. The computed joint moments can then be compared to experimentally recorded moments.

Medical students and residents can use models created with

SIMM

to study musculoskeletal anatomy and function. In addition to visualizing anatomical structures, students gain an appreciation for the interplay of muscle architecture and joint geometry.

Kinesiologists who record and analyze the motion of persons with movement disabilities can use

SIMM

to create three-dimensional animations of a person's movement. Movements, such as walking, can be quantitatively compared to normal movement to gain insight into the causes of movement deformities. Motion can also be analyzed in the context of optimizing athletic performance.

Human factors engineers who need to account for muscle strengths when designing products or work stations can use

SIMM

to study how posture effects muscle strength. Limits on joint ranges of motion can also be taken into account.

Biologists interested in animal movement can create models to quantify limb function. Investigating movement strategies in other species can provide insights needed to design machines that move.

Computer scientists who develop models of the human body for virtual environments can use

SIMM

to create the models and compare them with biomechanical data for verification.

Animators can use

SIMM

to develop realistic representations of human and animal movements. World objects can be added to provide a context for the animation.

1-19

Chapter 1: Introduction EVaRT 5.0 User’s Manual

For More Information

Please contact Motion Analysis Customer Support with any questions, problems, or feedback about the Motion Analysis

EVaRT

software. We can be reached at:

Motion Analysis Corporation

3617 Westwind Blvd.

Santa Rosa, CA, USA 95403

Phone: 707-579-6500

Fax: 707-526-0629 http://www.motionanalysis.com

For technical support and licensing information, please contact:

[email protected]

For information about sales, please contact:

[email protected]

1-20

Chapter 2

Quick-Start Tutorial for

Movement Analysis

Applications

Overview

Starting EVaRT

Project Initialization

System Calibration

Setup Analog

Marker Placement

Data Capture

Topic

Overview

This chapter provides a quick reference to begin using your motion capture system for Movement Analysis applications, and is intended for the more advanced motion capture system user. For Animation Production applications, refer to Chapter 3, Quick-Start Tutorial for Animation Production Applications.

Note:

This Quick-start Guide uses a Helen Hayes marker set and starts with a project file that is located in the

C:\Program Files\Motion Analysis\EVaRT 5.0\Samples\Helen Hayes in Otago

folder. The basic methodology outlined here can be generalized to other marker sets. Please be aware, that this data set is more complex since it utilizes two different marker sets (Static and Dynamic), and captures three different sets of data

(Static, Dynamic Template, and Dynamic movements).

Starting EVaRT

1.

2.

3.

Turn on the Host computer and login.

Turn on EagleHub(s) or Midas computer and turn on the cameras

(analog system only).

Turn on the Forceplate and EMG Amplifiers, if applicable. Make sure to zero the force plate(s) (see the manufacturer manual for more information).

2-1

Page

2-1

2-1

2-2

2-4

2-8

2-9

2-11

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

4.

Launch the

EVaRT

software by double-clicking the icon located on your computer’s desktop.

Project Initialization

Note:

1.

2.

Load a previous project (

File > Load Project

) that has a Helen Hayes static marker set (e.g.

Static.prj

).

Project files contain information about calibration, thresholds, masks, tracking parameters, marker sets, and templates.

By loading a previous project that contains all of this information, you will not have to re-enter it all each time you start a new capture session. You will only need to update the calibration.

Immediately save the project file in a new folder (

File > Save Project

As...

).

Create a new folder for the subject and save your project there.

This directory now becomes the default

EVaRT

directory.

Make sure that you do not write over previous projects. Separate projects are needed in order to run trials for that particular day. If calibration VC files are written over, then recreating the calibration parameters in Post Process mode will not be possible.

3.

In the

Setup > Cameras

sub-panel, choose the Camera Type.

2-2

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

Figure 2-1. Setup > Cameras > Camera Type

Under the Eagle/Hawk settings, choose

Frame Rate

and set the default shutter speed and brightness.

Under the Falcon Camera settings, the user selects the camera and the camera speed simultaneously. Leave the default Midas settings.

2-3

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

4.

Press

Connect to Cameras.

The first time you do this step, a message indicating that “X number of cameras were found, existing project has 2. Do you want to

modify project?” may appear. See Figure 2-2

. Press the

OK

button.

Figure 2-2. Connect to Cameras Status Pop-Up (Example)

System Calibration

Note:

System calibration should be done at a camera speed of 60 Hz.

Start the

Calibration Process

Using the (4-Point)

L-Frame

Calibration Square

1.

2.

3.

4.

5.

Place the calibration square device or four markers (L-shaped) on the floor or on the forceplate.

The Calibration Square markers are described under

Tools > Calibration Settings

. These markers have been placed in a particular orientation and precise distances apart in order to tell the software the origin and coordinate (XYZ) system of the lab/room.

Under the

Calibration > Calibrate

sub-panel, activate the Preview

Calibration check-box.

Press the

Run

button.

All the number buttons on the bottom of the

EVaRT

interface should turn yellow if all of them can see the L-frame.

Select

Layouts > 2 Panes: Top/Bottom

.

We want the 3-D Display window and 2-D Display window showing. These can be set by left-clicking in the window to make it active and then select

Data Views > 3D Display

or

2D Display

.

Check the 2-D views on all the cameras.

There should only be four markers in each camera view. If there are less, you may need to adjust the view of the camera or you can also adjust the threshold to see more markers. If you have

2-4

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

Continue the

Calibration Process

Using the

Calibration Wand

6.

7.

8.

more, you can mask out extraneous data points. To mask, press

Pause.

While in one of the 2-D views, press the middle mouse button and hold it, then drag a square over the bad data to mask.

Check the 3-D display and camera locations.

If it is not already set, right-click and select

Show Cameras

. All of the cameras should be in the correct place.

Optimize the camera positions and their orientation.

New camera positioning should be done at this point if needed.

Right-click in the 3-D view and select

Show Camera Field-of-

View

. You will probably have to change the length of the field of view to more than the default value of 4000 (4 meters). Try 9000

(9 meters).

Turn the capture volume on by right-clicking your mouse in the

3-D display and selecting

Show Volume

. This volume is a visual aid helpful in this process of aiming the cameras properly. The volume dimensions are entered under

Tools > Calibration Setting > Capture Volume

tab in the new window.

You camera field of view should cover the desired volume. Try and align edges of the volume box with edges of the camera field of view. This may require that one person moves the camera on 3 axes, while another person directs the movements.

Once the camera position is optimized, press the

Collect and Calibrate

button in the Calibration with Square field. If sound is enabled and you have speakers turned on, you will hear a sound.

1.

2.

3.

4.

5.

Remove the L-frame from the capture area. It will need to be completely out of view from all cameras.

Set the wand length in the

Calibration with Wand

field.

Make sure the wand length is set at 200mm or 500mm depending on the wand used.

Set the capture duration.

The wand capture duration should be around 60 seconds, or long enough to cover the capture volume. During the 60 seconds, 1/3 of the time should be spent waving the wand parallel to each axis: x, y, and z.

Press the

Collect and Calibrate

button in the Calibration with Wand field.

Begin waiving the wand to cover the capture volume as much as possible.

The object of this exercise is to cover the entire capture volume by waving the wand both horizontally and vertically through the cameras field of view. If you look at the 2-D fields of view, you should have only a small amount of white space. The better the coverage, the better the calibration.

2-5

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

Figure 2-3. Proper Wand Calibration Coverage

6.

When finished, and the Wand Processing Status window appears, you should uncheck

Heavily Weighted Seed

and then press the

Run

Again

button.

This will recalculate the calibration with more emphasis on the wand data, rather than the L-frame.

Figure 2-4. Wand Processing Status Window

2-6

7.

Check the calculated Focal Lengths.

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

Note:

8.

9.

The Focal Length for each camera is calculated and should be close to the value that is set on the lenses.

Check the 3D Residual Values:

The 3D Residual values should be fairly low depending on the type of cameras you have.The Standard Deviation should be approximately half of the 3D residual. Press the

Run Again

button until the values in the calibration processing window stop changing significantly.

When everything looks good and you are ready, press

Accept

. If the calibration still does not meet the desired values, you can press the

Reject

button. You may have to do one part or all of calibration again.

It is possible to calibrate with previously collected files.

10.

Save the project (

File > Save Project

).

When you press

Accept

in the step above, you will get two messages stating “Calibration has been saved”. This message indicates that the project is saved to a system folder. You need to select

File > Save Project

in this step, since the system folder will be overwritten each time a calibration is done.

2-7

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

Possible

Problems with

Calibrations:

How to Solve

Wrong placement or measurements of the L-frame calibration square.

Verify all measurements and x, y, and z axes that are set.

Check the brightness of the cameras and the use/non-use of masks.

Remember to limit the use of masks and make them as small as possible if they are in line from the camera through the intended capture volume. If any markers go through a masked area, the data will be ignored.

Too many extra images in wand are possible causes for a bad calibration. Watch out for anything reflective such as extra markers, reflective material on shoes, shiny floors, debris in carpeting, and sunlight coming in through windows.

If calibration problems persist, contact [email protected]

Setup Analog

1.

In the

Setup > Analog

sub-panel, right-click on the Name column.

Scroll down to

Channel type names

and select the type of Force

Plate that you are using or if you are looking at lower body muscles, select

Muscles

.

Figure 2-5. Setup > Analog Sub-Panel—Channel Type Names

2-8

2.

3.

4.

5.

If you are collecting forceplate data, then select one of the forceplate manufacturers and then select one of the FP1, FP2, FPx… choices available. These correspond to the number of forceplates you have available to use. This will automatically set 6 or 8 channel names

(forceplate dependant) with default voltage settings for each forceplate selected.

For muscles, it will only bring up the muscle selected. The user may also specify their own name for an analog channel (e.g. upper body muscles).

The Range setting also has a drop down menu of varying excitation voltages.These setting should match your hardware (forceplates,

EMG system, or other analog devices.)

Set your sampling rate at some value greater than the frame rate at which you are capturing video data.

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

6.

To activate the channel name, simply click in the On column corresponding to the analog channel. A check mark will appear when active.

Marker Placement

Note:

For the purpose of this illustration, this example uses the Helen Hayes

(both the static and dynamic) marker sets. The theory may be extended to different marker sets.

1.

Attach the reflective markers that are listed in Figure 2-6

and shown in

Figure 2-7

. Placement on bony points is ideal if available. Consult an anatomy book as reference for palpating these points.

Figure 2-6. Helen Hayes Marker Set List

2-9

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

Figure 2-7. Helen Hayes Marker Set Placement

Rear.Head

Offset

V.Sacral

Top.Head

Front.Head

L.Shoulder

R.Shoulder

L.Toe

R.Heel

L.Elbow

R.Elbow

R.Asis

R.Wrist

L.Asis

L.Wrist

R.Thigh

R.Knee

R.Knee.Medial

R.Shank

L.Thigh

L.Knee.Medial

L.Shank

R.Ankle

R.Toe

L.Heel

L.Toe

L.Ankle.Medial

R.Ankle.Medial

2-10

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

Data Capture

Capture a Static

Trial

1.

2.

3.

4.

5.

Go to the

Motion Capture > Output

sub-panel and activate the Raw

Video (.vc), Tracked Binary (.trb), and Analog Binary (.anb) checkboxes.

Type in a filename (e.g. Static). Do not use a number at the end of the filename. The trial number gets appended to the filename. If you need to have a number in the filename, make sure you follow it with an underscore (Static1_) otherwise your first trial will be interpreted by the software as trial 11 not 1.

Set the duration to be 1 second. Have the patient stand in the center of capture volume with arms raised parallel to the floor, thumbs facing forward.

Press the

Record

button.

This will produce a

Static1.trb

file and an analog and raw video file of the same name.

Identify a Static File

1.

Load

Static1.trb

.

Do this either by pressing the

Load Last Capture

button on the

Motion Capture > Output

sub-panel or go to

File > Load Tracks

File,

then select and load it.

Figure 2-8. Motion Capture > Output Sub-Panel—Load Last Capture Button

2-11

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

2.

3.

4.

5.

Loading a file will automatically bring you under the Post Process interface. Select the Identify sub-panel.

Press

Quick ID

. The Identifying window appears. Activate the

Rectify

check-box.

Identify each marker with the correct name.

Play the trial to make sure it is identified throughout the entire trial. If not, go to the frame you used for identification (usually frame 1) and press

Select Visible Frames

(shown in Figure 2-9

) located in bottom right of screen and then press

Rectify

. Check again by playing the trial.

Figure 2-9. Post Process > MarkerSets Sub-Panel

Click: All/None Button

2-12

XYZ Graphs in Bottom Pane

Select Visible Frames Button

6.

Make sure there are no unnamed markers or gaps in the data. If so, activate the XYZ Graphs or in the bottom pane.

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

Load a Walking

(Dynamic) Marker

Set

Create a Template from the First

Walking Trial

7.

Turn all markers on with the

Click: All/None

button, as shown in

Figure 2-9

.

Press the Select Visible Frames button again, then press the

Del_Un

button, which is located on the

Post Process

panel.

Finally, press the

Join Cubic

button.

Save this as a TRB file (

File > Save Tracks

).

1.

2.

Select

File > Load MarkerSet

.

Load the

Walk.prj

file that contains the Helen Hayes Dynamic marker set. This will be the same marker set minus the medial knee and ankle markers.

Save the project (

File > Save As Project

) with your new name (i.e.

Walk.prj), which will become the active project, as will be shown in the top, blue bar. This keeps the calibration for this capture session.

You should now have two project files in the subject’s directory.

2.

3.

4.

5.

1.

6.

7.

8.

9.

Select the

Motion Capture > Output

sub-panel. Activate the Raw

Video (.vc) and Tracked Binary (.trb) check-boxes.

Type in a filename (e.g. Walk).

Set the duration to be long enough to record one full step cycle.

Press the

Record

button. This will produce a

Walk1.trb

file.

Next, load the

Walk1.trb

file.

Done by either pressing the

Load Last Capture

button in the

Motion Capture > Output sub-panel or selecting

File > Load

Tracks File

.

Loading a file will automatically bring you to the Post Process tab.

Select the

Post Process

panel, and then press

Quick ID

. An Identifying window appears. Activate the

Rectify

check-box.

Identify each marker with the correct name by clicking in the 3D view. The stick figure will automatically be drawn as you identify the markers and will help to highlight mistakes (gaps in data, marker misidentifying, swaps, ghost markers).

Play the trial to make sure it is identified throughout the entire trial. If not, go to the frame you used for identification (usually frame 1) and press

Select Visible Frames

, located in bottom right of screen and press then press

Rectify

. Check again by playing the trial.

Gaps in data can be filled by using Join Cubic and/or Join Virtual functions.

2-13

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

10.

Press the

Create Template

button.

Figure 2-10. Create Template Interface

Extend Template

Collect the Walking

Data

Note:

The number of data frames should be at least 80% of the total number of frames. Save this

Walk1.trb

file (save over previous file). It now has the correctly identified marker names.

11.

12.

Save the tracks file by selecting

File > Save Tracks

.

Save project file by selecting

File > Save Project

.

The template becomes part of the project, yet the project still needs to be saved.

When you create a new template it uses only what is in the currently loaded tracks file to make the full template. When you extend the template, it also just uses what is currently loaded, but it doesn't throw out the existing template information. In effect, you are creating a new template from the original tracks file and the one you currently have loaded that has been combined into a single tracks file.

Extending a template will only increase the allowable range of motion in the linkages and never reduce them.

1.

2.

Go to the Motion Capture interface.

Collect the Walk trial. You may want to use a naming convention that adds a descriptor of the movement if your subject is doing multiple trials (e.g. Walk1.trb, Run1.trb, etc.).

If you are collecting force or EMG activity, activate the

Analog

(.anb)

check-box.

If you are using a Falcon/Midas system, you must start your analog data collection while connected to cameras and in pause mode. If you do not, you run the risk of being out-of-sync with your analog data.

The reason for this is that the A/D board and the Midas are clocked differently, and over a period of time, cumulative drift will occur between the analog and video portion of the Motion Analysis system.

This can be manually corrected in the Post Processing interface.

3.

Set the capture duration.

2-14

EVaRT 5.0 User’s Manual Chapter 2: Quick-Start Tutorial for Movement Analysis Applications

4.

5.

6.

7.

The duration should be representative of the length of the trial, typically 5-10 seconds.You can set the duration to be the maximum that you would ever expect, and then you could press the

Record

/

Stop

button for shorter, more typical trials.

Record the data.

Press the

Record

button.This will produce a

Walk1.trb

file as well as VCX and ANB files of the same name.

View the data.

To view analog data while collecting, choose the

2 Panes: Top/

Bottom

layout and set the top window to

3D Display

. Then set the bottom window to

Analog Display (Data Views > Analog

Display).

You can now view one or all of the analog channels in this window.

Edit the data.

Editing is done using the

Data Views > XYZ Graphs.

Save Tracks (

File > Save Tracks

)

This concludes this quick-start chapter for Movement Analysis applications (Helen Hayes/OrthoTrak marker set). If the Post Processing is to be done in the OrthoTrak software module, this requires TRB files for one static and up to several walking trials. The walking trials may also have their associated ANB files if analog (force and/or EMG) data has been collected.

2-15

Chapter 2: Quick-Start Tutorial for Movement Analysis Applications EVaRT 5.0 User’s Manual

2-16

Chapter 3

Quick-Start Tutorial for

Animation Production

Applications

Overview

Starting EVaRT

Project Initialization

System Calibration

Marker Placement

Data Capture

Topic Page

3-1

3-1

3-2

3-3

3-8

3-9

Overview

This chapter provides a quick reference to begin using your motion capture system for Animation Production applications, and is intended for the more advanced motion capture system user. For Movement Analysis applications, refer to Chapter 2, Quick-Start Tutorial for Movement Analysis Applications.

Note:

This Quick-start Guide uses an animation marker set and starts with a project file that is located in the

C:\Program Files\Motion Analysis\EVaRT 5.0\Samples\Animation Calibration

folder. The basic methodology outlined here can be generalized to other marker sets.

Starting EVaRT

1.

2.

3.

Turn on the Host computer and login.

Turn on EagleHub(s) or Midas computer and turn on the cameras

(analog system only).

Launch the

EVaRT

software by double-clicking the icon located on your computer’s desktop.

3-1

Chapter 3: Quick-Start Tutorial for Animation Production ApplicationsEVaRT 5.0 User’s Manual

Project Initialization

1.

2.

Load a previous project (

File > Load Project

) that has an animation marker set.

Project files contain information about calibration, thresholds, masks, tracking parameters, marker sets, and templates.

By loading a previous project that contains all of this information, you will not have to re-enter it all each time you start a new capture session. You will only need to update the calibration.

Immediately save the project file in a new folder (

File > Save Project

As...

).

Create a new folder for the subject and save your project there.

This directory now becomes the default

EVaRT

directory.

Note:

Make sure that you do not write over previous projects. Separate projects are needed in order to run trials for that particular day. If calibration VC files are written over, then recreating the calibration parameters in Post Process mode will not be possible.

3.

In the

Setup > Cameras

sub-panel, choose the Camera Type.

Figure 3-1. Setup > Cameras > Camera Type

3-2

EVaRT 5.0 User’s ManualChapter 3: Quick-Start Tutorial for Animation Production Applications

4.

Under the Eagle/Hawk settings, choose

Frame Rate

and set the default shutter speed and brightness.

Under the Falcon Camera settings, the user selects the camera and the camera speed simultaneously. Leave the default Midas settings.

Press

Connect to Cameras.

The first time you do this step, a message indicating that “X number of cameras were found, existing project has 2. Do you want to

modify project?” may appear. See Figure 3-2

. Press the

OK

button.

Figure 3-2. Connect to Cameras Status Pop-Up (Example)

System Calibration

Note:

System calibration should be done at a camera speed of 60 Hz.

Start the

Calibration Process

Using the (4-Point)

L-Frame

Calibration Square

1.

2.

3.

4.

Place the calibration square device or four markers (L-shaped) on the floor or on the forceplate.

The Calibration Square markers are described under

Tools > Calibration Settings

. These markers have been placed in a particular orientation and precise distances apart in order to tell the software the origin and coordinate (XYZ) system of the lab/room.

Under the

Calibration > Calibrate

sub-panel, activate the Preview

Calibration check-box.

Press the

Run

button.

All the number buttons on the bottom of the

EVaRT

interface should turn yellow if all of them can see the L-frame.

Select

Layouts > 2 Panes: Top/Bottom

.

We want the 3-D Display window and 2-D Display window showing. These can be set by left-clicking in the window to make it active and then select

Data Views > 3D Display

or

2D Display

.

3-3

Chapter 3: Quick-Start Tutorial for Animation Production ApplicationsEVaRT 5.0 User’s Manual

Continue the

Calibration Process

Using the

Calibration Wand

5.

6.

7.

8.

Check the 2-D views on all the cameras.

There should only be four markers in each camera view. If there are less, you may need to adjust the view of the camera or you can also adjust the threshold to see more markers. If you have more, you can mask out extraneous data points. To mask, press

Pause.

While in one of the 2-D views, press the middle mouse button and hold it, then drag a square over the bad data to mask.

Check the 3-D display and camera locations.

If it is not already set, right-click and select

Show Cameras

. All of the cameras should be in the correct place.

Optimize the camera positions and their orientation.

New camera positioning should be done at this point if needed.

Right-click in the 3-D view and select

Show Camera Field-of-

View

. You will probably have to change the length of the field of view to more than the default value of 4000 (4 meters). Try 9000

(9 meters).

Turn the capture volume on by right-clicking your mouse in the

3-D display and selecting

Show Volume

. This volume is a visual aid helpful in this process of aiming the cameras properly. The volume dimensions are entered under

Tools > Calibration Setting > Capture Volume

tab in the new window.

You camera field of view should cover the desired volume. Try and align edges of the volume box with edges of the camera field of view. This may require that one person moves the camera on 3 axes, while another person directs the movements.

Once the camera position is optimized, press the

Collect and Calibrate

button in the Calibration with Square field. If sound is enabled and you have speakers turned on, you will hear a sound.

1.

2.

3.

4.

5.

Remove the L-frame from the capture area. It will need to be completely out of view from all cameras.

Set the wand length in the

Calibration with Wand

field.

Make sure the wand length is set at 200mm or 500mm depending on the wand used.

You can also use previous CalWand VC files.

Set the capture duration.

The wand capture duration should be around 60 seconds, or long enough to cover the capture volume. During the 60 seconds, 1/3 of the time should be spent waving the wand parallel to each axis: x, y, and z.

Press the

Collect and Calibrate

button in the Calibration with Wand field.

Begin waiving the wand to cover the capture volume as much as possible.

The object of this exercise is to cover the entire capture volume by waving the wand both horizontally and vertically through the cameras field of view. If you look at the 2-D fields of view, you should have only a small amount of white space. The better the coverage, the better the calibration.

3-4

EVaRT 5.0 User’s ManualChapter 3: Quick-Start Tutorial for Animation Production Applications

Figure 3-3. Proper Wand Calibration Coverage

3-5

Chapter 3: Quick-Start Tutorial for Animation Production ApplicationsEVaRT 5.0 User’s Manual

6.

When finished, and the Wand Processing Status window appears, you may uncheck

Heavily Weighted Seed

and then press the

Run Again

button. See Figure 3-4

.

This will recalculate the calibration with more emphasis on the wand data, rather than the L-frame.

Figure 3-4. Wand Processing Status Window

7.

In this example (

Figure 3-4 ), not all the cameras can see the seed

device, so not all cameras have calibration values. Press the

Extend

Seed

button to start the calibration process for the cameras that did not see the seed.

Figure 3-5. Wand Processing Status Window—After Extend Seed

3-6

8.

9.

Check the calculated Focal Lengths.

The Focal Length for each camera is calculated and should be close to the value that is set on the lenses.

Check the 3D Residual Values:

EVaRT 5.0 User’s ManualChapter 3: Quick-Start Tutorial for Animation Production Applications

Possible

Problems with

Calibrations—

How to Solve

Note:

10.

The 3D Residual values should be fairly low depending on the type of cameras you have.The Standard Deviation should be approximately half of the 3D residual. Press the

Run Again

button until the values in the calibration processing window stop changing significantly.

When everything looks good and you are ready, press

Accept

. If the calibration still does not meet the desired values, you can press the

Reject

button. You may have to do one part or all of calibration again.

It is possible to calibrate with previously collected files.

11.

Save the project (

File > Save Project

).

When you press

Accept

in the step above, you will get two messages stating “Calibration has been saved”. This message indicates that the project is saved to a system folder. You need to select

File > Save Project

in this step, since the system folder will be overwritten each time a calibration is done.

Wrong placement or measurements of the L-frame calibration square.

Verify all measurements and x, y, and z axes that are set.

Check the brightness of the cameras and the use/non-use of masks.

Remember to limit the use of masks and make them as small as possible if they are in line from the camera through the intended capture volume. If any markers go through a masked area, the data will be ignored.

Too many extra images in wand are possible causes for a bad calibration. Watch out for anything reflective such as extra markers, reflective material on shoes, shiny floors, debris in carpeting, and sunlight coming in through windows.

If calibration problems persist, contact [email protected]

3-7

Chapter 3: Quick-Start Tutorial for Animation Production ApplicationsEVaRT 5.0 User’s Manual

Marker Placement

Note:

For the purpose of this illustration, this example uses a typical animation marker set. The theory may be extended to different marker sets.

1.

Attach reflective markers following the markers listed in Figure 3-6 .

Placement on bony points is ideal if available. Consult an anatomy book as reference for palpating these points.

Figure 3-6. Typical Animation Marker Set

1

13

15

7

4

11

8

5

2

16

9

17

19

21

Note-When placing markers on end segments, the markers should not form a line and should not have mirror symmetry. Thus, thumb and hand markers should never be the same distance from the wrist marker and should be well separated.

20

19

28

29

Note left right asymmetry

30

31

37

36

18

34

32

33

39

35

38

40

41

1

2

4

3

16

7

10

6

22

12

27

24

23

25 26

13

Note single shoulder marker

14

Head and Neck

1. TopHead

2. L_Head

3. B_Head

4. R_Head

5. F_Head

Arms and Hands

11. RBicep

12. RElbow

13. RWrist

14. RPinky

15. RThumb

17. LBicep

18. LElbow

19. LWrist

20. LPinky

21. LThumb

Back and Root

22. MidBack

23. LowBack

24. RootOffset

25. Root

Shoulders and Sternum

6. TopSpine

7. RShoulder

8. FRshoulder

9. FLshoulder

10. ShoulderOffset

16. LShoulder

Pelvis and Hips

26. BRHip

27. BLHip

28. FRHip

29. FLHip

Legs and Feet

30. RThigh

31. RKnee

32. RAnkle

33. RHeel

34. RMidfoot

35. RToe

36. LThigh

37. LKnee

38. LAnkle

39. LHeel

40. LMidfoot

41. LToe

40

41

38

39 33

35

34

32

3-8

EVaRT 5.0 User’s ManualChapter 3: Quick-Start Tutorial for Animation Production Applications

Data Capture

Load an Animation

Marker Set

Create a Template from the First

Walking Trial

1.

2.

Select

File > Load Marker Set

.

Load the

MarkerSetBody.prj

file that contains the animation marker set.

Save the project file (

File > Save As Project

) with a new name (i.e.

Dave.prj

), which will become the active project, as will be shown in the top blue bar. This keeps the calibration for this capture session.

You should now have three files in the project folder.

1.

2.

3.

4.

5.

6.

7.

8.

Select the

Motion Capture > Output

sub-panel and activate the

Raw

Video (.vc)

(if you're doing a live capture) and the

Tracked Binary

(.trb)

check-boxes.

Create a range-of-motion (ROM) file. Type in a filename (i.e. DaveROM) and set the duration to be long enough to record one full step cycle. If you're using the example data, the program will know how long to record. For more information on the ROM files, refer to

“Building a Template from the Range of Motion Trial” on page 9-5 .

Press the

Record

button. This will produce a

DaveROM1.trb

file.

Load the

DaveROM1.trb

file.

Done by either pressing the

Load Last Capture

button in the

Motion Capture > Output

sub-panel or selecting

File > Load

Tracks File

.

Loading a file will automatically bring you to the Post Process tab.

Select the

Post Process

panel and then press

Quick ID

. An Identifying window appears. Activate the

Rectify

check-box.

Identify each marker with the correct name by clicking in the 3D view. The stick figure will automatically be drawn as you identify the markers and will help to highlight mistakes (gaps in data, marker misidentifying, swaps, ghost markers).

Play the trial to make sure it is identified throughout the entire trial. If not, go to the frame you used for identification (usually frame 1) and press

Select Visible Frames

, located in bottom right of screen and press then press

Rectify

. Check again by playing the trial.

Gaps in data can be filled by using Join Cubic and/or Join Virtual functions.

3-9

Chapter 3: Quick-Start Tutorial for Animation Production ApplicationsEVaRT 5.0 User’s Manual

9.

Press the

Create Template

button.

Figure 3-7. Create Template Interface

Start Collecting

Motion Data

10.

11.

Save the tracks file by selecting

File > Save Tracks

.

Save project file by selecting

File > Save Project

.

The template becomes part of the project, yet the project still needs to be saved.

You are now ready to collect data for this subject.

3-10

Chapter 4

Planning a Motion Capture

Session

Topic

Overview

Studio or Lab Preparation

Prior to the Capture Session

Job Assignments and Tasks During the Session

Additional Equipment

Motion Capture Terminology

Motion Capture Session Sequence of Events

Capturing the Data

Overview

The motion capture process starts by collecting raw video data of the subject. The success of the final motion data will depend not only on the quality of the subject’s performance but also on the organization skills and experience of the

EVaRT

operator. The quality of the

EVaRT

data can be greatly affected by the events leading up to and during the motion capture session.

An efficient motion capture session can ultimately save time and money.

Although this chapter is geared towards animation, some information may be helpful for both animation and biomechanics. What follows are suggestions that can help make the motion capture session run smoothly.

Studio or Lab Preparation

At least a day before the capture session, the

EVaRT

user should know the capture volume required and the nature of the motion capture project.

This information is essential for an efficient motion capture session.

Knowing the capture volume allows for the advanced selection of the appropriate marker size for the session.

It may be appropriate to use different capture volumes for the different moves of a capture session. Changing the capture volume size and optimizing this volume could take up to one hour, so this switchover should be scheduled during a break. The approximate volumes can be set up ahead of time using tape on the floor to mark the capture volume boundaries and the position for the tripod legs (if used) of each camera.

4-1

4-3

4-4

4-5

4-6

4-7

Page

4-1

4-1

4-2

Chapter 4: Planning a Motion Capture Session EVaRT 5.0 User’s Manual

If more than one subject will be performing in a capture session, it is a good idea to mark out a capture volume practice area away from the actual capture area. This will allow the next subject to practice before motion capture.

On the

EVaRT

host workstation, create the appropriate directories and project files. For batch edit work in Post Processing, a separate file folder for each project and its associated capture files is strongly suggested.

Make sure there is enough room on the

EVaRT

workstation’s hard disk. If you know the number of trials you are going to capture and the approximate length of each trial, you can estimate the amount of hard disk space you will need. Use some form of backup medium (e.g. CD-ROM, Zip disk) to back up previous data and clear space on the hard disk for the new trials.

Prior to the Capture Session

Several days prior to the capture session, schedule a visit by the subject and any producers or directors involved in the motion capture session. If the subject has not worked with reflective markers, this will allow time to become familiar with marker placement and to practice in the marked-out capture area.

You will want to specify the most desirable type of clothing for the session. Remember, your goal is to capture the fine details of the movement of the body, not the movements of clothing on the body. The rule is to apply markers to skin whenever possible. The areas on the body that present the greatest potential problems are shoulders, the rear neck, sternum, mid back, and the root. A tank-top shirt may be used to expose the shoulder, neck, and sternum. The root marker should be placed low on the spine in an area below the belt line where there is usually very little clothing movement. If a mid back marker is used, the shirt should be rolled up and taped to expose the back. If skin cannot be exposed for all marker placements, then tight fitting clothes or a motion capture body suit should be worn.

Have the subject perform some of the motion capture moves within the capture area. If markers have been placed on the subject, go ahead and capture some data. This would be an ideal opportunity to use a stopwatch to time the duration of each move. These trials can give the

EVaRT

user an indication of potential tracking problems, and if this data is taken all the way to the animation software, it will allow the artist to see how well the data fits their models.

Finally, instruct the subject to speak up during the motion capture if there are any problems with the markers. If markers become loose, they will need to be reconnected more securely with tape or rubber bands.

4-2

EVaRT 5.0 User’s Manual Chapter 4: Planning a Motion Capture Session

Job Assignments and Tasks During the Session

Director

Camcorder

Operator

Scribe

EVaRT Operator

The director ensures that everyone involved is prepared for the capture session and controls the session, including instructing and critiquing the subject’s performance.

The video recording from the camcorder can be very important documentation to aid in choosing the best takes. A video recording may also be useful for post-production promotions. An optional reference video capture is available. Refer to “Digital Video Option (EVaDV Software)” on page 6-21 .

The camcorder should not be allowed to run during the entire capture session. This means that someone should be assigned to start and pause the camcorder for each take.

The camcorder operator should slate each of the takes. This involves recording the take number, the

EVaRT

trial name and number, and any other relevant information on a slate board. After starting the camcorder and before each take, the slate board is held up in front of the camcorder for a few seconds. The audio on the camcorder can also be used. When the slate is held up in front of the camcorder the operator can say the take number, the

EVaRT

filename, and any other information necessary. When the take is complete, the director can make audio comments on the quality of the take.

Someone should be assigned to take notes and to fill in the Motion Capture Log. A sample of this log is found in Appendix M, Useful Blank

Forms .

The take number,

EVaRT

file name, and the duration of the take and any comments from the director or subject should be recorded. The scribe can also do a time check using a stop watch to get the length of data capture for each new move.

The

EVaRT

operator must make certain that the motion capture data is clean and trackable. The

EVaRT

operator must make sure the camera calibration is good and that raw calibration data is collected at various times throughout the session as insurance. This is particularly important when there are several people around and a camera might get bumped accidentally. The operator should watch for reflections, changing light conditions, such as sunlight coming through a window, or other external variables which may affect a capture.

4-3

Chapter 4: Planning a Motion Capture Session EVaRT 5.0 User’s Manual

Additional Equipment

Props

Camcorder

Still Camera

Slate Board

Markers, Tape,

Pre-tape Liquids,

Rubberbands

Backup Media

If the motion capture session requires the use of props, this must be known to all parties well in advance. The type of prop and its use are very important because reflective markers may have to be attached to the prop as well as the subject. One marker may be used to track position, but as many as three markers may be required to show all rotations of the prop.

Many props that would ordinarily seem simple become very difficult to deal with during a motion capture session. A good example is the use of a ball as a prop. If a small ball is only being held, one marker may be used to track position. If a large ball is being bounced, three markers may be required to show all rotations.

Reflective or glossy material should

not

be used in the construction of props, and very large props may occlude the subject’s markers. Remember, the design of the prop and how it affects the subject’s movement are more important than the prop’s physical appearance. Props may also be assigned separate templates (see “Multiple Tracking Objects” on page

9-7 ).

Used to completely document each trial, a camcorder will allow the producer to rank the trials of a move and will also give the animation artist something to use as a reference for the completed animation. See also

“Digital Video Option (EVaDV Software)” on page 6-21 .

Photographs of the subject, with markers attached, will help the artists understand the correspondence between the marker data and the actual figure.

A slate board and chalk or grease-pen board will provide an easy way to relate the camcorder record to the

EVaRT

data.

An adequate supply of reflective markers, double stick tape, paper tape, and “Tuff Skin” or “New Skin” should be available. For rough and tumble sessions, the best method to adhere markers is by using Velcro TM on a skin-tight motion capture suit. However, markers can be applied directly the skin.

If markers must be placed directly on the skin and the subject will be performing athletic moves in which perspiration might be a problem, pretape liquids like “Tuff Skin” or “New Skin” can help make double-stick tape adhere better. These products must be applied to dry skin and allowed to set for a minute or two before the marker is attached.

Rubberbands looped around the marker and limb also work well to stabilize the markers. Rubberbands can be looped together to increase diameter and prevent restriction of blood flow. Rubberbands can be used around the elbows, wrists, hands, knees, ankles and toes.

Spare CD-ROMs, Zip disks, or some other backup medium should be available for backups and data transfers.

4-4

EVaRT 5.0 User’s Manual Chapter 4: Planning a Motion Capture Session

Motion Capture

Body Suit

Camcorder Tapes

Music Player

Stop Watch

Sample Form

A motion capture body suit with Velcro attachments for markers provides a quick way to prepare a subject for motion capture. The use of the body suit is especially effective when subjects are involved in rough or contacttype motion capture sessions, common in animation applications.

Depending on the length of the capture session, spare video tapes should be on hand.

Either a CD or tape player can provide musical accompaniment. Music helps calm and smooth out the subject’s performance not only with dance, but athletic moves as well.

A stop watch is handy for calculating the duration of each new move.

You should decide at the outset whether you will build a hierarchical skeleton. If you decided to, there are two software methods available:

SkB

(Skeleton Builder)

and

Calcium/Si.

Blank forms to help you define skeleton parameters for each of these methods can be found in Appendix M, Useful Blank Forms . You may want to copy one of these forms for recording your project measurements.

Motion Capture Terminology

Some terms that are useful to a motion capture session are

moves, trials

,

and

takes

.

Move

A

move

is an event or routine performed by the motion capture subject. A move can be as simple as a neutral stance position, or as complex as a 2 person, 30 second dance routine. The director and subject will work from a

move list

.

Trial

Multiple

trials

of a move should be taken. The number of trials depends on the complexity of the move, the subject’s performance, and quality of the

EVaRT

raw data. Usually, three trials per move is adequate. It is important that the director or subject’s comments about the quality of the trial (which trial was the best) be recorded on the Motion Capture Log.

Knowing which trial of a move is the best will allow the

EVaRT

user to track only the best trial.

Take

A

take

is the master number used to relate what is on the camcorder’s video tape to the

EVaRT

filenames and trial numbers. The

take number

is displayed on the slate board and on the Motion Capture Log. Every new image recorded on the video tape should have a new take number. This should include calibration collection, initialization and T-pose/Init pose stance positions. You should never re-use or redo a take number. If a data collection is aborted for some reason, e.g. a marker fell off, then the

EVaRT

filename and trial number can be overwritten, but the take number should change.

4-5

Chapter 4: Planning a Motion Capture Session EVaRT 5.0 User’s Manual

Motion Capture Session Sequence of Events

The Day Before:

1.

Optimize the camera positions and orientation to the capture volume.

2.

3.

4.

5.

6.

Calibrate the volume of the capture area.

Determine the correct marker size to use.

EVaRT

raw data should show 2 lines or greater per marker.

Setup the

EVaRT

project with the correct markers, virtual markers, linkages, segments, etc.

If possible, collect and track the markers on a person to verify that the tracking parameters are optimal.

Verify that there is enough space on the workstation’s hard disk. If there is not enough space, back up the previous files and then erase them from the hard disk.

7.

Organize the markers, tape, and props to most efficiently facilitate the session.

The Day of the

Motion Capture

Before the subject arrives:

1.

2.

3.

Load the

EVaRT

project.

Optimize the threshold settings.

Collect calibration data sets (both seed and wand).

Subject Preparation

1.

Ensure subject’s clothing is appropriate.

2.

Allow the subject to warm up.

3.

Attach the markers according to predetermined placements.

Note:

Asymmetrical marker placement on the subject is critical for obtaining the best marker data.

4.

With the markers in place, take still photos of the subject from the front, side, and rear view.

Note:

If you are taking photographs, do not use the flash attachment on the still camera while you are collecting data. A flash during data collection can corrupt the data.

5.

6.

Allow the subject to practice in the capture volume with the markers on.

Prepare for the calibration collection. Explain to everyone the importance of not bumping the camera tripods.

4-6

EVaRT 5.0 User’s Manual Chapter 4: Planning a Motion Capture Session

Capturing the Data

Calibration

Note:

This section provides a general overview of the calibration process. For complete calibration information, refer to Chapter 8, Calibration Panel.

Collect the Square

(Seed) Calibration

1.

2.

3.

4.

Fill out the Motion Capture Log and slate board for the first square

(seed) calibration. This would be Take 1 and an

EVaRT

filename, for example “

CalSeed

”.

Prepare the

EVaRT

system for data collection. Press the

Collect and

Calibrate

button to trigger the event button.

Verify that the camera buttons turn yellow after the Seed calibration is complete.

Remove the calibration seed device (calibration square) from the capture volume.

Collect the Wand

Calibration

For best results it is recommended that you collect and use wand calibration data.

Prepare the

EVaRT

system for wand calibration. The duration of the wand calibration is directly correlated to the capture frame rate. A typical duration for a small capture volume is 30 to 60 seconds. Large volumes with ten or more cameras can take 120 to 180 seconds, and very large volumes may take up to 240 seconds.

Collect and verify that the wand calibration data is good. It may be necessary to reposition or move cameras and to retake both the seed and the wand calibration data if one or more cameras has large areas without wand calibration data. You will also want to uncheck the

Heavily

Weighted Seed

check-box, and press

Run

again. Keep pressing

Run

after it finishes, until the calibration numbers stop changing.

Your wand data should cover the entire capture volume. A common method of ensuring better wand data is to use a 1/3 method. That is, hold the wand markers in alignment along each axis (X, Y, and Z) for 1/3 of the wand capture session.

Collecting Trial

Data

Subject

Initialization

The type of subject initialization depends on the application.

In animation applications it is the “T-Pose” or “Init Pose” trial

For

OrthoTrak

it is the Static Trial

For

KinTrak

it is the Neutral Trial

In general, the procedure is as follows:

4-7

Chapter 4: Planning a Motion Capture Session EVaRT 5.0 User’s Manual

Capturing the

Moves

Collecting

Calibration

“Insurance” Data

Wrapping It Up

Note:

1.

2.

3.

4.

5.

6.

7.

Have the subject stand in the capture volume with the markers on. On the

EVaRT

system. Look for any reflection and light source that might interfere with the capture and correct the problem.

Prepare the

EVaRT

system by entering a filename. If the subject’s name is Jane use something like JaneInit1. For this initialization file, use a duration of 2-3 seconds. The take number should be set to 3.

Click on

Collect

to arm the event button. You can enter a long duration (e.g. 20 seconds) and then press the hand-held event button a second time at the end of the move.

Update the slate board with the new take number and

EVaRT

filename. Make the same entries in the Motion Capture Log.

When everybody is ready, the director can say “roll video”, then slate the video. Now the

EVaRT

operator gives the signal for the subject to start and presses the event button to start the data collection. The event button must be pressed a second time to stop data collection at the end of the move.

For initialization, use the T-pose/Init pose. In this pose the subject faces forward and raises both arms straight out from their sides with the thumbs oriented up. Perform this motion within the duration of the capture time.

Pause the camcorder.

Collect two trials of this “initialization” move.

1.

2.

3.

4.

5.

6.

7.

8.

9.

Before capturing, have the subject practice each move.

Enter a duration

longer

than the estimated length of the move.

Enter an

EVaRT

filename, the duration of data capture, the trial number and trigger the Event button.

Update the slate and the Motion Capture Log.

The director should ask if everybody is ready and then say:

“Roll video”, “Slate video.”

The

EVaRT

operator presses the event trigger button and the subject begins the trial. When the trial is finished, the event trigger button is pressed again to complete the capture.

Comments on the quality of the trial should be entered into the

Motion Capture Log and on the audio of the camcorder.

The

EVaRT

operator quickly reviews the raw data and looks for any problems.

This process is usually repeated for 3 trials of each move that is scheduled.

Usually only one trial is tracked. The other trials are there for insurance and to allow the end user to pick the best trial.

For insurance, it is a good idea to periodically collect raw calibration data whenever there is down time. As the number of people increases in the capture studio, the chance for bumping a camera (if tripods are being used) increases and “insurance” calibration data suddenly becomes very valuable.

After all the trials have been collected, perform the following to wrap up the process:

4-8

EVaRT 5.0 User’s Manual Chapter 4: Planning a Motion Capture Session

1.

2.

3.

4.

5.

6.

Collect the last calibration trial.

Backup all the

EVaRT

trials on a CD-ROM, Zip disk, or other backup medium. Label and store the tape in a safe place.

Remove the video tape from the camcorder and set the safety tabs on the tape to prevent being recorded over.

Consolidate and make copies of the motion capture logs and forms.

If necessary, give the video tape and Motion Capture Logs to the director so the best trials of each move can be indicated.

Place logs and offset forms in a binder. Clear plastic inserts can be added to hold the still photos. The binder along with the video tape will provide important information to both the

EVaRT

user tracking and editing the data, and for the artists who will apply the final data to the model.

4-9

Chapter 4: Planning a Motion Capture Session EVaRT 5.0 User’s Manual

4-10

Chapter 5

Camera Setup

Topic

Setting Up a Motion Capture Laboratory

Deciding On the Optimum Number of Cameras

Setting Up the Cameras

Eagle Camera Physical Dimensions

Hawk Camera Physical Dimensions

Hawk-i Camera Physical Dimensions

Overview of the System Calibrating Process

Placing the Calibration Square

Marker Sizes and Maximum Distances for Eagle/Hawk

Cameras

Troubleshooting Eagle and Hawk Camera Problems

Relationship Between Capture Volume and Marker Size

Page

5-1

5-2

5-11

5-17

5-18

5-19

5-20

5-22

5-23

5-24

5-29

Setting Up a Motion Capture Laboratory

Camera placement is the most important aspect of setting up your motion capture laboratory. If properly done, good camera placement will reward you with highly accurate and consistent results, and greatly reduced editing time.

Optimum

Laboratory

Conditions

Fluorescent lights are the best ambient light when red or notch filters are used on the motion capture cameras.

Windows should be covered with curtains to eliminate direct outside light.

Carpeting or other non-shiny floor surfaces are preferable to tile flooring which can reflect opposite ring lights.

For analog cameras (Falcon, Cohu, etc.), power outlets should be located within ten feet of each camera’s tripod, or extension cords may be used.

Recommended

Supplies

A stepladder—for adjusting the cameras/tripods.

Masking tape—to mark the floor when measuring the capture space and setting up the cameras.

Reflective markers—to attach to the subject and also enough to place on the floor when adjusting the cameras.

Other supplies include surgical tape, electrode collars for applying markers to people and gaffer’s tape (black masking tape).

5-1

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Deciding On the Optimum Number of Cameras

There are several objectives to consider when deciding how many cameras should be used and where they should be placed.

1.

2.

There should be a sufficient number of cameras to insure that, at all times, all markers will be visible by at least two, and preferably three, cameras. In general, the number of cameras must be increased when:

• the motion of the subject becomes less constrained

• the number of subjects or objects increases

• the capture volume increases

As more cameras are used, each camera should view only a portion of the capture volume to achieve higher accuracy and prevent too many cameras from seeing any one marker. The only requirement is that all

4 markers on the Calibration Square should be visible in at least 1/2 of the cameras. You can then use the Extend Seed function (see

“Extending the Seed Calibration” on page 8-18 ).

Note:

When more than 5 or 6 cameras see the same marker, the accuracy of tracking is not increased and computation time increases.

3.

Camera views should not include areas outside the capture volume to ensure the highest possible spatial resolution.

The number of cameras in a typical motion capture setup can be as few as

3 or as many as 64. The following provides some guidelines for deciding on the number of cameras to use. In the following figures, all measurements are in meters.

6 Cameras

For motion capture involving only one subject, where the occlusion of markers is not a problem, six cameras may be adequate. This configuration is often used for gait analysis and other similar biomechanical applications. The two end cameras are often tilted so that the long axis of the view areas is vertical. For optimum results, all cameras should be about

2.5 meters above the floor. See

Figure 5-3 .

8 Cameras

As a wider range of motion is allowed, the probability of markers being occluded increases to the point that eight or more cameras are required.

This is the minimum recommended configuration for animation applications. Cameras should be about 3 meters above the floor. See

Figure 5-4

.

10 Cameras

In an elongated capture space, ten cameras may prove beneficial. The first

8 cameras should be placed about 3 meters above the floor as in the 8 camera setup. The two additional cameras (9 and 10) should be placed 5 meters above the floor at each end of the long dimension of the capture volume and will probably have longer focal length lenses than the other

cameras. See Figure 5-5

.

12 Cameras

As the capture volume becomes more elongated, twelve cameras may be required. The first 8 cameras should be placed 3 meters above the floor.

Cameras 10 and 11 should be placed as 5 meters above the floor on the ends of the capture volume but closer to the center than cameras 9 and 12.

Therefore, one end of the long volume will be covered best by cameras 9

5-2

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

14 Cameras

16 Cameras

More than 16

Cameras

Capture Volumes

Between Eagle and

Hawk Cameras

and 11, while the other end will be covered best by cameras 10 and 12.

See

Figure 5-6

.

When the sides of the capture volume are too long to be adequately covered by four cameras on each side, an additional pair of cameras with wide angle lenses can be placed in the center of each long side of the capture volume. The first 8 cameras should be 3 meters above the floor. Cam-

eras 9 through 14 should be 5 meters above the floor. See Figure 5-7

.

To use more than 14 cameras effectively it is usually necessary to break the capture volume into two overlapping sections across the long axis.

Each section is calibrated separately with its own square. Two squares can be used simultaneously or one square can be placed in two carefully measured positions in sequence. Every camera must see at least one of the squares in its entirety. All cameras should be placed 3 meters above the

floor. See Figure 5-8 .

For more information and an example, refer to

“Overview of the System

Calibrating Process” on page 5-20

.

As capture volumes increase in size, more than 16 cameras may be required. It is best to consider the capture volume as two or more overlapping regions. For large square shaped capture volumes, up to 32 cameras

can be used with the space broken into four regions. See Figure 5-9

.

Eagle and Hawk cameras use the same high-powered ringlights and have the same limits for marker distances. The difference between the two cameras is that you can use smaller markers with the Eagle cameras

(about one-half the size of the Hawk markers).

5-3

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

16 Camera, Two-

Tier Setup

F o r a n e x a m p l e o f t h i s 1 6 c a m e r a , t w o - t i e r s e t u p , o p e n t h e

MAS_16Camera_2Tier.prj

file in the following directory:

C:\ProgramFiles\MotionAnalysis\EVaRT50\Samples\LargeVolumes

Also refer to

Figure 5-1 .

Figure 5-1. 16 Camera, Two-Tier Setup

Tier 2: 4 Cameras

Tier 1: 12 Cameras

5-4

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

28 Camera, 3-Tier

Setup

F o r a n e x a m p l e o f t h i s 2 8 c a m e r a , t h r e e - t ie r s e t u p , o p e n t h e

Spectrum_28Camera_3Tier.prj

file in the

C:\ProgramFiles\MotionAnalysis\EVaRT50\Samples\LargeVolumes

directory. Also refer to Figure 5-2

.

Figure 5-2. 28 Camera, 3-Tier Setup

Tier 3: 8 Cameras

Tier 2: 8 Cameras

Tier 1: 12 Cameras

Ideal Capture

Volume Sizes

Calculating the ideal volume size for a specific camera setup can have many factors involved and it can become a very hard question to answer.

However, a good starting point would be to assume a two person, full body capture area in an ideal space (no restrictions on camera placement, etc.). For this we suggest the figures listed in

Table 5-1

.

Table 5-1. Ideal Volume Sizes for Specific Eagle and Hawk Camera

Setups (with standard lenses a

)

Number of Cameras

6

10

14

16

32

Dimensions (m)

5 x 2

7 x 5

9 x 6

13 x 6

13 x 11

Area (m

10

35

54

78

143

2

)

a.Standard lenses are 18-35 mm Zoom for Eagle cameras, and 6 mm C-

Mount for Hawk cameras.

5-5

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

So how is it that 16 cameras give you 4 times the capture area of 8 cameras (you might ask)? Mostly because when using a small number of cameras you end up wasting a lot of the usable viewing cone of each camera.

Using more cameras allows for more efficient usage of each individual camera.

Table 5-1 is for Eagle and Hawk cameras. For Falcon cameras, you

should multiply each dimension (each side) by about 0.8, which gives you

64% of the capture area compared to the same number of Eagle and Hawk cameras.

Figure 5-3. Typical 6 Camera Setup

Note

: Capture volumes may vary depending on room size and the distance from the camera to the capture area.

5-6

EVaRT 5.0 User’s Manual

Figure 5-4. Typical 8 Camera Setup

Chapter 5: Camera Setup

Figure 5-5. Typical 10 Camera Setup

5-7

Chapter 5: Camera Setup

Figure 5-6. Typical 12 Camera Setup

EVaRT 5.0 User’s Manual

Figure 5-7. Typical 14 Camera Setup

5-8

EVaRT 5.0 User’s Manual

Figure 5-8. Typical 16 Camera Setup

Chapter 5: Camera Setup

5-9

Chapter 5: Camera Setup

Figure 5-9. Typical 32 Camera Setup

EVaRT 5.0 User’s Manual

5-10

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Setting Up the Cameras

The key to placing cameras around the capture area is to position them where they will yield the highest resolution without excluding any part of the adjacent capture volume. In other words, if you plan to track 2 gait cycles, do not set up an area suitable for 4 gait cycles. When the tracking volume is increased, the quality and accuracy of the tracking data will decrease.

First, you will want to measure the room to establish the center of the tracking area.

An Example Eagle or Hawk Camera

Setup

For example, if you have a 10 x 15 meter room and you are using 8 Eagle or Hawk digital cameras:

1.

2.

3.

4.

5.

Measure in from the walls 5 and 7.5 meters. This should be the center of a 10 x 15 meter room.

Mark the center of the room or tracking area with a piece of masking tape.

Find the corners of the actual capture volume. For optimum tracking, the length and width of the capture volume should be no more than about half the room dimensions.

Position the cameras evenly around the capture area. Place the cameras above the top of the capture space, looking down, to prevent cameras from seeing an opposing camera’s ring light.

For most gait analysis installations, a height of 2 meters should be sufficient. For a larger capture area (e.g. full body or sports analysis), the cameras may need to be raised higher.

If a camera must view an opposing camera, use the mask function in

EVaRT

to block the offending image. Refer to “Creating and

Clearing Masks” on page 7-9 .

Position the cameras so that they are equally spaced when viewed from the center of the capture area.

Beware of making the capture area too large. The resolution and the quality of the data may be compromised.

Adding cameras low. Good results can come from adding cameras or positioning the cameras low (on the floor), looking up at the capture subject. This is also effective for capturing markers as the subject is stooped over or lying on the floor. Opposite camera ringlights can be masked out if necessary.

5-11

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

6.

7.

One by one, adjust each camera so it sees as much capture area as possible.

To see the camera view, right-click in the 3D display and then select

Show > Show Camera Field of View.

See

Figure 5-10 .

Figure 5-10. Show Camera Field of View

5-12

Right-Click Menu

In 3-D Display

You can adjust the depth of the camera view by moving the slider in the

Camera Depth of Field function (usually located in the upper-left corner of the interface when the Show Camera Field of View function is activated). This does not change the depth of view the camera will have. It only provides a visual aid to determine if an object at a particular distance will be in the camera’s field of view.

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Figure 5-11. Show Volume

8.

To see the capture volume, right-click in the 3D display and select

Show Volume

.

5-13

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

9.

To See the camera coverage in the volume, select the

Show Camera

Coverage

check box in the

Calibration > Refine

sub-panel, and then right-click in the 3D view and select

Show Volume

. Select Show

Field of View again to turn the camera rays off. See

Figure 5-12

.

Figure 5-12. Show Camera Coverage

Show Camera Coverage

5-14

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

10.

11.

12.

For the example capture area of 5 by 7 meters, 2 meters high, most

Eagle and Hawk camera lenses can be set to a focus of 00 (infinity) and an f-stop that is wide-open (smallest number).

Set the shutter speed in the

EVaRT

software so the markers are bright and have a good threshold setting (usually about 500).

Place the Calibration Square in the center of the taped area. A useful convention is to place marker #1 on the square closest to camera #1.

Note:

The cameras need at least 20 minutes to warm up before collecting calibration or trial data.

Tracking With More

Than 8 Cameras

As the subject moves from one region to the next in a multiple region capture volume,

EVaRT

has no problem as the subject leaves the view of some cameras while entering the view of others. The only requirement is that at least two (preferably three or four) cameras can see the subject at all times.

For additional cameras to be effective, they must be sufficiently far apart so that the rays from a given marker to the two adjacent cameras subtend a large enough angle to yield good positioning data.

Using Many

Cameras in a Small

Volume

It is possible to use eight or more cameras effectively in a relatively small volume if there is sufficient height. We suggest placing half the cameras at a moderate height and the other half as high as possible. You may need to experiment to obtain the optimum camera adjustment for your lab or studio.

5-15

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Adjusting

Camera View for

Increasing

Height

If your capture volume is too high for your Eagle or Hawk cameras, you may turn the cameras on their side (just as photographer may turn their camera on its side for increased height). Note that your camera width coverage will decrease. You may turn the cameras on their side, up to 89

° without having to make any changes to the software settings. If you turn the camera 90

°

or greater, you will need to select the Alternate setting for the particular camera(s). This is done in the

Calibration > Calibrate

subpanel. Select

Details > Lenses/Orientation

, then change the setting from

Normal

to

Alternate

. If the camera is hanging upside-down, you will need to use the Alternate position.

Note:

Do not leave the cameras set too close to 90

°

(i.e. 85

°

to 95

°

) since it may appear Normal or Alternate and result in non-repeatable calibrations.

Figure 5-13. Lenses/Orientation Window

5-16

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Eagle Camera Physical Dimensions

The following diagram illustrates the physical size and weight of the

Eagle digital camera. The tripod mounting points are the holes used to hold the tripod bolt. There are four tripod mounting points on each camera.

Figure 5-14. Eagle Camera Physical Dimensions

196.25 m m (18-5

5 mm le ns)

189.60 mm

147

.70 m m

Tripod Mounting Points = 1/4 in. diameter x 20 threads/inch

Camera Weight = 2.2 kg with lens (4.9 lbs.)

5-17

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Hawk Camera Physical Dimensions

The following diagram illustrates the physical size and weight of the

Hawk digital camera. The tripod mounting points are the holes used to hold the tripod bolt. There are four tripod mounting points on each camera.

Figure 5-15. Hawk Camera Physical Dimensions

124.50

mm (6 mm le ns)

189.60 mm

147

.70 m m

Tripod Mounting Points = 1/4 in. diameter x 20 threads/inch

Camera Weight = 2.1 kg with lens (4.7 lbs.)

5-18

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Hawk-i Camera Physical Dimensions

The following diagram illustrates the physical size and weight of the

Hawk-i digital camera. The tripod mounting points are the holes used to hold the tripod bolt. There is one tripod mounting point on each Hawk-i camera.

Figure 5-16. Hawk-i Camera Physical Dimensions

12

4.5

0 m m

189.60 mm

14

7.7

0 m m

Tripod Mounting Points = 1/4 in. diameter x 20 threads/inch

Camera Weight = 2.1 kg with lens (4.7 lbs.)

5-19

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Overview of the System Calibrating Process

Three-dimensional tracking is performed in two stages: a seed and a wand calibration. A relationship must be established between real-world positions (object-coordinates) and the corresponding image-coordinates from the camera view. This is called calibrating the system. When a target is visible in two or more camera views, there is sufficient information available to track the targets in three-dimensional space.

The calibration of a given camera’s view is completely dependent on the camera lens focal length and the position and orientation of the camera with respect to an arbitrary reference frame called the object-referenceframe. A change of any sort, which alters the relationship between the object-coordinates and image-coordinates, must be followed by a fresh calibration. This includes accidently bumping a camera tripod.

The calibration process calculates eleven calibration coefficients which implicitly define the configuration of a particular view. The calibration coefficients, together with the image-coordinates of a single target, are sufficient to define the path of an optical ray from the target to the camera through the object-space. If rays from two cameras intersect in space at a specific time, they define the 3D position of a target at that time. Therefore, the tracking process is one of intersecting optical rays generated from different views of the same event.

EVaRT

employs a “best fit” tracking algorithm using only good camera views.

The Calibration

Coordinate

System

In order to calibrate the system, you must first decide on the location of the origin and orientation of your object-reference-frame. This is determined by the Calibration Square. All results generated by the tracking process are referred back to this reference frame.

The selection of an object-reference-frame is arbitrary. However, judicious selection is advised. In most cases, it is advisable to align one axis of the frame with the axis of gravity and another with the predominant direction of motion. Remember, that all targets will be tracked with respect to the object-reference-frame, and that the units used to locate the control points (mm, cm, inches, etc.) will be the same units used in the tracking process.

For computer animation users, a commonly used coordinate system convention is called “Y-up”, with the Y axis pointing up, the +Z axis normal to the direction of motion and the +X axis oriented from the person’s right side to left side. Looking at the frontal view, you would see a normal X-Y plot (Y-up, X-right) and the +Z coordinate sticks out of the screen.

For biomechanics applications, it is common to use a coordinate system with +Z up, +X in the direction of forward motion, and +Y toward the subject’s left side.

The coordinate convention you use is your choice. Be sure that:

The coordinates of the calibration square are entered into your project file(s) correctly.

5-20

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

The Calibration Square is oriented correctly in the room when you collect the calibration trial. The position of the Calibration

Square determines the orientation of your calibration.

Note:

The International Society of Biomechanics (ISB) has officially adopted the convention that the Y axis should point up. This has the advantage that in both 2D and 3D studies, the Y axis is up. However, many studies and software packages use the Z-up coordinate system favored by mathematicians.

Control Points

Once a reference frame has been selected, you must provide a number of calibration markers with known locations, which can be used for control purposes; hence, these calibration markers are known as control points.

The control points serve much the same purpose as the simple scale widely used for two-dimensional studies—they are, in fact, a three-dimensional yardstick representing the X, Y, and Z dimensions.

Motion Analysis offers a calibration square with four retro-reflective spheres. The relative position of the spheres have been accurately measured.

Place the calibration square at the origin (or at an accurately measured point) of the laboratory’s test area. When placing the calibration square, consider the direction of motion to be studied, position of force plates, etc. You can change the orientation of the calibration square by making adjustments to the Origin Offsets table. This is located in the

Calibration

> Details > Calibration Settings Origin Offsets

tab. Refer to

Figure

5-17

.

Figure 5-17. Calibration > Details > Calibration Settings Origin Offsets

5-21

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Placing the Calibration Square

Mark the floor area with tape where the motion is to take place.

Set Marker #1 of the Calibration Square at the desired origin of the capture volume.

If the orientation of the coordinate system is not important, the square should be rotated so as many control points as possible can be seen by all cameras.

Check the video monitor to see if any of the four points on the

Calibration Square are merging. The Calibration Square should be seen by at least half of the cameras to give a good calibration.

The other half of the cameras can be calibrated using the wand, with an Extend Seed menu item. You may need to adjust the cameras at this point.

Figure 5-18. Placing the Calibration Square In the Capture Volume (Z-Up)

1

8

7

2

3

4 Y

4

Z

2

3

1

X

5

6

5-22

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Marker Sizes and Maximum Distances for Eagle/Hawk

Cameras

The limiting factor in what size marker works in what volume is the distance the marker is viewable (and usable) from the camera.

"Min 3 lines" below is the minimum number of scan lines (or pixels) to allow in calculating a 2D centroid. As a rule, the more lines, the cleaner the 2D data and resulting 3D data. But once you get above 3 lines, the data is very clean and going more lines per centroid does not generally make the data any better.

Table 5-2. Marker Size and Maximum Distance for Eagle Digital Cameras

Eagle Camera

Marker Size

6 mm (1/4 in.)

12 mm (1/2 in.)

19 mm (3/4 in.)

25 mm (1 in.)

Min 3 Lines

4.2 m

8.0 m

10.7 m

12.1 m

Distance (m)

Min 2 Lines

7.0 m

12.0 m

15.0 m

16.0 m

These are empirical tests taken from an Eagle camera with the 20-35 mm zoom lens, at a capture rate of 60 Hz, 100% brightness, and a threshold =

500.

Note:

Hawk cameras will require markers that are approximately 50 to 100% larger than those listed in

Table 5-2 .

When using 1/2-inch markers, the useful distance for VERY CLEAN data

(Min. 3 lines per centroid) is about 8 meters. Going to a minimum of 2 lines per centroid, takes the usable distance to 12 meters. Going to the 3/4 inch marker gains another 3 meters (to 15 meters). A note of interest is that going to a one-inch marker does not give a big boost in distance.

Other factors come into play, such as the inverse square law about light falling off as we get further from the source. Above 16 meters, the markers are "self extinguishing".

Our experience indicates that if you go beyond about 10 meters in any direction of the capture volume (length or width), it is best to have a second tier of cameras in the middle of the longer dimension. This is a big benefit for multiple person captures as it minimizes editing time needed since you get lots of good solid 3D marker points.

5-23

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Troubleshooting Eagle and Hawk Camera Problems

If Any Cameras Fail to Respond

Motion Analysis sets the default network address in the software to

10.1.1.199. Please note, some computers have multiple network cards installed in them. Please make sure they are labeled so there is no confusion.

If you or your IT department has changed the network address for your system or your cameras, please make note of this for reference as it will save you time in the future.

If you see the error shown in Figure 5-19

, there can be multiple reasons why.

Figure 5-19. Unable to Connect to Cameras Error

The following are some steps to try and fix the problem, starting with the simplest and progressing to the more complex.

1.

Under the

Setup > Cameras

subpanel, verify the Eagle Network

Address in the Eagle Network Address box.

Figure 5-20. Eagle Network Address Box

5-24

If nothing has been changed this should have a network address of

10.1.1.199. Try and connect to cameras again. If this does not fix the issue please move on to the next step.

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

2.

3.

Make sure that the Network cable going from the back of the computer to the EagleHub is securely installed. Motion Analysis uses the on-board network port for the Eagle Network. If you purchased a computer from another source, this may not be how your system is setup. Please take note of this when checking the connections, as it will be useful when talking to Motion Analysis Customer Support staff.

a.

b.

Unplug the network cable from the back of the computer and plug it back in.

Do the same for the connectors on the EagleHub. You should hear an audible “Click” when inserting it back into the jack.

If Windows Updates has been recently run (they may be running in the back ground), there may be a possibility that the Microsoft Windows Firewall was either installed or turned on. This will need to be turned off, as well as any other Firewall software installed on your motion capture computer. Because of the nature of the digital cameras, it is required that the network coming in to the computer, on a particular IP address, is open for streaming data. If your facility requires a Firewall to be installed for their network, it will need to be configured to leave the Eagle Network untouched and open.

You can turn off the Windows Firewall by going to the

Start > Settings

menu in your Windows desktop and then selecting

Control

Panel > Security Center

. This is a feature in Windows XP, Service

Pack 2 and later software.

5-25

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

4.

Double-check to make sure the Network Address that is set in

EVaRT

is the same as the Network Address that is assigned to your Network

Interface Card (NIC). To do this follows these steps:

a.

b.

c.

From your desktop, select

Start > Settings > Control Panel >

Network Connections

. You should have a “Local Area Connection” and possibly an “Eagle On-board Network” (or Hawk if so stated). There may also be “1394 IEEE Connection”, this is for

Fire Wire and can be ignored.

Right-click the appropriate network connection for the Eagle Network and select

Properties

from the drop down menu.

Under the General tab, scroll down and select

Internet Protocol

(TCP/IP)

and click on the

Properties

button.

Figure 5-21. Internet Protocol (TCP/IP) Properties Selection

5-26

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup d.

You should have “Use the following IP address” selected, if not, please make sure you have selected the proper network connection. You may close this window and return to Step 4b. The IP

Address should be 10.1.1.199 and the Subnet Mask should be

255.255.255.0. If either of these is incorrect, please change them.

Figure 5-22. IP Address and Subnet Mask Address

5.

If the software has still not connected to the Camera Network, use the

DOS interface in Windows to Ping the cameras to verify if there isn’t a hardware failure.

a.

To Ping a camera, select

Start > Run

from the Windows desktop.

In the pop up window, type in cmd. This will launch the command prompt.

b.

Type in Ping

10.1.1.201

and press

Enter

. If the request times out (it will try 4 times) try doing the next number,

10.1.1.202 and so on. Your cameras should be set to 10.1.1.201

for camera number 1 and 10.1.1.202 for camera number 2 and so on (unless changed by you or your IT personnel). If the cameras do not respond then you may need to use your Eagle Test Cable to determine the Camera Network address. The Eagle Test cable is the black cable about 1 meter in length, one end plugs into the

Aux port on the back of the camera, the other end has a VGA port, a COMM port and a BNC connector. Follow the steps found in Appendix A-23 through A-24; this will display the Network address of your camera.

5-27

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

The Ping command should return a message similar (but not exactly the same) as follows

:

Pinging 10.1.1.201 with 32 bytes of data:

Reply from 10.1.1.201: bytes=32 time=20ms TTL=128

Reply from 10.1.1.201: bytes=32 time=20ms TTL=128

Reply from 10.1.1.201: bytes=32 time=20ms TTL=128

Reply from 10.1.1.201: bytes=32 time=20ms TTL=128

Ping Statistics for 10.1.1.201:

Packets: Sent=4, Received: 4, Lost=0 (0% loss),

Approximate Round trip times in milliseconds:

Minimum=0ms, Maximum=242ms, Average= 128ms

If you are getting a message that says:

Pinging 10.1.1.201 with 32 bytes of data:

Request timed out.

Request timed out.

Request timed out.

Request timed out.

Ping Statistics for 10.1.1.201:

Packets: Sent=4, Received: 0, Lost=4 (100% loss),

Approximate Round trip times in milliseconds:

Minimum=0ms, Maximum=0ms, Average= 0ms

Then the camera is not responding to Ping requests.

6.

This last step involves trying to determine if there is a camera, cable, or connection that may be causing the system to not identify the cameras. The best way to do this is to unplug all of the cameras from the

EagleHub. Then, plug each camera in, by individually (do not connect any other camera cables to the hub). It does not matter which

RJ45 port you connect the camera network cable. The same applies to the power connector. The only other connection going into the Eagle-

Hub is the cable coming from the back of the Host computer (

EVaRT

tracking computer).

After you have plugged-in a camera, click-on the

Connect to Cameras

button on the Real-Time Dashboard. If this works, unplug this camera and set it off to the side or label it as good. Move on to the next camera doing the same, and so on, each time remembering that there should only be one camera plugged into the EagleHub at a time.

This will help you narrow down if there is a conflict.

After running through each individual camera, it is very useful to power down the EagleHub to help clear out any stored data that may be in its memory. You may also want to take the time to write down IP address for each camera. If you run into a problem with multiple IP

5-28

EVaRT 5.0 User’s Manual Chapter 5: Camera Setup

Socket Error

Function Not

Found In Library

Error

Addresses being the same, this could be the problem. Each camera needs to have it's own independent IP Address. No two addresses can be the same. This makes each camera unique and will help the system identify them.

When connecting to the cameras on the RealTime Dashboard, if you encounter a SOCKET ERROR, you will need to verify the following:

• that the Ethernet connector on the back of the Host Computer is working properly.

• that the Ethernet cable running from the EagleHub (or switch connected to multiple EagleHubs) to the EVaRT Host Computer is connected.

If you get the message ERROR: Function not found in li-

brary

, you have an older version of a library that is not working or is needed. If you do not have the A-D option, check the

C:\Winnt\System32

directory and rename the file

nidaq32.dll

to be

nidaq32.dll.old

. Then close and relaunch

EVaRT.

This applies for Windows 2000 operating system. If you are using a different operating system, you will need to do a search for the file

nidaq32.dll

.

Relationship Between Capture Volume and Marker Size

Note:

The following is for Midas-based cameras only (i.e. Falcon, Cohu,

Pulnix, etc.). Eagle and Hawk cameras set marker size parameters through the

EVaRT

interface.

Listed in

Figure 5-23 are Optimal (Highest Accuracy), Large Volume, and

Extended Volume capture areas for the 6, 8, and 10 Falcon camera systems. Larger volumes require more cameras. Different shaped capture areas are also possible.

At the extremes, volumes will vary with ceiling heights and can vary with optical conditions including external lighting.

5-29

Chapter 5: Camera Setup EVaRT 5.0 User’s Manual

Figure 5-23. Guidelines for Selecting Marker Size for Falcon Cameras

Highest

Accuracy

Large

Volume

Extended

Volume

Marker Size

Eagle = 3/8-inch (9mm)

Hawk = 1/2-inch (12.5mm)

Falcon = 3/4-inch (19mm)

Eagle = 1/2-inch (12.5mm)

Hawk = 3/4-inch (19mm)

Falcon = 1-inch (25mm)

Eagle = 3/4-inch (9mm)

Hawk = 1-inch (12.5mm)

Falcon = 1 1/2-inch (19mm)

Normal Capture Volume meters # Cameras

2.5 x 2.5

2.5 x 3.5

2.5 x 5

3.5 x 3.5

3.5 x 6

3.5 x 8

5 x 5

5 x 8

5 x 11

6

8

10

6

8

10

6

8

10

1/8'' and 1/4'' markers are also available for smaller volumes such as face, hand, or foot capture volumes.

5-30

Chapter 6

The EVaRT User Interface

Topic

Getting Acquainted With the User Interface

PRJ Files

Viewing Sample Data

Digital Video Option (EVaDV Software)

Real Time Dashboard

Join Virtual

Post Process Dashboard

Post Process Toolbar

Zooming, Rotating, and Translating

6-35

6-39

Selecting Markers, Virtual Markers, Linkages, and Segments 6-41

Time Code 6-41

Page

6-1

6-13

6-14

6-23

6-23

6-33

6-34

Getting Acquainted With the User Interface

Before using

EVaRT

it is necessary to become familiar with the interface and the names of the tools and controls that will be used throughout this manual. The major components are as follows:

5.

6.

7.

8.

1.

2.

3.

4.

9.

10.

The Graphics Panes in the center of the screen.

The Menu Bar in the upper left corner.

The Directory List below the Menu Bar.

The Mode Buttons along the top left of the screen.

The Sub-Panel Buttons along the top right of the screen.

The Sub-Panels on the right side of the screen.

The Real Time Dashboard along the bottom of the screen.

The Post Processing Dashboard replaces the Real Time Dashboard.

while the program is in Post Processing mode.

The Status Bar Messages in the lower left corner.

The Information Center in the lower right corner.

6-1

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

An image of the interface is shown in

Figure 6-1

. Note that multiple 3D

Displays can be rendered simultaneously.

Figure 6-1. EVaRT Interface in Real Time Mode

Menu Bar

Directory List

Mode Buttons Sub-Panel Buttons

Graphic Panes

Status Bar Messages

Real Time Dashboard

Panels

Information Center

(Cells 1 through 5)

6-2

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Menu Bar

Figure 6-2. Menu Bar

The Menu Bar selects the primary menu items for

EVaRT

functionality.

These include file management, layout control, data views, tools, and help.

File Menu

Load Project...

Loads a project file (.prj) from the current working directory.

Save Project

Saves a project file (.prj) to the current working directory.

Save Project As...

Provides a method to save the current project (PRJ) file under a different name.

6-3

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Select Raw Video File (Live)...

Loads the VC (Raw Video) files from a capture session.

Load Raw Video Files (Post)...

This loads the raw data into the post process screen and is used for the Refine Tracks feature (see “Refine Tracks” on page 12-3 ). The data is the centroids calculated during the load.

Load Tracks...

Loads the TRB or TRC (Tracks) files from a capture session.

Save Tracks

Allows the user to save the current Tracks (TRB or TRC) file with the current filename.

Save Tracks As...

Provides a method to save the current Tracks (TRB or TRC) file under a different name.This function will only allow you to change the file name, not the file type. To change the file type, you must use the

Export.xxx

File...

menu item.

Trim Capture W/Options...

Provides frame and marker options when saving Tracks (TRB or TRC) files. The

Trim Capture

feature allows you to specify which directories the new files will be saved to, as well as frame and marker management options. Refer to

Figure 6-3 . The software remembers two folder names,

the current folder, and the export folder.

6-4

EVaRT 5.0 User’s Manual

Figure 6-3. Trim Capture w/Options Interface

Chapter 6: The EVaRT User Interface

Export HTR File...

Exports a file in HTR format, which is differently organized than the

HTR2 format.

Export C3D File...

Exports marker positions and analog data in an open sourced file format viewable by many different software packages.

Export CRC File...

CRC (Centroid Row Column) data are the 2D data points in ASCII text format. Usable by advanced users who want to reconstruct 3D positions of markers using their own software, in post process mode only.

Export Forces File...

Exports ASCII files containing forceplate data. This uses your current forcepla.cal file and converts the raw forceplate data into calibrated forces. The units used are Newtons and Newton-Meters.

Convert .anc File...

Converts binary formatted Analog data (.anb) into an ASCII viewable format (.anc).

Skeleton Definitions

Import Skeleton (.mod) File...provides a method for bringing a new MOD file into the current PRJ file.

MOD files are used for Deep Solver applications. The name of the MOD file will match the marker set name as defined in the project.

6-5

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Layouts Menu

Data Views Menu

Export Skeleton (.mod) File...provides a method for saving a MOD file for use in a different PRJ file.

Create Skeleton from HTR File...brings in the skeleton definitions from the animation package.

Load Marker Set...

Can be used to load a previous marker set into a new or newly calibrated project.

Load Calibration...

Can be used to load a new set of calibration information into a project.

Load Analog Setup...

Can be used to load a new set of analog setup information (forceplates,

EMG, and others), into the current project.

Load .ini Preferences...

If the end user changes the look or feel of the software (different color scheme, sounds etc.) then when they move to a new version of the software (i.e. new release), they can load up their preferred 'look' and have that as the default start. You need to use save the .ini preferences (see below) to make this .ini file.

Save .ini Preferences...

If the end user changes the look or feel of the software (different color scheme, sounds etc.) then they must save their changes using this tool.

Creates an INI file that can be loaded onto different computers or on new versions of the

EVaRT

software. See

Load .ini Preferences...

. above.

Exit

Allows you to exit the

EVaRT

software. Make sure you have saved the current project and tracks files.

The items in this menu are generally self explanatory.

The

EVaRT

interface can accommodate up to four simultaneously open data views and can be resized with the mouse to fit the panes however you desire. You can view combinations of six different graphical panes, which include:

Color Video F1

Shows the live action in a capture volume or a replay of an AVI file. You will need to have a DV (Digital Video) camera connected to the Host computer’s 1394 Firewire port to see the live video in this window.

6-6

EVaRT 5.0 User’s Manual

Tools Menu

Chapter 6: The EVaRT User Interface

2-D Display F2

Shows a 2D image of the markers and their centroids.

3-D Display F3

Displays a moving 3D stick figure showing named markers, virtual markers, linkages and/or a skeleton.

XYZ Graphs F4

Displays graphs depicting the marker’s positions in each frame.

Analog Display F5

Displays analog data graphs representing the force plate’s output.

Skeleton Graphs F6

Displays hierarchical translations and rotations of skeletal segments.

Analysis Graphs F7

Calculates and displays angles between markers, distances between markers, and position, velocity, and acceleration of a marker, or groups of selected markers. Results can be saved as ASCII Time Series (

.ts

) files. For more information, refer to Appendix K, Analysis .

Graphs Only (Ctrl+G)

Hides the side panels and Post Process/Realtime Dashboard to maximize the graphic panes.

Quickfiles

The Quickfiles display provides a window within the

EVaRT

interface to navigate the file structures and contents of the directories used for motion capture data sets. Specific file types can be chosen for display, which allows the user to filter out unwanted file types.

The Tools menu gives access to various functions within the

EVaRT

software. Many of these functions are also available in other sections of the user interface.

Calibration Setup

The Calibration Setup defines the parameters for the system calibration

(e.g. capture volume up-axis, calibration units, etc.). For complete information, refer to “Details... Button (Calibration Settings Window Tabs)” on page 8-4 .

6-7

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Edit Thresholds

This activates the threshold slider used to block out excessive noise in the

2D camera view. For more information, refer to “Adjusting Thresholds” on page 7-9 .

Time Lines

Shows the time line of the data for each marker, indicating any breaks in the stream of data. More information can be found in “Time Lines” on page 10-26 .

Batch Processing Options

This provides an overview of the tracking, identifying, and solving configuration in

EVaRT

. For complete information, refer to “Batch Processing” on page 9-13 .

Virtual Marker Definitions

This sets the definition markers that are used to support a particular virtual marker. For complete information regarding Virtual Markers, refer to

“Virtual Markers” on page 11-6 .

Pose ID Options

This is used to automatically identify markers based on a Model Pose that you create when you make a template. The result is that if you use the same marker set repeatedly, you will not have to ID the new person each time the marker set is used. This is the same function as using the

New

Subject...

button that is found on the Real Time dashboard. For complete

information, refer to “New Subject Button” on page 6-24 .

6-8

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Forceplate Forces

Selecting this feature will display the forceplate measurements in numerical values. This works when you are live and connected to the cameras, or when you are simulating Real Time from VC files and you are Post Processing mode.

Figure 6-4. Forceplate Forces

Figure 6-5. Colors Form

Colors...

The Colors form allows you to choose RGB colors for the markers, Background, Foreground, Real-Time Floor, and Post Floor for your project.

Refer to

Figure 6-5 .

6-9

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Figure 6-6.

To change the color for any item, just click on the colored circle and a color palette window opens up. This lets you choose the color blend for that marker. The colors are stored in the

EVaRT50.ini

initialization file that is loaded from the

EVaRT

launch folder. They can also be stored in your personalized .ini file.

Misc

Replace Loaded Analog Channel Names

—Occasionally, there are situations where an incorrect analog channel name can occur. Examples of this can be seen when analysis software (like OrthoTrak) requires specific muscle names for the analysis. In these cases the ability to go back and rename the problem analog channel is required.

To Rename Analog Channels, follow these steps:

1.

2.

3.

4.

5.

6.

7.

Load a Project file.

Load a Tracks (.trb/.trc) file.

Select

Data Views > Analog Display.

In the existing project file, go to

Setup > Analog

, and change the name of the analog channel you wish to modify.

Save the project. You may want to save it as a different project name.

Go to the Tools menu and select

Misc > Replace Loaded Analog

Channel Names

.

To see the replaced name, reload the project file you are working with. This will show the changed name in the Analog display sidebar.

Note:

To save this change to the analog file, you must follow the next steps exactly.

8.

Select

File > Trim Capture W/Options

.

6-10

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Help Menu

9.

10.

11.

Under the Trim Capture W/Options, choose

Save Selected Frames

.

Make sure you have highlighted all the frames that you want by using the middle mouse button to highlight an area in the Post Processing window, or by using the Select All Frames button in the lower right corner.

Press the

Export Trimmed Capture

button, and type in the filename you want or keep the current one.

If you have more files that need to have the analog channel names replaced, you will need to repeat steps 2 through 10.

Record

—The record function under the Tools menu item starts the recording of a data capture. It is the same function as the Record button found on the

Motion Capture > Output

sub-panel and the F12 function key.

The Help menu provides information about the software, along with shortcuts, interface sub-panel information, and the on-line manual.

Directory List

The Directory List allows you to easily navigate through frequently used directories, making it easier to address and manage motion capture sessions. Refer to

Figure 6-7 .

Figure 6-7. Directory List

Directory List

6-11

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Mode Panel

Buttons

These buttons are arranged to guide you through a motion capture session in a phase-oriented order. Refer to

Figure 6-8

. The first three mode buttons (Setup, Calibration, and Motion Capture) activate Real-Time mode and present you the necessary tools to successfully capture motion data.

Figure 6-8. Mode Panel Buttons

Mode Panel Buttons

Sub-Panel

Buttons

Status Bar

Messages

Real Time

Dashboard

The fourth button, Post Process, activates Post Process mode and transforms

EVaRT

into a tracked data editing tool.

The final two buttons, Model Edit and User Apps, are mode-less function buttons that present various tools without switching the program between the Real-Time mode and Post Process mode. Model Edit is used to define markers to create linkages.

These buttons give you access to the various tools specific to the different phases of the motion capture session.

This feature, located in the lower left corner of the user interface, provides the status and confirmation of the software in its current processing state.

Refer to

“Real Time Dashboard” on page 6-23 .

6-12

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Information

Center

The Information Center gives the following information for the current motion capture project (from left to right):

Cell 1: Number of frames in the current data set

Cell 2: Frame rate in frames per second

Cell 3: Up axis (e.g. Y up, Z up)

Cell 4: Calibration units (e.g. mm)

Cell5: Analog sample rate (samples/sec)

If you leave the mouse pointer over the message, its definition will pop up.

PRJ Files

Important

PRJ files, or project files, are the main files used in gathering all

EVaRT

motion capture data. Every motion capture session must have a project file containing all system settings, equipment parameters, and other information related to the project. This file contains both equipment parameters common to many different setups and calibration values unique to one particular session. Among the data found in a project file are:

• the camera setup

• the marker set

• calibration setup and results

• linkages between markers

SkB (Skeleton Builder) segment definitions, coordinate systems, and hierarchies (optional)—refer to Chapter 12, Skeleton Types

MoCap Solver segment definitions, joint types, and hierarchies

(optional)

• camera type and parameters

• tracking parameters

In most cases, you will begin a session by loading an existing project file, editing it as necessary, and saving it in the directory where the motion data is to be saved. Any time you calibrate the system or edit project parameters, you should save the project file to disk to retain the new information.

Note:

Multiple PRJ files should be saved in the same file directory with care.

Project files contain ASCII data and it may be useful to view them using any text editor, however, you should not edit them in a text editor as that can result in a corrupt file.

6-13

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Viewing Sample Data

To become familiar with

EVaRT

, we will start by replaying some sample data. This is done by loading a sample project found in the Samples directory.

Sample Data Set

1.

From the Menu Bar, select

File > Load Project...

2.

Navigate to:

Program Files\Motion Analysis\EVaRT50\Samples\GolfSwings with Temper.

3.

Double-click on

Body_Club_Merged.prj

to load the project.

Having loaded the sample project, we will now load the related data files.

1.

2.

Select

Raw Files

on the Real Time Dashboard.

Click on

GolfTemper1.vc1

.

Click the

Run

button on the Real Time Dashboard. At this point the action on the screen is a simulation of a live motion capture session.

EVaRT

is processing the data from the stored raw video file, color video file, and analog file generated by the force plates. If this were an actual real-time capture session, the action on the screen would be similar, but the data would be coming directly from the cameras and force plates.

Note:

Loading a VC file or TRB file will automatically load any saved ANB files captured during this session of this particular project. The ANB files are comprised of analog forceplate data in this example.

Having loaded all of the related data files, we can now exercise all of the six different Graphics Panes available to us. We will now look at four simultaneously.

1.

2.

3.

From the Menu Bar, select

Layouts > 4 Panes

.

Left-click on the empty lower-left pane. This action will select this pane.

Press

F1

on the keyboard or choose

View > Color Video

.

Note:

If you are interested in the Color Video option, contact your Motion

Analysis sales representative.

4.

5.

6.

7.

Left-click on the empty upper-right pane.

Press

F2

on the keyboard or choose

View > 2D Display

.

Left-click on the empty lower right pane.

Press

F5

on the keyboard or choose

View > Analog Display

.

The displays can be controlled by hand if you click

Pause

, click on the

FIFO slider on the Real Time Dashboard, and then drag from side to side.

The “First In First Out” FIFO slider can only manipulate the 256 frames of data that are currently stored in the FIFO memory space (but not all of the data in the data set).

Figure 6-9

should be similar to what you see on your screen.

6-14

EVaRT 5.0 User’s Manual

Figure 6-9. Viewing the Pre-Recorded Data

Chapter 6: The EVaRT User Interface

EVaRT

displays data somewhat differently when in Post Process mode.

For instance, the analog displays become static graphs rather than having the oscilloscope style seen during collection and replays of raw data. The

2D Display and Skeleton Graphs become entirely unavailable but the

XYZ Graphs become available. 3D stick figure images can be rendered for all of the data set rather than just the 256 frames available in Real-

Time mode.

Post Process mode allows you to edit the tracked data generated during a motion capture session. Editing can be performed upon groups of markers or one marker at a time.

4.

5.

6.

7.

1.

2.

3.

From the Menu Bar, select

File > Load Tracks File...

Double-click on

GolfTemper1.trb

.

Leave the 3D figure currently in the top pane in the 3D Display.

From the Menu Bar, select

Layouts > 2 Panes: Top/Bottom

.

Left-click on the bottom pane. This action will select this pane.

Press

F4

on the keyboard or choose

View > XYZ Graphs

.

Select marker 15 on the Markerset panel (right side).

6-15

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

The two Graphics Panes used, often simultaneously, during a Post Process editing session are the 3D Display and the XYZ Graphs shown here. Notice that the Post Process Dashboard has replaced the Real Time Dashboard. The data shown in the XYZ Graphs represents the X, Y, and Z coordinates of the selected markers throughout the capture period.

Figure 6-10. Viewing Tracked Data in Post Process Mode

6-16

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Pop-Up Menus

The Graphics Panes have viewing options and associated tools that can be accessed through pop-up menus. In all cases, the pop-up menus are activated with a click of the right mouse button while the pointer is in the display region.

3D Display Pop-Up

Menu

1.

2.

3.

4.

From the Menu Bar, select

Layouts > 1 Pane

.

From the Mode Buttons, left-click on the Post Process button in order to be in the Post Process mode.

If the 3D Display is not visible, press

F3

on the keyboard or choose

View > 3D Display

from the Menu Bar.

With the right mouse button, click on the 3D Display.

The 3D Display pop-up menu and descriptions of the tools are shown in

Figure 6-11

. These options are recorded in your INI file and are reloaded when you launch

EVaRT

.

Figure 6-11. Post Process 3D Display With Pop-Up Menu Items

Show options—cascading menu

Quick ID the markers sequentially

Marker ID the selected marker

ID marker(s) based on current template

Rectify marker(s) over the selected frame range

Hide selected marker(s) from view

Show selected marker(s)

Make selected marker(s) unnamed

Create Template

Cut data in selected frames from the selected marker(s)

Cut data outside of the selected frames from the selected marker(s)

Exchange data between two markers over the selected frames

Smooth selected marker(s) over the selected frames

Join selected marker(s) over selected frames using cubic splines

Join selected marker(s) over selected frames using linear interpolation

Create a temporary virtual marker to fill in missing marker data

Undo last action

Search data set for spikes and/or gaps as defined in the Options panel

6-17

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

To see the 3D Display options:

1.

2.

Choose the cascading

Show

item and another pop-up list will appear.

Several of the view options will have check marks next to them indicating they are active. All of the Show items in the Show list are considered User Preferences and get stored to the

EVaRT.ini

file when you exit the program.

Choose items from the Show options menu and see the effects. These options are saved in your INI file settings and are reloaded when you launch

EVaRT

.

Figure 6-12. 3D Display Right-Click Pop-Up View Options

Displays all markers in 3D display; Options... for length

Displays all links between markers in the 3D Display

Displays the prj file name over the marker cloud

Displays the ID numbers for all markers

Displays the marker names (over the marker)

Displays the marker trajectories (in PP mode only)

Displays the virtual markers

Displays the skeleton segments when they are defined

Displays a selected skin

Displays the RGB/XYZ orientation for each bone

Displays the force vectors off the forceplates

Displays the marker locations as defined in the T-pose

Displays the model pose as defined in Create Template

Displays the motion capture cameras in the 3D Display

Displays which cameras can see the selected marker(s)

Displays the camera view for the selected camera(s)

Displays the virtual floor in the 3D Display

Displays the capture volume; Calibration Details...

Displays the Digital Video layered with the 3D Display

Centers the display on the selected marker

3D Display center follows the selected marker

Mirrors marker set from left to right hand coordinates

Normal display, adjust with mouse

Show front view

Show side view

Show top view

Rotate around the capture area by pressing Play

Toggle display to view from selected markers

6-18

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

2D Display Pop-Up

Menu

The 2D Display renders raw camera data as blobs and/or marker centroids. You can choose to see the centroids either with or without lens correction. To see the marker data as viewed from any one of the cameras or multiple cameras simultaneously,

1.

2.

Press

F2

on the keyboard or choose

View > 2D Display

from the

Menu Bar.

Choose one or more cameras with

Ctrl

+ click or

Shift

+ click on the green camera buttons on the Real Time Dashboard or press

All On

.

To see the 2D Display options, right-click on the camera view 2D Display.

Figure 6-13. 2D Display with Pop-Up View Options

Deletes selected mask

Deletes all masks in the 2D display for the selected camera

Draws a mask around all items in the field of view

Allows the capture of raw data with any masks in 2D display

Toggle black raw data blobs

Toggle red, raw centroid crosses

Toggle centroids corrected from lens distortion

Toggle marker names

Toggle marker numbers

Toggle outline of volume floor

Leave smeared paths of markers

6-19

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Analog Display

Pop-Up Menu

For users collecting analog data from force plates, an Analog Display provides graphs of output from up to 64 analog channels. You can view any combination of channels at the same time. As a convenience, the Analog

Display allows you to resize the label panel on the left side of the screen to accommodate long channel names. To open the Analog Display and modify the number of visible channels:

2.

3.

4.

5.

6.

1.

Press

F5

on the keyboard or choose

View > Analog Display

from the

Menu Bar.

With the right mouse, click on the Analog Display.

From the pop-up menu, choose

Visible Channels...

Left-click on any one of the check marks in the Visible column.

Press

Shift

+ click in the

Visible

column to toggle multiple channels.

Click directly on the

Visible

header cell to toggle all of the channels at once.

Figure 6-14. Analog Display With Pop-Up View Options and Channels Table

Shows the Channels Table

Toggle all channels

Corrects for time match problems between analog and video data

Channels Table

6-20

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

XYZ Graphs Pop-

Up Menu

Post Process mode allows you to edit tracked data. The XYZ Graphs displays the positions of each marker in each frame. It also lets you select and edit those markers in any frame. A complete discussion of editing tracked data can be found in Chapter 10, Post Processing Panel . To see the

XYZ Graphs and the pop-up menu of tools and view options:

1.

2.

Press

F4

on the keyboard or choose

View > XYZ Graphs

.

When in Post Process mode, right-click on the XYZ Graphs

.

Figure 6-15. Post Process XYZ Graphs With Pop-Up View Options and Tools

Zoom into the current frame range

Zoom out from the current frame range

Reset the amplitude display

Auto scale to visible channels

Applies uniform scale to all three X, Y, and Z panels

Select all frames in the data set

Show Residuals and Cameras plots

Quick ID the markers sequentially

Marker ID the selected marker

ID marker(s) based on current template

Rectify marker(s) over the selected frame range

Hide selected marker(s) from view

Show selected marker(s)

Make selected marker(s) unnamed

Create Template

Cut data in selected frames from the selected marker(s)

Cut data outside of the selected frames from the selected marker(s)

Exchange data between two markers over the selected frames

Smooth selected marker(s) over the selected frames

Join selected marker(s) over selected frames using cubic splines

Join selected marker(s) over selected frames using linear interpolation

Create a temporary virtual marker to fill in missing marker data

Undo last action

Search for spikes and/or gaps; as defined in the Options panel

6-21

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Skeleton Graphs

Pop-Up Menu

The Skeleton Graphs pop-up menu offers a single item menu which opens

the Skeleton Graphs Control panel as shown in Figure 6-16 . This control

panel allows you to control the segments that are shown in the graphs and control the chart style and colors of the display. You must have a skeleton defined with either

Calcium/Si

or

SkB

(Skeleton Builder) for this to work. A skeleton defines joint centers and segments according to the rules of the modeling. The

Calcium/Si

software is used to generate Solver skeletons.

EVa 6.0

or

7.0

can be used to define

SkB

skeletons.

Figure 6-16. Skeleton HTR Graphs Pop-Up Menu—Control Panel

HTR Order (Euler

Angle Order)

This sets the rotation sequence, with the first letter being the axis of the angle (flexion/extension point).

Figure 6-17. HTR Order (Euler Angle Order)

6-22

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Digital Video Option (EVaDV Software)

The color Digital Video option allows you to record a time-matched Reference Video along with your motion capture trial on a separate computer.

With this option, you will record a time-matched color video AVI file with the same trial name in your motion capture folder. A separate computer is used in order to not burden your

EVaRT

Host computer, which is an issue if your computer is too slow for the number of markers being tracked. For single person captures, you may connect the DV Camera directly to the

EVaRT

Host computer. In this case, the

EVaDV

software is not needed. It is built into the

EVaRT

software. You can run

EVaDV

on one or more computers and then capture multiple AVI files (multiple views). They will all have the same AVI file name. You may experience a small delay in frames from the

EVaRT

software and the

EVaDV

software when capturing. The Color Video display has a pop-up menu with one item, Adjust Frame Offset. This allows for time-matching data streams.

Note:

The

EVaDV

software option is not to be confused with the AVI function in the

Motion Capture > Output

sub-panel. This function creates an AVI file within the

EVaRT

project file.

Real Time Dashboard

The Real Time Dashboard is available when

EVaRT

is in Real Time mode as opposed to Post Process mode. When you are capturing data in realtime, this dashboard provides the controls to manage a motion capture session. It also supports the replay and tracking of previously recorded data with a simulation of real-time from the raw VC camera files.

Note:

To help you distinguish between the two modes (Real Time and Post Process), the dashboard and the floor color changes.

Figure 6-18. Real Time Dashboard

Configure

Motion Capture

Connect to Cameras

Current Frame #

Time Code

Reset IDs

Join Virtual Camera Buttons

New Subject

FIFO Slider

6-23

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

EVaRT

Dashboard

Camera Button

Colors

The Real Time Dashboard camera buttons inform you of the following:

Green

Light green

Yellow

Light yellow

White

Dark grey

Camera is completely calibrated

Camera is completely calibrated and is selected

Camera has undergone seed calibration but not wand calibration

Camera has undergone seed calibration and is selected

Camera is not calibrated

Camera is inactive—A right mouse click on the camera number will enable and disable that camera.

Note

—Camera #1 must remain enabled when you collect data to make sure you have a selectable VC1 file.

Tracking

The Tracking check box triangulates (tracks) the markers from frame to frame. You might want to uncheck this if your computer is not fast enough to calculate the marker coordinates or if your system is not calibrated for any reason. You can still collect raw VC files and track them at a later time. Collecting raw VC files is the highest priority thread in the motion capture Record mode to ensure that you do not lose your raw data.

Identifying

The Identifying check box identifies and names the tracked markers according to the current template. If you do not have a template, it is best to disable this function to keep the software from attempting to ID the data.

Skeleton

The Skeleton check box is set to calculate the skeleton using the currently active skeleton model.

Reset IDs Button

The Reset IDs button forces the current template to be used for that specific frame. It is used when a marker is misidentified. Press this if you see markers that are incorrectly identified. If this fails to fix the problem, you may need to create a new template or adjust the marker set to be less symmetric. After pressing the

New Subject...

button and with the Pose ID window open, the Reset ID button will use the current marker pose to identify the markers.

New Subject

Button

This feature allows the software to automatically identify markers based on a Model Pose that you create when you make a template. The result is that if you use the same marker set repeatedly, you will not have to ID the new person each time the marker set is used. The marker identification is automatic and instant, saving you time.

The Model Pose has its own kind of generalized template that is used to automatically identify a new person when they appear in the field of view.

It saves the steps of using the Quick ID feature to identify a new person in order to make a template for them. The Auto ID feature works in the

6-24

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Note:

Real-Time mode when you are connected to cameras or it works when you are tracking the data from VC files after the collection.

The current template is size specific, so a new person or a new arrangement in the markers will not generally work for automatically identifying the markers.

When you click

Update Template

, it also updates the Model Pose.

To use the

New Subject

button, use the following procedure:

1.

2.

Get a Range-of-Motion Trial.

a.

Get a good range of motion trial for your current tracks in Post

Process, Quick ID, and edit so there are no mistakes or marker switches. The data does not have to be highly complex, but it should represent the minimum and maximum stretching for all limbs. Jumping Jacks are a good example of the kind of dynamic motion that has worked well and does not obscure the markers or require editing. For simple walking motion, a single walking trial will suffice.

b.

Select one frame that represents a standard or neutral pose position. This can be with the arms down or the arms out, feet apart or together, but where no markers will be hidden. You will want it to be a standard position that the next person will be able to repeat quickly and simply. Have them face a certain direction that will also be easily repeatable for the next person. (along the +X axis for example).

Create a Template.

a.

Select

Post Process > Create Template.

b.

c.

Select Body Template and check the box

Include current frame as the Model Pose

and select the correct range of frames where you have good data.

Save your project file which now has a new feature called the

Model Pose stored in it. You may want to use the word “Pose" in the project file name to distinguish it from earlier versions without the pose, but that is not necessary. The normal template will also work for this person.

Note:

You can choose to see the Model Pose in your current project by rightclicking in the 3D display and then selecting

Show > Show Model Pose

.

6-25

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Figure 6-19. Create Template Window—Include Current Frame as the Model Pose

Getting Auto ID to

Work: Tuning and

Updating the

Template

If the person who was used to create the template moves some markers, or if a new person comes out with the same marker configuration (but in slightly different locations), you will want to update the template to the new marker locations. This will make sure that your RealTime tracking and the Template ID and Template Rectify functions in Post Process will be at an optimum performance level.

Activate the Motion Capture panel and then select

Connect to Cameras

(or select the VC files) and then select

New Subject...

. This will bring up the Pose ID dialog box (below) and the Model Pose stick figure appears in the 3D panel.

Figure 6-20. New Subject... Interface

6-26

At this point the old Template ID feature is not working, but instead the generalized automatic Pose ID feature is looking to identify the unnamed

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

markers. As soon as the ID is recognized and catches, you will see two stick figures; the static one from the Model Pose and one that is the newly

ID-ed person that is moving. If the person is not ID-ed right away, have them face the same direction and assume the same general posture as seen in the Model Pose. The Auto ID feature works as long as the person is facing within about 45

°

of where the Model Pose was recorded.

The status display in the lower left tells you how fast the ID process took.

A small number is a fast ID, a bigger number would be slower, but still working. Using the Reset IDs button on the lower right will force the software back to the Pose ID if something gets switched and you want to correct it.

Figure 6-21. Pose ID in Message Center and Reset IDs Button

Pose ID in Message Center

New Subject Button Reset IDs Button

Updating the Template also updates the Model Pose, so before clicking the

Update Template

button, you should again get your new person into something close to the Pose position. The changes can be saved in your project file if you want.

After you Update the Template, the template is then re-sized to the new person's limb lengths and marker placements. Note that the changes in the lengths, as recorded in the range of motion trial is still saved, so that you will not need to do another range of motion. The new template should work well for many sizes and marker adjustments using the same marker set.

Recommended

Procedure

1.

2.

3.

Create your own library project file for the marker set. This library contains your markers set with the template created from your range of motion TRB files, and your Pose ID.

When creating the library project file, start with the range of motion

TRB type file in the neutral position, facing +X, arms down, feet slightly apart to show all the markers. That would be frame 1 to make the Pose ID easy to find. When you create the Template for your library file, you would have Frame 1 selected as the current frame and you would check

Include Current Frame as Model Pose

for frame number 1. You should only need to do this once per marker set.

When the New Subject comes into the volume, they should be standing in the Pose position facing the same +X

1

and you select

New

Subject...

The PoseID will show and you should then select

Pause

.

The stick figure should snap to, looking like the Model Pose.

1.Facing +Z or any other direction also works as long as Pose ID is used in the same orientation

6-27

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Using the Pose

ID Feature

Get a Range of

Motion Trial and

Make a Template

4.

5.

Update the template, then select

Pause.

Check that the ID is correct and that the position is similar to the Pose. You can use the FIFO slider to make adjustments. Save the project file and then select

Run

.

Select

Update Template

again.

This feature allows the software to automatically identify markers based on a Model Pose that you create when you make a template. The result is that if you use the same marker set repeatedly, you will not have to ID the new person each time the marker set is used. The marker identification is automatic and instant, saving you time.

The Model Pose has its own kind of generalized template that is used to automatically identify a new person when they appear in the field of view.

This generalized template depends on the person facing the same direction as the stored Pose ID and having the markers in the same general locations with respect to each other. You save the Pose ID from one frame of data and is saved in your project file. It saves you the steps of using the

"Quick ID" feature to identify a new person in order to make a template for them. The Auto ID feature works in the RealTime mode when you are connected to cameras or it works when you are tracking the data from VC files after the collection.

The following are basic steps on how to use the Pose ID function.

1.

2.

3.

You will first want to obtain a good range of motion trial and set it as your current tracks in Post Process.

You will then need to Quick ID and edit the trial, if needed, so there are no mistakes or marker switches. It does not have to be overly complex, but it should represent the minimum and maximum stretching for all limbs. Jumping jacks is a good example of the kind of dynamic motion that has worked well and does not obscure the markers or require editing. For simple walking motion, a single walking trial will be sufficient.

Select one frame that represents a somewhat standard or neutral pose position. This can be with the arms down or the arms out, feet apart or together, but where no markers will be hidden. You want it to be a

"standard" position that the next person will be able to repeat quickly and simply. Have the subject face a certain direction that will also be easily repeatable for the next person. (along the +X axis for example).

6-28

EVaRT 5.0 User’s Manual

Figure 6-22. Create Template Interface

Chapter 6: The EVaRT User Interface

Create a Template

Getting Pose ID to

Work: Tuning and

Updating the

Template

1.

2.

3.

4.

Select

Post Process > Create Template

.

Select

Body Template

.

Activate the

Include current frame as the Model Pose

check-box and select the correct range of frames where you have good data.

Save your project file which now has a new feature called Pose stored in it. You may want to use the word Pose in the project file name to distinguish it from earlier versions without the pose, but that is not necessary. The normal template will work for this person.

If the person who was used to create the template moves some markers or if a new person comes out with the same marker configuration (but in slightly different locations), you will want to update the template to the new marker locations. This will make sure that your RealTime tracking and the Template ID and Template Rectify functions in Post Process will be optimum. The steps are as follows:

1.

Select

Motion Capture > Connect to Cameras

(or select VC files) and then select

New Subject…

. This will bring up the Pose ID dialog box and the Model Pose stick figure appears in the 3D Display.

6-29

Chapter 6: The EVaRT User Interface

Figure 6-23. Pose ID Dialog Box

EVaRT 5.0 User’s Manual

Figure 6-24. Status Display

At this point the old Template ID feature is not working, but instead the generalized automatic Pose ID feature is looking to identify the unnamed markers. As soon as the ID is recognized and catches, you will see two stick figures, the static one from the Model Pose and one that is the newly ID-ed person that is moving. If the person is not IDed right away, have them face the same direction and assume the same general posture as seen in the Model Pose. It has been found that the Auto ID feature works as long as the person is facing within about

±

45 degrees of where the Model Pose was recorded.

2.

The status display in the lower left tells you how fast the ID process took. A small number is a fast ID, a bigger number would be slower, but still working. Using the Reset ID button on the lower right will force the software back to the Pose ID if something gets switched and you want to correct it.

3.

4.

Updating the Template also updates the Model Pose, so before clicking the Update Template, you should again get your new person into something close to the Pose position. The changes can be saved in your project file.

After you have updated the template, the template is re-sized to the new person’s limb lengths and marker placements, but the changes in the lengths, as recorded in the range of motion trial, is still kept. So you do not need to do another range of motion and the new template should work well for many sizes and marker adjustments using the same marker set.

6-30

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Raw Video

Button

The Raw Video button will tell the program to simulate a live motion capture session from previously captured Video Camera (VC) files. It also allows you to tack and record to TRB or TRC trials for which you have raw

VC files.

The Connect Cameras button will activate all the cameras used in a motion capture session.

Connect

Cameras Button

Run Button

The

Run

button will start the streaming of live camera data or start the simulation of a motion capture session from existing raw VC files.

The

Run

button has the following functions:

1.

2.

3.

4.

If you are connected to the cameras, it starts the data steaming from the cameras. You are able to record the Raw Video VC files as set in the motion capture Output sub-panel. Check your 2D views to be sure the cameras, masks, and thresholds are all set properly.

If

Enable Tracking

is checked, you will see the marker data appear in the 3D Display. This requires that the system has been calibrated. You are then able to record VC files, and TRC or TRB files.

If

Enable Identifying

is checked, you will see the colored markers and the stick figures in the 3D Display. This requires that the system is calibrated and a template is defined and operating. You are then able to record VC files, and TRB or TRC files.

Tracking from Raw Video files—If you are not connected to the cameras, you have the full range of option 2 or option 3 above from your previously collected Raw Video VC file.

6-31

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Time Code

Counter

The Time Code Counter displays the frame number in HH:MM:SS:Frame

(hour, minute, second, frame) format. If you have the optional Time Code

Reader card installed in your computer, this displays the current Time

Code value when you are in the Motion Capture mode and connected to the cameras.

Frame Counter

The Frame Counter displays a count of the total number of frames in the data set.

Camera Buttons

You can select each camera by clicking on it’s respective numbered button that is listed across the Real Time Dashboard. Clicking on a camera button will either activate or de-activate that camera for setup features.

Right-Click Camera

Buttons

Right-clicking on any of the camera buttons will open a function menu with various commands for that specific camera. The menu and a descrip-

tion of each command is shown in Figure 6-25 .

Figure 6-25. Camera Buttons Right-Click Menu

Enables the selected camera to capture data (if disabled)

Disables the selected camera from capturing data

Enables the camera to capture and display data in Real Time

Disables the camera from data collection in Real Time only

Camera numbers are sorted in a counter-clockwise order

Camera numbers are sorted by IP address starting from lowest

Deletes the selected camera from the project file

6-32

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Join Virtual

Join Virtual is an extremely powerful editing tool used to fill gaps in marker data with simulated data based on the relationship (positional interpolation) with other markers on or near the particular problem segment. This positional interpolation is defined by Virtual Markers.

Figure 6-26. Join Virtual Check Box

The concept behind the Join Virtual and the Virtual Marker definitions are the same and are much more stable and more useful than the classic Rigid

Body data filling mechanisms. The reason is that you get to choose two sets of three markers, in decreasing importance, that determine the replacement data. These three markers are:

1.

2.

3.

the Origin Marker the Long Axis (Y) Marker the Plane (XY) Marker

The two sets of virtual marker definitions allow you to continue generating virtual marker data if one of the definition markers is not being tracked. For the Join Virtual function to work properly, you will need a minimum of four different support markers among the six spots to fill. If you are in Streaming mode from cameras or VC files, the first definition set is used. If you are in Post Process mode, you may choose which definition set works best.

Figure 6-27. Virtual Marker Definitions

When running live, the Join Virtual tool only uses the first VM Join definition of the two that you are allowed. However, 4 passes are made over the list on each frame so that if a definition depends on another then after the first pass the second marker is reconstructed so that the first marker can be reconstructed on the second pass. It also works this way in Post

Process mode when you have multiple markers selected and do a Join Virtual function.

6-33

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Streaming vs.

Post-Processing

The Real Time Join Virtual function (in the Real Time dashboard) eliminates what might be seen as a possible “pop” on the frame when the real marker re-appears. At that time, the Virtual Marker filling the gap is no longer used. This is the only solution for streaming (using the Join Virtual check box when you are recording the data either live or from raw video files).

In the post-processing Join Virtual mechanism, the offsets between the marker to join and the Join Virtual Origin Marker are measured both at the start of the gap and the end of the gap, and a linear interpolation is used for all in between data points. The result is always a perfectly fluid transition on both ends of the gap.

The Join Virtual mechanism is a powerful tool in creating and editing data quickly with good results. It is the result of working with our customers to define and develop techniques to get good motion capture data quickly and efficiently.

Post Process Dashboard

The Post Process Dashboard is available when

EVaRT

is in Post Process mode as opposed to Tracking mode. After you have generated and saved tracked data, this becomes available to help manage a data editing session. It controls the range of visible frames and the range of selected frames to be edited. It also provides several controls for playing through the tracked data and choosing a current frame. This dashboard is described further in Chapter 10, Post Processing Panel .

Figure 6-28. Post Process Dashboard

Play/Pause Play Speed Frame #

Move to

Lowest/Highest Frame

Active

Frame Selectors

Selected Frames—Low

Visible Frames—Low

Move 1 Frame

Time Code

Selected Frames—High

Visible Frames—High

6-34

EVaRT 5.0 User’s Manual

Post Process Toolbar

Figure 6-29. Post Process Toolbar

Chapter 6: The EVaRT User Interface

All Markers Radial

Button

Selected Markers

Radial Button

Template ID

Template Rectify

Create Template

Make Unnamed

Rectify Unnamed

All of the Identifying tools are accessible using hot keys, panel buttons, and right mouse menu items on the 3D Display and XYZ Graphs.

Allows you to select all markers at once for Identifying.

Allows you to select specific markers for Identifying.

Uses the template to ID all markers in the current frame.

Uses the template and continuous tracks to ID markers thorough time.

Refer to “Building a Template” on page 9-5 .

Make Unnamed will specifically move a marker’s data into the first unnamed marker slot. It is important to know that the data is not deleted by this operation.

Makes unnamed markers into contiguous paths to follow through the capture sequence. For more information on the Rectify functions, refer to

“Rectify” on page 6-37

.

6-35

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Quick ID

Identifies the selected marker, identifying all markers one by one, according to the list. It will normally select with auto incrementation (Auto Increment).

Marker ID is the same as Quick ID, without the auto increment feature.

Marker ID

Exchange

Hide Markers

Exchanges the XYZ coordinates of two selected markers.

Hides selected 3D view markers.

Unhide Markers

Rectify

Unhides the hidden selected 3D view markers.

Re-identifies missing makers (gaps) in a determined frame range. For

more information on the Rectify functions, refer to “Rectify” on page

6-37

.

Rigid Body Rectify

Uses the selected markers to ID unnamed markers through the capture sequence. For more information on the Rectify functions, refer to

“Rectify” on page 6-37

.

Rigid Body Rectify and Template Rectify assume that all the current marker identifications are correct. They are intended for continuing the identification process without undoing previous work.

Rigid Body Rectify is a tool that could be considered a "stand-alone" tool.

It does not use anything from the marker set definition at all. When the tool is activated:

1.

2.

3.

The selected markers are dynamically turned into a "Rigid Body" definition and measured

The previous frame and the current frame are then used to predict the next frame

Identify the frame

This stops when less than three markers of the original selected markers is identified.

Note:

If one or more markers are already correctly identified, then that can help prevent errors.

This has been used to identify the entire body.

2.

3.

4.

5.

1.

Select ALL the markers (minus the obscured ones). The starting frame must be identified manually.

Press

Rigid Body Rectify

Go forward to the frame where the misidentification occurred

Make unnamed

Repeat steps 2 through 5.

Options

Sets the sliders, zoom, and search options.

6-36

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Rectify

Cut

Copy

Paste

Cut Outside

Smooth

Calculate Virtual

Markers

Join Cubic

Join Linear

Join Virtual

Exchange

Search

Used for cleaning up the Initial Pose for making a template when you have no template to start with. Takes ALL markers on the current frame

(regardless of the All vs. Selected radial button), measures the linkages on the current frame and uses those measures to automatically sort markers into the correct marker slots.

Characteristics of Rectify:

Uses all markers, Named and Un-named

Works only on the Highlighted XYZ Selected Time Range

Uses the Named marker linkages and XYZ path continuity

It will switch Named markers (Named markers are not automatically locked)

Adjusts Linkage lengths dynamically to fit the data (including mistakes)

Uses the

Motion Capture > Tracking > Identifying Parameters

function (typical)

Cuts the data within the selected frames inclusive of the endpoints.

Copies selected markers in selected frames.

Pastes data with the Current Frame being the first frame of the paste region.

Cuts the data outside of the selected frames exclusive of the endpoints.

Smooths data within the set frames with the selected filter type. The filter selection is found in the

Post Process > Options

form. For more infor-

mation, refer to “Filters” on page 10-4

.

This calculates the virtual markers based on the parameters set. For more information, refer to “Virtual Markers” on page 11-6 .

Calculates the values to place in the gaps with a cubic spline. If you manually select the endpoints of the gap before executing the join, the function will fill the gap with a linear interpolation because the second derivative at the endpoints equals zero.

Selecting this will automatically fill the gap with linearly interpolated data.

For all information regarding Join Virtual, refer to

See “Join Virtual” on page 6-33.

Exchange requires that exactly two markers are selected. The data is exchanged between the markers within the selected frames.

Finds gaps and/or spikes throughout the data set. The current frame will be set to the first gap or spike found in either the first selected marker on

6-37

Chapter 6: The EVaRT User Interface

Delete U_n

Options

RB Join

Undo

EVaRT 5.0 User’s Manual

the marker list or all of the markers. See the Post Options form for settings.

Deletes all unnamed markers.

The Options button opens a form that lets you set the sliders, zoom, undo, and search (for gaps and/or spikes) options.

The Acceleration at Spikes function will indicate the frames in which a marker has experienced an acceleration greater than or equal to the selected value. The indicator appears as a carat (V) at the top of the XYZ

Graphs.

The Memory Gauge lets you know when you computer is running out of memory to store edits in the undo buffer.

The rigid body join feature has been created for rigid objects with 4 or more markers per segment. For rigid or semi-rigid objects such as swords, spears, head markers, torso markers, multiple markers on a basketball, it is convenient to use this feature to join across missing marker data. You must select a starting frame where all markers that you select are all present and part of a rigid body. You then select a range of frames on which you wish this to operate. Select

RB Join

and it automatically joins across the missing marker data.

Undo retrieves data affected by the most recent Edit or ID function and places it back into the data set.

EVaRT

supports ten levels of undo. This feature can be disabled or cleared on the Post Options form. If you get the message that an Undo function may not execute, you may need to clean your Undo buffer. This can be found in the

Post Process > Options >

Undo

section.

6-38

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Zooming, Rotating, and Translating

Zooming and translating a display can occur in both the 3D Display and the XYZ Graphs. Rotating only occurs in the 3D Display. Choosing

Help > Hot Keys and Tips

from the Menu Bar will bring up an online table describing how these features work.

Zoom

—In the 3D Display, zooming is accomplished if you:

1.

2.

3.

Hold the

Alt

key down.

Hold both the

left

mouse and

middle

mouse buttons down.

Move the mouse forward or left to zoom out and backward or right to zoom in.

Rotate

—In the 3D Display, rotating is accomplished if you:

1.

2.

3.

Hold the

Alt

key down.

Hold the

left

mouse button down.

Move the mouse in any direction.

Translate

—In the 3D Display, translating is accomplished if you:

1.

2.

3.

Hold the

Alt

key down.

Hold the

middle

mouse button down.

Move the mouse in any direction.

In the XYZ Graphs, time zooming is done in terms of frames (time) or amplitude. If you want to zoom in frames, there are two methods. 1) If one or no frames are selected, zooming is done relative to the current frame. 2) If two or more frames are selected, zooming is done relative to the selected frames.

Time Zoom

Method 1

1.

2.

3.

4.

5.

6.

7.

Click on the

Post Process

button among the Mode Buttons.

Press

F4

or choose

View > XYZ Graphs

from the Menu Bar.

Left-click on

None

in the lower right corner below the marker list.

Now, no frames are selected.

Left-click anywhere on the XYZ Graphs to set the Current Frame which is indicated by the red line.

To zoom in, press the “Zoom Frames In” hot key (default is

I

) or by right-clicking in the XYZ Graphs and selecting “Zoom Frames In” from the pop-up menu.

To zoom out, press the “Zoom Frames Out” hot key (default is

O

) or by right-clicking in the XYZ Graphs window and selecting “Zoom

Frames Out” from the pop-up menu.

Unzoom time: Double-click on the Time Slider on the Post Process

Dashboard to zoom out completely making all frames visible. The

Post Process Dashboard Visible boxes will now have a

1

and the highest frame number displayed.

6-39

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Time Zoom

Method 2

Amplitude Zoom

The second method of zooming in frames is described as follows:

1.

2.

3.

Hold the

middle

mouse button down in the XYZ or the Analog

Graphs.

Drag the mouse to the right or left to select any number of frames.

To zoom in, press the “Zoom Frames In” Hot Key (default is I) or by right-clicking in the XYZ Graphs window and selecting “Zoom

Frames In” from the pop-up menu.

Zooming amplitude is done relative to the closest data point and frame nearest to the location you initially click on. You can optionally zoom into the data in the Current Frame regardless of where your mouse cursor is on the screen. This option is a User Preference and it can be set by launching the Options Form from the Post Process panel.

1.

2.

3.

Hold the

Alt

key down.

Hold both the

left

mouse and

middle

mouse buttons down.

Move the mouse forward and backward.

In the XYZ Graphs, translating is accomplished if you:

1.

2.

3.

Hold the

Alt

key down.

Hold the

middle

mouse button down.

Move the mouse in any direction.

6-40

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Selecting Markers, Virtual Markers, Linkages, and Segments

Markers can be selected by the following:

1.

2.

3.

Clicking on the markers seen on the 3D Display

Double-clicking on the markers seen in the XYZ Graphs

Clicking on the markers listed on the Marker Grids (which are a part of the Post Process panel, the Model Edit > Markers sub-panel, and the Model Edit Tree View)

Note:

All of the conventional

Shift

+ click and

Ctrl

+ click techniques to select multiple items are supported in this software.

Time Code

SMPTE Time Code and EVaRT

Overview

Using the Time

Code Reader

Option with Eagle and Hawk Cameras

Using the Time

Code Reader

Option with Non

Eagle and Hawk

Cameras

SMPTE Code reads as HH:MM:SS:FF, which is Hours:Minutes:Seconds:Frame. Frame numbers are 0 to 29 in NTSC and 0 to 24 in PAL.

When you capture at a higher motion capture rate such as 60 or 120 Hz, there are multiple motion capture VC frames per color video frame. The software takes care of that so if you record your VC files at 60 Hz, the VC frame advances twice for every single frame advance in the color video when you play it back or step through the data. The SMPTE time code is visible on the Real Time Dashboard.

Eagle and Hawk digital cameras can use the Time Code Reader (PCI version) card, installed into the

EVaRT

Host computer. It reads the LTC

(Longitudinal Time Code) from the RCA audio connector on the Time

Code card, creating a

trialN.tc

file (time code) when you collect a

trialN.vcX

dataset. It is automatic if you have the Time Code Reader option

(card) installed in your

EVaRT

computer. There is a BNC type connector on the card as well; it appears that the Time Code Reader will genlock to the black burst video signal, but that is not needed.

EVaRT

reads the current time code when the data collection is started and time stamps it into the TC file. The current time code also displays on the Post-Process Dashboard.

A simple test program called

TimeCodeReader.exe

is distributed with the latest

EVaRT

releases for Eagle and Hawk camera users. It is a standalone program which launches, and in a small window reads the current value of the Time Code Reader in the

EVaRT

Host computer. It is useful for testing to see if the Time Code reader is working. Without a card installed, it just leaves a blank display. With a time code reader card installed, it displays the current time code, static or not. When the time code starts to advance, you can immediately see it.

On Midas-based systems, the Time Code Reader card is an ISA type card that plugs into the NT Midas box. On the

EVa 6.0

and

7.0

CD, there is a folder under

RTMIDAS

called

DOCS

which has information about turning

ON and OFF the Time Code reader option. See

TimeCode_Notes.doc

.

This feature is not supported on older DOS Midas systems. When it is turned on, a TC file is created in the capture folder as with the Eagle and

Hawk option above. There is a sample program for NT Midas systems

6-41

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

Using the Time

Code in Post

Processing

Time Code and the

Digital Video

(EVaDV) Option

that reads and displays the time code. It can be found in this EVaRT release distribution CD in the

Tools\RTmidas

folder with the name of

MACTCReader.exe

.

If you load a TRB or TRC file that has an associated TC file, then the

Post-Processing Dashboard will lock the time code onto the time code display. You can step forward or backwards in time or push the play button and the time code reads accurately. If you switch back to the Motion

Capture panel and are connected to the cameras, you will see the current time code. If you are in the Motion Capture panel and are not connected to the cameras, but using Raw Video Files, you will see the Time Code associated with the Raw Video File.

The Digital Video option can be used with the Time Code. In our current software, the only way to record the time code is with the Time Code

Reader card. Some cameras have Time Code capability within the camera, but those time codes are not recorded with the DV (Digital Video) option. The Time Code must be connected to the Time Code Reader card to have a TC file created and hence be time coded.

Live Video

Backdrop

The Live Video Backdrop allows you to set your streaming live video as the backdrop to your 3D display. To activate this function, right-click in the 3D display and select

Show Video

.

Figure 6-30. Live Video Backdrop

6-42

EVaRT 5.0 User’s Manual Chapter 6: The EVaRT User Interface

Unload Tracks

Button

This button provides a quick method to unload, or not save any changes to, the Tracks files which you have edited.

Figure 6-31. Unload Tracks Button

Unload Tracks Button

6-43

Chapter 6: The EVaRT User Interface EVaRT 5.0 User’s Manual

6-44

Chapter 7

Setup Panel

Topic

Getting Started

Cameras Sub-Panel

Creating and Clearing Masks

Going Live

Adjusting Thresholds

Analog Sub-Panel

Misc Sub-Panel

Getting Started

Before using

EVaRT

, you must configure your software to match the overall system. Setup Mode provides tools to do this. The camera settings do not need to be reset before each and every motion capture session but they do need to be reset after changes are made to the cameras.

1.

2.

3.

4.

5.

6.

Choose

File > Load Project...

from the Menu Bar and load a recent or sample project.

Choose

Setup

from the Mode Buttons.

Choose the Cameras sub-panel from the panel buttons if it is not already open.

Select the correct camera type from the Camera Type drop down list.

Change the camera type and frame rate to correspond to your hardware and the desired frame rate.

Leave the Sub-sampling Rate set to the default for normal data collection.

Click

Connect to Cameras

on the Real Time Dashboard if your cameras and connections are fully operational.

The system is now ready to go live with the

Run

button.

Page

7-1

7-2

7-12

7-12

7-12

7-13

7-16

7-1

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Cameras Sub-Panel

Eagle/Hawk

Camera Settings

To view the settings mentioned in this section, you must have the Camera

Type set to

Eagle/Hawk

and you must have the

[Eagle Support]

item in your

mac_lic.dat

file.

Figure 7-1. Cameras Sub-Panel

7-2

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

Frame Rate

Shutter Speed

Sets the frame rate of the Eagle digital camera to any number ranging from 0.1 to the maximum frame rate. The number does not have to be an integer; it may be set to 59.05, for example.

Note:

The maximum supported frame rate for the Eagle digital camera system is

500 Hz. Please contact Motion Analysis Customer Support for information and technical advice for using frame rates higher than 500 Hz.

Sets the shutter speed of the Eagle digital camera ranging from 0 to

2000

µ s. This pulse is issued is in conjunction with the timing of the strobe (ring light) pulse. There are 1024 different levels of shutter speed control.

Using Hardware

Sync

This is selected when you are capturing data at high frame rates (greater than 500 Hz) and the cameras are wired together with the Hardware Sync cable

Using Sunlight

Filter

When activated, this feature eliminates large blobs (targets) and one-pixel blobs in the camera hardware caused by typical outside lighting. The tracking parameter “Max Horizontal Lines per Marker” gets set as the max size allowable target in horizontal pixels.

Genlocking Master

Camera

This is selected when the master camera is synced to an external video source, either NTSC or PAL To enable the feature, you must have a license feature installed in your mac_lic.dat license file that looks something like the following line:

[Eagle Genlock] 9c3856f6 782cb125

Please contact [email protected] if you need this item.

To turn it ON, check the box called

Genlocking Master Camera

in the

Setup > Cameras

sub-panel. When this is done, the Master Camera

(which can be any of your Eagle, Eagle-i, Hawk or hawk-i cameras) must have an analog video signal (black burst or other signal) applied to the camera. This is done using the BNC connector of the 2 meter long Eagle

Test Cable that came with the Motion Analysis system. It connects to the master camera using the AUX connector on the back of all the MAC digital cameras. Failure to connect the video signal to the Master camera will show up when you press the Run button. The slave cameras will send data but the Master camera will not.

This feature is available on all Motion Analysis digital cameras and can be set to any multiple of the NTSC or PAL frequencies that your mocap camera will allow. So you can capture at 59.94 (NTSC frequency) or

119.88 (2X NTSC) or 179.82 (3X NTSC) or higher if your motion capture camera will allow it. For Eagle/Eagle-i cameras and NTSC genlock sync, you need to set the camera frame rate to 59.94 Hz (on the Cameras sup-panel). For PAL genlock sync, set the camera frame rate to 50 Hz (or

100 or 150 or 200 or so on). The slave cameras will follow the master camera without any extra wiring.

7-3

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Brightness

Current Camera

Information

Sets the brightness of the ring lights for all cameras.

Displays the selected camera number and its corresponding IP address, software version number, along with the date and time the Eagle software was compiled.

Changing the

Camera’s IP

Address

You can change the IP address in this box for any camera at any time. You need to make sure that you do not use duplicate numbers though. It is recommended that you use the same IP address number scheme as used when the cameras are first shipped (10.1.1.xxx). The last three digits should be any number between 1 and 250. In the event that your local area network is set to a 10.1.1.xxx IP scheme, you can also use 10.1.2.xxx for the Eagle camera network (Eagle Host computer, EagleHubs, Eagle cameras, etc.).

Set as Master

Sets the selected camera as a master camera. A master camera generates synchronized pulses to the rest of the cameras within the system so that all camera shutters are opening and closing at the same rate.

Note:

Only one master camera can be set for each system.

The master camera is set as follows:

1.

2.

3.

4.

In the Real Time Dashboard, select the camera number button of the camera which you would like to set as a master camera.

Click on the

Set Master Camera

button.

If the camera is not turned on or working, select another camera in the

Real Time Dashboard, and then press

Set Master Camera

again.

If you have an analog sub-system, the A-D sync cable must be connected from the master camera to the A-D Interconnect box. See Figure B-3 on page B-6 .

Any camera may be designated as a master camera, but only one at a time.

7-4

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

Loading New

Camera Software...

This allows you to select and upload new software, when available, into the selected camera.

Note:

The software loaded into the camera must be of the form

rom_date.bin

.

The specific date in the file name may vary.

Note:

Hawk-i cameras require a different version of the

rom.bin

software from the other digital cameras. Loading of incorrect software may cause your

Hawk-i camera to cease functioning. The

rom.bin

file for the Hawk-i cameras is found in a folder of its own, under the Camera Software directory.

To ensure of the latest software release, the

rom.bin

file will have the date and time of the program in the file name (e.g.

rom_Jun_23_2006.bin

).

Install the new software as follows:

1.

2.

Obtain the latest

rom_(date).bin

file from Motion Analysis Corporation by means of either an FTP site, e-mail, or disk.

Copy the

rom_(date).bin

file into the following directory:

C:\Program Files\Motion Analysis\EVaRT50\Camera Software

Figure 7-2. Camera Software Directory

3.

4.

Return to the Calibration sub-panel in the

EVaRT

user interface.

Select the camera (on the Real Time Dashboard) you wish to load the new

rom_(date).bin

software into and click on the

New Camera

Software...

button in the Cameras sub-panel.

7-5

Chapter 7: Setup Panel

Figure 7-3. Download New Camera Software Verification Dialog

EVaRT 5.0 User’s Manual

5.

6.

Manually type in the Unit’s IP address prior to clicking on the

Download

button.

Navigate to the

Camera Software

directory with the

[..]

and select

rom_(date).bin

.

Figure 7-4. New Camera Software Interface, Loading the rom.bin File

7-6

7.

Click

Download

and wait about two minutes for the Writing to Flash operation to finish.

Note:

If the message “Send Failed” appears, ignore and press

Download

again.

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

8.

9.

10.

Once the software has been loaded, the cameras may blink oddly.

After this, you will need to reboot all the cameras by cycling the power on the EagleHub.

Run the

EVaRT

software and click on

Connect to Cameras

.

Finally, check all cameras for the new software version number.

Note:

For Eagle and Hawk Camera users, the camera software with this release is:

\CameraSoftware\Eagle_Eagle 4_Hawk\rom_June_23_2006.bin

For hawk-i camera users, the camera software is:

\CameraSoftware\hawk-i_ONLY\rom_Hawki_June_23_2006.bin

Do I need to update camera software?

Yo u c a n c o n t i n u e t o u s e t h e E a g l e o r E a g l e - 4 o r H a w k

rom_May09_2005.bin

or later software. Reasons to upgrade to the new camera software include:

1.

2.

3.

Genlock to multiples of NTSC or PAL video

Hardware sync for Eagle cameras and frame rates above 500 Hz

Mixed camera environment with hawk-i cameras

Old Camera, New

Camera

Compatibility

Issues

Cameras with different revisions of the CPU board installed may exhibit problems with different versions of the rom.bin software.

How to tell which cameras have the OLD CPU board and which cameras have the NEW CPU Board is easy:

ALL MAC DIGITAL Cameras with the SILVER backplate (where the connectors are) have the newer CPU boards (CE approved) and require the newer camera software, dated

May 9, 2005

or later.

Cameras with the BLACK backplate have the older CPU card (non

CE approved) can use either the NEWER or the OLDER rom.bin software.

The

rom_Mar_11_2004

software was released with the

EVaRT 4.2

software.

The

rom_May_9_2005.bin

software was released with the

EVaRT 4.4

software

Dedicated Interface for Eagle Cameras

Reboot All

Cameras

Allows you to input the IP address for the network interface card (NIC) of the host computer. This is the IP address for the NIC that is connected to the Eagle cameras. There must be a dedicated NIC for this purpose. Other connections to local area networks (LAN) must be done on a different

NIC to avoid network traffic on the Eagle camera network and to keep the

Eagle system working properly.

Reboots the cameras (cycles the power). This is used when changing the camera’s IP addresses. Note that the camera and software will not recognize the change in IP address until the camera has been rebooted.

7-7

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Falcon and

Other Camera

Settings

Note that if you change the camera type in this sub-panel, you will need to perform a new calibration, both square and wand. Changing the sub-sampling rate does not affect the system calibration.

Sub-sampling Rate

The Sub-sampling Rate is the actual rate of raw data collection. Frames may be skipped over during the data collection if the sampling rate does not match the hardware rate.

Video Display

Options

The Video Display Options will affect the image displayed on the Falcon

Threshold Monitor.

Show Video

Shows the grey scale video image direct from a camera (for analog cameras only).

Show Threshold

Show and Use

Masks

Shows a black and white binary image indicating those picture areas above and below the specified video threshold level (for analog cameras only)

Shows the masks or areas designated to receive no marker information.

7-8

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

Eagle and Hawk Camera Display Codes

The Eagle and Hawk digital camera displays indicate which mode the cameras are operating. Note that hawk-i and eagle-i cameras do not have the display feature.

Table 7-1. Eagle and Hawk Camera Display Codes

Code

Master Camera

Standard Camera

Image Description

This display code, with the active LEDs in the four corners, indicates that the camera has been set as a master camera. The ringlights are ON, which indicates that

Connect To Cameras

on the

EVaRT

interface is active.

Yellow (Red and Green ON) numbering indicates that the camera is in an idle state

(powered up but not connected).

Red number displays indicate that the camera is either disabled, out of sync, or that there is a hardware problem within the camera.

This display code has no active LEDs other than those set for the number display

LEDs. This indicates a standard operating camera.

7-9

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Table 7-1. Eagle and Hawk Camera Display Codes

Image Code

Camera Ready for rom.bin Download

Description

This display code, with the active LEDs in a slash through the display number, indicates that the Eagle camera is ready to accept a new

rom_date.bin

file. This display code will be go away after the new software has been installed and the camera is rebooted.

Press Download

This display code, with active LEDs in an arrow and rectangle pattern, indicate that the Download button, in the Download

FTP window, is ready to be pressed.

Rom.bin Download in

Progress

This display code indicates the progress of the download process. The number of activated LEDs will increase as the download process nears completion.

7-10

EVaRT 5.0 User’s Manual

Table 7-1. Eagle and Hawk Camera Display Codes

Image Code

Rom.bin Download is

Complete

Chapter 7: Setup Panel

Description

Camera ON, not

Connected to

EVaRT

This display code, with both green and red

LEDs activated for the number display, indicates that the camera has been powered ON, but is not connected to

EVaRT

.

7-11

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Creating and Clearing Masks

Masks are rectangular regions that you designate to receive no marker information. Masks allow you to block out fixed light sources that cannot be physically removed from a camera’s view. Masks are created on the 2D

Display by clicking and dragging the middle mouse button and appear as hatched regions.

Masks can be cleared by using the right mouse pop-up menu item in the

2D Display and choosing either

Delete Mask

or

Delete All Masks

. Note that you can use masks after the VC data has been collected by disconnecting from the cameras, creating the mask, and then loading or selecting the raw files. This applies to the calibration files (CalSeed and CalWand) as well.

Going Live

After having configured the software to the system and connected

EVaRT

to the cameras:

1.

2.

3.

Place the Calibration Square on the floor in the capture volume. The orientation of the Calibration Square determines the directions of your global X, Y, and Z axes.

Press

F2

on the keyboard to open the 2D Display. The view seen by one or several cameras will be displayed. To select multiple cameras, press

Shift

+ click or

Ctrl

+ click on any of the camera buttons on the

Real Time Dashboard or click

All On

.

Click the

Run

button on the Real Time Dashboard.

Adjusting Thresholds

1.

2.

Choose

Tools > Edit Thresholds...

from the Menu Bar.

Slide the Threshold slider on the floating Thresholds control until the markers on the floor appear on the screen.

Figure 7-5. Threshold Slider

3.

4.

Select a camera.

If you are not seeing any blobs on the screen, choose

Motion Capture

from the Mode Buttons. Then in the Tracking Panel, set the Min. Horizontal Lines per Marker to

2

and Max Horizontal Lines per Marker to

100

.

7-12

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

5.

6.

Mask out any unwanted light sources by creating a mask with the middle mouse button held down. Delete masks by clicking on a mask and pressing

Delete

on the keyboard or right-clicking in the 2D Display and selecting

Delete Mask

.

Repeat Steps 4 through 5 for all cameras.

Analog Sub-Panel

For users who have integrated force plates into their motion capture system, you will need to configure the EMG or other analog source signals for the analog signals to be collected properly. This is done by following these steps.

1.

2.

Choose

Setup

from the mode buttons.

Choose

Analog

from the sub-panel buttons.

Figure 7-6. Analog Setup Grid

Note:

In general, a multiple of the frame rate is recommended. For longer captures (more than 5 seconds), this is required.

3.

4.

5.

To open a list of force plate names, right-click anywhere on the Analog sub-panel grid and choose

Channel Type Names

from the popup menu. The built in names include Kistler, AMTI, Bertec and Muscles.

To edit a channel’s name, left-click in its row in the Name column.

Left-click on the arrow that appears in the cell and select a name.

Alternatively, you can simply left-click in the cell and type in a name directly.

7-13

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

6.

7.

8.

9.

Left-click in the channel’s row in the Active column if you wish to make it active. You can also click in the Active column’s title cell or right-click on the Analog sub-panel grid and select

Activate All

Named Channels

. Both actions activate all named channels.

Left-click in the channel’s row in the Range column and click on the arrow to select a voltage range. The range must match the output of your analog device.

Select the correct sample rate for your system from the Sample per

Second drop list at the bottom of the panel.

To select EMG muscle names, right-click in the Analog sub-panel and select

Channel Type Names > Muscles.

For any particular analog channel number, left-click in the Name column and scroll through the drop-down menu for the EMG muscle name you want (see

Figure

7-7

).

Note:

The Forceplate and EMG muscle names are consistent with the names used in the

Orthotrak

Gait Analysis and KinTrak software from Motion Analysis.

Figure 7-7. EMG Muscle Name Selection

7-14

10.

11.

To save the entries in the Analog sub-panel, choose

File > Save

Project

from the Menu Bar.

If you would like to give the file a new name or save it to a different directory, choose

File > Save Project As...

.

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

Shifting

Selected Analog

Data

This series of functions is used to correct the time-shift-delay in certain telemetered EMG channels. These are introduced by Noraxon EMG units

(model 2400 and later), which have a 15ms delay in their signal transmission protocols. As a result, the EMG signals in multi-source (EMG and

Forceplates) analog data collected by the Motion Analysis system become non-synchronous with the motion data.

EVaRT

allows the user to time shift the EMG data in the analog channels using the following steps:

1.

2.

3.

4.

5.

6.

7.

8.

Load a Project File

Load a Tracks file (.trb/.trc).

Select

Data Views > Analog Display

.

In the Analog Display window, right click and select the Visible

Channels option.

In the Analog Display window, right-click and select the

Shift the

Data

option.

In the Visible Channels dialog, select only the channels you want shifted (typically the EMG data).

In the Shift the Data dialog, select the

Selected Channels

option.

Set the value for the “Shift the data this number of samples”. This can be calculated by the following formula:

# of samples to shift = (Analog Sample Rate) x (Time Delay)

Example

For an analog sampling rate of 1200 samples/second and a time delay of

15 ms (15 x 10

-3

seconds) the calculation would be:

# of samples to shift = (1200 samples/sec) x (15 x 10

-3

sec) = 18 samples

1.

To input this frame shift correctly, enter the number as negative value

(–18).

Negative values indicate a shift to the left (decreasing the delay), positive numbers indicate a shift to the right (increasing the delay).

2.

3.

To change the entered value from red to black, press

Enter

.

Click on

Apply

. When it prompts you with "Would you like to rewrite the analog file?", select

Yes

.

Figure 7-8. Shift the Analog Data Dialog Box

7-15

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

Misc Sub-Panel

From Raw Video

Files (Options)

CPU Speed

Real Time

Loop Raw File

Processes files as fast as the CPU will go.This is a feature when you are not connected to the cameras, but you are using Raw Files. If you click on the Run button (which causes the program to perform all the checked functions: Track, Identify, Skeleton), the difference is that CPU speed will go as fast as possible, which means if you have a fast enough computer, it is done is less time than the capture time.

Slows down the software to pause between frames if there is available time. If you select Real Time speed, it will wait between frames (and update the display) so that the motion being tracked is not faster than realtime playback. If the time to process the data is slower than realtime (lots of markers, fast frame rate), then the selection does not matter (it tries to keep up with real time, but there is not enough time). When you are Connected to Cameras, the CPU Speed setting does not matter.

Loops the file to continuously repeat in Run mode. Note that if you record to a trb file with this feature ON, the trb file will contain all of the loops as well.

Skeleton Engine

The Misc sub-panel ( Figure 7-10

) provides additional setup options and functions that are in development stages.

No Skeleton

Calculation

Skeleton Builder

(SkB)

Turns off any skeleton calculations for the Motion Capture and Post Processing panels

.

Fill in SkB Gaps

Calcium Solver

1.2.10 (or later)

Turns on the skeleton builder definitions (if present in the project file) and allows the skeleton to be calculated. To see the skeleton in the Motion

Capture or Post Processing panels, you must have the Show Skeleton feature turned on in the 3D window (right-click in the 3D display and select).

All

EVaRT

users can calculate the

SkB

skeletons after they have been defined, but creating and editing skeletons requires a separate software license.

If the markers for a bone segment disappear, the position information for that segment is automatically filled in based on its previously measured position in relation to its parent. This is helpful in live performance situations to help make the animated character's motion appear more smooth when the markers that define the segment disappear.

Checking this option tells the software to use the Calcium Solver type skeleton. The version number of that software is listed after its name and it can be loaded independently with a different

solver.dll

. You must have the Show Skeleton feature turned on in the 3D window and you must have

7-16

EVaRT 5.0 User’s Manual Chapter 7: Setup Panel

SIMM OrthoTrak

Model

Create Orthotrak

Model

clicked the Calculate Skeleton "Bone Button" in Post Processing to have the skeleton calculated. Note that

SkB

skeletons are calculated automatically in Post Processing, but you must press the Big Bone button to calculate the Calcium Solver skeleton. To calculate the skeleton, you must also have a model file (the .mod file) of the same name as the project file in your current directory. Model files are created and edited in the

Si 2.0/

Calcium

software. Any user can run and calculate the Calcium Solver skeletons, but it takes a separate license to edit and create the model files.

The skeleton can be calculated in the Motion Capture panel from either live camera data or in the simulated realtime mode when you "Disconnect

-Use Raw Files".

This skeleton calculation uses the Calcium Solver type skeleton and requires that you use the anatomically named marker set defined in the

OrthoTrak

software. The

OrthoTrak SIMM

basic marker set uses fixed names like L.Shoulder, L.Wrist, and L.Knee for prominent marker locations. There are several required markers and many more optional markers. If you use this marker set, you do not need to use the

Si/Calcium

software for creating a model (.mod) file. You must also use an additional

"Static" trial with inside (medial) knee and ankle markers. You load the

Static project file, load the Static Trial, click this

SIMM OrthoTrak

button and then the Big Bone (calculate skeleton) button becomes active. Then load the Walking (motion) trial and click the

Big Bone

button on the Post

Processing screen. The Calcium Solver skeleton is calculated. To see the skeleton in the 3D window, select

Show Skeleton

. The skeleton can be calculated in the Motion Capture panel from either live camera data or in the simulated realtime mode when you "Disconnect -Use Raw Files".

The Create OrthoTrak Model button brings up the dialog box as shown in

Figure 7-9

. It uses a pre-defined SIMM model, which is pre-defined in a

.jnt file and is generally part of the

EVaRT

software release. It also uses the current Tracks file with the Static + Dynamic markers to create a scaled model of the person with the correct bone lengths. This allows the user to calculate the Skeleton with the Post-Processing Calcium Bone button or in the Motion Capture mode in real time. The skeleton engine that is used is the very high-quality Calcium skeleton engine which iteratively fits the "marker cloud" of named markers around the pre-defined skeleton structure using a least mean squares, "global optimization" method. The joints are pre-defined in the SIMM-type .jnt file and modification of this file requires detailed knowledge of the SIMM (Software for Interactive

Musculoskeletal Modeling) software. The steps for calculating a new

OrthoTrak-SIMM model are as follows:

1.

2.

3.

4.

Select

File > Load Project

with all marker names used in the

DYNAMIC or Walking trials.

Select

File > Load Tracks file

with all DYNAMIC markers.

Select

Setup > Misc > Skeleton Engine

and click-on the button

Create OrthoTrak Model...

Select

Model Definition

and browse to the correct .jnt file. The

OrthoTrak marker set uses the mocap.jnt file (the default). Jack uses the Jack.jnt file. Note that it is easiest if you put copy the .jnt file to the folder where your

EVaRT

executable resides and you will not need to navigate to another folder.

7-17

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

5.

6.

7.

Select

Init Pose

. Browse to the .trb file where you have all of the

Static + Dynamic markers. For the OrthoTrak model, this is the trial with both legs with the medial ankle and knee markers applied. For the Jack software there is only one marker set for both static and dynamic trials. Use the standing Init trial with the legs together facing the +X direction, arms out, elbows bent at 90 degrees, thumbs pointing towards each other. Click-on

Create

. This creates the scaled model with the correct bone lengths for this person and saves this in a

.jnt file with the same name as your project name in your current data folder. In the future, if you load the project file, the associated .jnt file also gets loaded. If there is a personal.dat file in the folder with the init pose file, Solver will read it when scaling the model. Hip center offsets, ankle and knee diameters, and foot lengths are read from this file, if they are present.

In the Post Process panel, in bottom left corner you now have the Calcium Bone button. Clicking on this button solves the skeletal segments for your walking trial with the correctly scaled bone lengths for this person. At this point, you can use the 3D window right-click

Show Menu to show or un-show: Skeleton, Skin and Skeleton Axes.

Also, the

File > Export HTR

function allows you to save the segmental data into an HTR file.

For additional Dynamic Trials of the same person, select

File->Load

Tracks file with DYNAMIC markers

. Note that the Bone button still remains in the bottom-left corner of the

EVaRT

interface. Press the

Bone button and a correctly scaled Skeleton gets created. At this point, you can go to

File > Export HTR

and you will have a file with the segmental data. Also, you can use the 3D window right-click

Show Menu to show or un-show Skeleton, Skin and Skeleton Axes.

Note, that this IS the way you have to run the create skeleton option.

Once you have done the 'Create OrthoTrak Model' step, you can run a real time capture with a skeleton being generated on the fly. Also, if you save the project file in step 3, you can Load the project file with

Dynamic trials and start at step 5 above. Also, after you have create the OrthoTrak model, you can check the box in the Motion Capture mode

Calculate Skeleton

and the skeleton will be calculated in real time when you are Connected to Cameras or Disconnected and using

VC files to simulate real time performance. It takes more CPU resources to calculate these, so pay attention to the

Task Manager >

Performance

tab to see if your computer is fast enough.

7-18

EVaRT 5.0 User’s Manual

Figure 7-9. Create OrthoTrak Model Dialog

Chapter 7: Setup Panel

Model Definition: A SIMM defined JNT file must be selected. The

mocap.jnt

file is the one used for the OrthoTrak marker set. Other joint files exist for special applications, such as specific ergonomic models.

Init Pose: Select the static trial that applies additional medial markers to help define the joint centers. This is the static trial in the OrthoTrak marker set and sometimes called the Init Pose with the animation software packages.

Other Misc Sub-

Panel Functions

Disable Sound

Effects

Load Another

Tracks File

Frame Offset

Number of Marker

Slots

This turns the sound features of

EVaRT

on and off.

This merges another track trial into memory so you can view two different data trials in the 3D window. A Frame Offset lets you offset in time the merged data. Only positive offsets are allowed. It is meant for visualization of multiple data sets in the 3D window. The first trial that is loaded must have enough frames to wholly accommodate the additional trials as additional memory for any additional frames is not allocated and can cause unpredictable results. You can make one trial that has enough frames to fully accommodate all of the merged data frames and load that track data set first with the

File > Load Tracks

menu item. The merged data does not get marker names or linkages unless you have a project and marker sets defined for the additional tracks.

This offsets the first tracks file with respect to the second. The first tracks file loaded must have enough frames to accommodate all additional tracks files. If not, unpredictable results may occur. This is intended for viewing multiple trials. To see the stick figures, multiple marker sets must exist in your current marker set.

This function depicts the number of marker slots the software will use.

EVaRT

allocates memory based on this number. Earlier releases of

EVaRT

would require a Heavy Duty version for a large number of marker slots. This latest version will variably assign memory according to the number of marker slots used. The maximum number of marker slots that

EVaRT

will allow is 1500. The default number of marker slots is set to

7-19

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

192, which uses 512 MBytes of memory. From 193 to 500 markers, the amount of memory used is 750 MBytes. From 501 to 1500 markers, the amount of memory used is 1000 MBytes.

Note:

Upon changing the number of marker slots, you must relaunch

EVaRT

.

Forceplates—Force

Vector Scale

This adjusts the scaling factor of the force vector that is deployed in the 3-

D display.

Forceplates—

Autozero Forces

Streaming Options

When active, this check-box automatically zeros all named forceplate channels. This does not affect the raw data in the .anb files. This does affect the following:

1.

Force Vector Display in the 3D window

2.

File > Export Force File

function

3.

Tools > Show Forceplate Forces

function

This works when connected to the cameras and uses the first analog sample to set the zero value. This is set when the user selects Run or Record.

In Post-Processing, it uses the first analog sample of each forceplate channel in the .anb file to set the zero values.

This option streams all motion capture and post process playing data to the NIC address specified. This must be a NIC (Ethernet) address of the

EVaRT

host computer. If there are multiple NIC cards in the host computer, you must indicate which card will be used to stream the SDK2.

7-20

EVaRT 5.0 User’s Manual

Figure 7-10. Misc Sub-Panel

Chapter 7: Setup Panel

7-21

Chapter 7: Setup Panel EVaRT 5.0 User’s Manual

7-22

Chapter 8

Calibration Panel

Topic

Calibrating Your System

What is the Square and Wand Calibration?

Calibration Sub-Panel

Calibration Files

Refine Sub-Panel

Calibration from Previously Collected Files

Tips for Getting More Accurate Data

Extending the Seed Calibration

Post Processing Square and Wand Data

Calibrating Your System

A new calibration must be performed whenever:

• camera positions have changed

• the coordinate system orientation has changed

• the units of measure have changed

• you have changed the camera setup for your Falcon camera

(Eagle and Hawk cameras do not require recalibration when you change the frame rate.)

It is imperative to complete an accurate calibration in order to collect high quality motion data.

Calibrating your system is a two step process. First, the seed calibration is done by employing the Calibration Square. The exact positions of these markers must be known. Next, a wand with precisely located markers is waved around throughout the capture volume by somebody wearing no reflective material. Wand calibration ensures that a direct measurement of an object of known size has been made by all cameras throughout the entire capture volume.

This process locates the exact positions of your cameras and accounts for any geometric distortion the camera lenses may have, as well as accurately measuring the camera lens focal-lengths. The importance of this information is so great that a new calibration must be completed if a camera is moved or even accidentally bumped.

Page

8-1

8-2

8-3

8-13

8-15

8-18

8-18

8-19

8-22

8-1

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

What is the Square and Wand Calibration?

In

EVaRT

, a four-point square and wand calibration takes the place of the cube and wand calibration that was done in previous versions of

EVa

software. This method has proven to be very robust and is extremely accurate.

This method requires only four markers and the wand.

Note:

If using an L-frame, orient your markers in the same directions as illustrated. Care should also be taken in placing the 4 points on the floor as this determines the global axes and the orientation of the volume displayed in

EVaRT

. The points on the 3-point axis must be in a straight line and the spacing of point 2 must be close to 1/3 of the distance between points 1 and 3.

8-2

EVaRT 5.0 User’s Manual

Calibration Sub-Panel

Figure 8-1. Calibration > Calibrate

Chapter 8: Calibration Panel

Protect Lens

Correction

This locks the lens corrections coefficients for all cameras as saved in your project file. Once you have set your lenses’ focus and zoom factor, the lens distortion maps should not change and they need not be calculated with each wand calibration. With

Protect Lens Correction

checked, the wand calibration will converge more quickly. So if you do not change the lenses, it is a good idea to leave this box checked for all your calibrations AFTER you have completed a good wand coverage and good wand calibration. The results of the successful wand calibration are stored in your project file and your System Calibration file. The System

Calibration file can be saved after each wand calibration. It is the default calibration that is used when your launch the

EVaRT

software.

8-3

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

For a very accurate calibration, do the following steps:

1.

2.

3.

4.

Uncheck

Protect Lens Correction

.

Collect a very thorough wand calibration making sure to cover the corners of all the cameras. Uncheck

Heavily Weighted Seed

and then select

Run Again

until the numbers stop changing. Accept the results.

Check

Protect Lens Correction

.

Disconnect from cameras. Select the

WandCal.vc1

file and press the

Calibrate > Calibration > Calibration with Wand Calibrate

button

AGAIN. Let it run once, un-check

Heavily Weighted Seed

and press

Run Again

until the numbers stop changing, then Accept the results.

The above procedure uses the first Wand Calibration to determine the

Lens Distortion mappings and uses the second processing of the Wand calibration to refine the calibration. Subsequent wand calibrations can be run like steps 3 and 4 if you do not change the lens settings.

Details... Button

(Calibration

Settings

Window Tabs)

Calibration Frame

Click the

Details

button, located in the upper-right corner, in the Calibration sub-panel.

The tabs for the calibration settings window, shown in Figure 8-2 on page

8-5

, are defined as follows.

Where you enter the measurements of the Calibration Square. You can make your own Calibration Square by placing four markers on the floor and measuring their locations with a tape measure. Measurements should be within 1 mm. See

Figure 8-2

for reference alignment using the Z-up calibration method. Selecting a different calibration up-axis will show the correct view on how to set up your calibration square.

8-4

EVaRT 5.0 User’s Manual

Figure 8-2. Calibration Frame Tab

Chapter 8: Calibration Panel

8-5

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Origin Offsets

This allows for translation and rotation from the origin. The Calibration

Square may then be positioned anywhere in the motion capture area. This is useful for two possible reasons:

1.

2.

All cameras do not see the calibration frame, but you want to use it to position the cameras. In this case, you can move the calibration frame to where it can be seen and enter the Origin Offsets measured from the true origin to the (temporary) location of the calibration frame.

You want a different location for an origin for any reason.

Figure 8-3. Origin Offsets Tab

8-6

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Lenses/Orientation

This indicates the focal length and positions of each camera as used in the

Collect Calibration Square

button. If you use the

Preview Calibration

check box to position and orient your camera, the focal length entries should be nominally correct (e.g. 6 for 6 mm lenses). In the Preview Calibration and Calibration with Square functions, the

Show > Show Camera Field of View

cone is determined only by what you put in this table.

After the wand calibration, the actual focal length of the lens is calculated exactly and can be stored in your project file.

Note:

The

Calibrate Wand

option calculates the actual focal lengths, but does not update the table.

Figure 8-4. Lenses/Orientation Tab

8-7

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Capture Volume

This displays the rectangular capture volume according to your measurements and helps to provide a visual reference of the volume to the operator. It does not affect the tracked data in any manner. It is for display purposes only.

Figure 8-5. Capture Volume Tab

Square (Seed)

Calibration of

Cameras

Using Preview

Calibration to

Position or Aim

Your Camera

Select

Calibration > Details,

and a

Calibrations Settings

window pops

up as shown in Figure 8-2

. The Calibration Square should be laid out on the floor exactly as is in this figure. This can be done with four separate loose markers, or it can be done using the wand and a single loose marker, placed at the end of the wand handle. The distances from the origin are measured and are entered into the

Measurements

spaces. Observe the right-hand rule and make sure that you enter the data correctly. In the Z-

Up example in Figure 8-2 , points 1, 2, and 3 would be at +X, and point 4

would be at +Y coordinates, but adjust accordingly to your Calibration

Up Axis. The vertical distances are the distance from the center of the markers (centroid) to the floor. Click on the other tabs and fill in the values accordingly. The Lenses tab should reflect the type of lenses you have in your camera (e.g. 6 mm, 17 mm, 20 mm etc.). The values for the lenses need only be approximations within a factor of two. The actual focal lengths are calculated when you process the wand data. Once you have completed filling in the details, press

Apply

and return to the Calibration window in

EVaRT

.

Select

Calibration > Calibrate Box

, check

Preview Calibration

and then press

Run

. The cameras that see four individually defined markers will instantly adjust to their approximate positions in

EVaRT

(as in

Figure

8-6 ). If a camera does not see all four markers or sees more than four markers, it will be displayed at the origin, facing down as in Figure 8-6 .

This camera will not have a seed calibration, which is acceptable. Refer to

8-8

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

“Extending the Seed Calibration” on page 8-19

. If the camera is not seeing all of the points, first try one or more of the following steps:

1.

2.

3.

4.

Adjust the threshold to see four centroids.

Insert Masks to eliminate stray data points.

Move the camera position so that it sees four defined markers. In

Figure 8-6

, a poorly positioned camera will be shown as a camera situated at the origin.

If one camera is seen in the exact opposite position in the room, the orientation (up/down) must be changed in the Lenses tab in the Calibration Settings menu. This usually occurs when cameras are tilted more than 90

°

or mounted upside-down.

Figure 8-6. Poorly Positioned Camera 4 Results in a Non-Seeded Camera (Camera #4)

Example: The camera does not see the Calibration Square. Click on the camera to identify it.

8-9

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Once you have all of your cameras positioned and oriented correctly, press the

Collect and Calibrate

button (see

Figure 8-7 ). This creates the

CalSeed.vcX

(vc1, vc2… vcN) files and consists of one frame of data stored in your current data folder. Your camera buttons at the bottom should now be yellow in color, indicating that all cameras are seeded (see

Figure 8-8

). Fully calibrated cameras show up as Green, but this does not happen until after wand calibration is completed.

Figure 8-7. Properly Seeded Cameras

8-10

All camera positions should be reasonable approximations of their actual positions in the room.

This completes the square part of the Square and Wand Calibration.

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Wand

Calibration of

Cameras

1.

2.

3.

In the Wand Calibration box on the right hand side, set the wand length to your wand size. Make sure that you are using only a threepoint wand.

Set the duration of the trial. The duration should be sufficiently long enough to wave the wand through most of the volume that you want calibrated. Smaller volumes take less time to complete.

Click the

Collect and Calibrate

button and start waving the wand side to side and up and down through the volume. You want to spend about 1/3 of the data collection time with the wand parallel to each of the three X, Y, & Z axes.

Note:

It is recommended that you view the wand movement through the volume at least for the first few times. To do this you must select

Layouts > Top/Bottom

. One window should reflect the 3D view and the area where it is possible to see the wand waving through the volume. The other is the 2D view where the individual camera coverage of the wand in the volume is seen. To show all cameras, press the

All On

button. In the

2D view, right-click and select

Smear Display

. This allows the wand to draw its paths across each camera view. A good wand calibration will fill most of the 2D Display.

Figure 8-8. Seeded Cameras; 3-D and 2-D View (all cameras on)

8-11

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

4.

5.

6.

Once the wand calibration duration has been completed, the program starts to determine the volume calibration, and a screen comes up with a series of numbers that decrease as the calibration nears the

actual wand length and focal lengths of the camera lenses (see Figure

8-9

).

At the bottom of the user interface, a progress bar ticks toward 100%.

Once completed, the camera lenses should be very close to what was installed on the camera body (e.g. 6.023 mm for 6 mm lenses). If this is the case, and the wand length is very close (e.g. 500.06 mm for a

500.00 mm wand) to the original wand length, then the calibration is complete. Uncheck

Heavily Weighted Seed

and press

Run

again.

Press

Save Project

.

Figure 8-9. Wand Processing Status

Extend Seed

Button

Run Again Button

Accept Button

Reject Button

Stop Button

This button will seed the cameras that were not seeded during the Seed

Calibration, based on the wand data. After extending the seed (clicking

Extend Seed

), all the camera buttons should be yellow. Click

Run Again

to complete the wand calibration for all cameras.

This button continues the refinement of the wand calibration. If the num-

bers in Figure 8-9 continue to change, click

Run Again

for a more precise calibration.

Click this button if the calibration numbers look sufficient.

Click this button if the calibration numbers are unacceptable. Fix the problem, then select

Run Again.

A typical fix, without having to recollect the data, is that you can create a mask in one of the Raw Video

Cal-

Wand.vcX

files.

If you need to stop in the middle of a calibration, or the numbers are not improving during a lengthy calibration click the

Stop

button.

8-12

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Note:

The Stop button is not a cancel button. You can click

Stop

and then

Accept

if you wish to keep the calibration numbers.

Floor Calibration

It is necessary to perform a floor calibration if your floor is uneven.

1.

2.

3.

4.

After the Seed and the Wand calibration is processed, place several markers on to the floor.

Press the

Connect

button and then the

Run

button. You should see the unidentified markers on the floor.

Enter the distance from the floor (your y=0 plane) to the center of the markers (maybe 20 mm) and press the

Collect and Calibrate

button in Floor Calibration.

It will tell you how much the calibration origin was moved and rotated with a six number display which stands for the XYZ and yaw, pitch and roll adjustments.

From VC Files

If you are doing this from VC files, you will need to do the following:

1.

2.

3.

Select the

CalFloor.vc1

file from

File > Load Raw Files

(with the cameras disconnected).

Then press the

Run

button so that you see the unnamed markers on the floor in the 3D view.

Then press the

Collect and Calibrate

button in Floor Calibration.

CalFloor.vcX

(optional)

You would not normally need this, but it is there to “level the floor” if needed. It is typical to take a single walk cycle and copy and paste it into

100 cycles. If the Calseed device is slightly tilted up or down, this can cause the stick figure to be walking above or below the floor at the ends of the cycles. To correct for this, you can spread 4 or more markers on the floor and press the Calibrate button in the Floor Calibration box.

Note:

Make sure no other markers (ghost or otherwise) are visible in the 3D display, as it will tilt your new virtual floor to average them in as well.

Face Calibration

Refer to Appendix D, Capturing Facial Motion .

Calibration Files

Calseed.vcX

The following are the types of calibration files generated in your selected capture folder:

Calseed.vcX

files (one for each camera) get written when you press the

Calibrate

button in the Calibrate with Square box and when you are connected to the cameras. If you are not connected to the cameras, you can use the Disconnect-Use Raw Files item and select the

Calseed.vc1

file to re-process the Calseed files. This is a kind of simulated realtime mode that allows you go back and process the

Calseed.vcX

files and evaluate your data files. When you press the

Calibrate

button in the Calibrate with

Square box, it completely removes all of your calibration information and replaces it with the seed or approximate calibration for each camera.

8-13

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Calwand.vcX

TrialN.cal

Note:

vcX means the set of files that end in vc1, vc2.... vcN if you have N cameras.

Calwand.vcX

files get written when you press the

Calibrate

button in the

Calibrate with Wand box and when you are connected to the cameras. If you are not connected to the cameras, you can process the data in the simulated real time mode as above. The software uses the current system calibration (which is normally the results of the seed calibration, but can be otherwise) and refines the calibration. The calibration includes the exact location and orientation of each camera with respect to the origin, the lens distortion parameters for each camera, and other details about the cameras. At the successful completion of the wand calibration, the software asks if you want to save the system calibration. A

Yes

answer means that a file called

SystemCal.prj

gets written to the system directory.

Other uses for the

SystemCal.prj

file are when you launch

EVaRT

, the software automatically reads the

SystemCal.prj

file and when you exit the

EVaRT

program, it automatically writes the

SystemCal.prj

file into the system folder. The intent is so that you can launch the

EVaRT

software and it will remember its last good calibration without having to load any files. If you load a PRJ file or load a CAL file, it will overwrite the calibration information in memory with the contents of the PRJ or CAL file. Both contain calibration information, but the PRJ file also has the marker set information and template information.

It is a good practice to use

TrialN.cal

for every capture you make. Every time you collect a trial in the

Motion Capture > Output

sub-panel, the system writes out the current calibration to a file that has the same name as your trial name, but with a

.cal

extension. This is normally not needed, but will allow you to load up the calibration at the time of the capture with the

File > Load Calibration...

menu item. If you changed the calibration for some reason and you know you were calibrated when the trial was collected, you can load up that as the current calibration in the software at a later time.

8-14

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Refine Sub-Panel

After you calibrate the system, you can use the Refine sub-panel to either improve your calibration or to fix the calibration if a camera gets bumped or the cameras have moved slowly over time. When performed well, the

Refine Cameras function can greatly improve the accuracy of the system, and will fix a bumped camera, all within 60 seconds. It is one of the most powerful tools in

EVaRT

. To use it, check the

Refine Camera Positions

check box press the

Run

button when connected to your cameras. Have a subject move about in the capture volume. Like with wand calibration, you must cover the entire capture volume and field of view for each camera with any marker data. If you get good coverage, then refined calibration will be very good. If you get only partial coverage, then you the results may be worse than not doing the refine.

Figure 8-10. Refine Sub-Panel

Camera Details in 2D View

The panel on the bottom of the Refine sub-panel displays the details of the selected camera in the 2D view. You can move and rotate the cameras in the 3D display by changing the X, Y, Z coordinates, and the Elevation,

Azimuth, and Roll angles. This can be helpful to see if moving the cameras will help with seeing the volume better.

8-15

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Refine Procedure

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

Connect to cameras (or select a Raw Video file).

Select and activate the

Refine Camera Positions

check box.

Press the

Run

button on the Real Time Dashboard.

Press

All On

on the Real Time Dashboard so that all camera views are displayed on the 2D Display.

Right-click on the 2D Display and select

Smear Display

. This will show you how much of each camera’s field of view is being filled by the wand over time.

Start with the subject in one corner and as soon as the subject starts to move, check the check box for Refine Camera Positions.

Have the subject walk around the capture volume filling the entire volume. The subject is acting like a wand calibration. Have the subject walk with both arms slightly out so that all markers are easily identified. Then, have the subject walk the perimeter of the room, spiraling into the middle.

As soon as the subject has filled the room and reached the middle, press

Pause

.

A table of correction values will appear for all cameras included in the Refine sub-panel. The first three columns are the position changes of the cameras since the original calibration and the second three columns are the rotational changes. Changes of more than 1 mm are often significant and can result in a better calibration.

Click

OK

and save as a new project.

Press

Run

to start

EVaRT

again. You should now see lower residuals and fewer ghost markers.

There are a few things to be concerned about. Since this behaves just like a wand, and if the subject does not fill the volume during the Refine trial, the calibration can be poor and even worse than before the Refine. If the subject spends a lot of time in one area and not much in another, it can also be poor. If the markers are not identified, then there is no information for the Refine, so ensure that the subject is identified. Save the Refine as a new project (e.g.

Refine.prj

) so that if the new calibration is not as good as before, you can go back to the previous project file. The other thing to remember is that the Refine does not guarantee that the scale is exactly maintained. It just optimizes the camera locations to track the markers better. In effect, the scale of the room may change slightly. The reality is that Refining once does not change the scale. It is not recommend to do many Refine Cameras in a row to improve the results, since this may change the scale.

Note:

You do not need an identified stick figure to refine, you only need to see

3D unnamed (or named) markers to refine the calibration.

You can use any 3D data points to Refine the calibration. There does not need to be a template or a stick figure. Any data that fills the volume will be sufficient.

8-16

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Show Camera

Volume

Show Camera Volume is useful for telling you about how well your cameras are aimed and how much camera overlap you have. To see the camera volume:

1.

2.

Select the

Show Camera Coverage

check box in the

Calibration >

Refine

sub-panel.

Right-click in the 3D view and select

Show Volume

.

Figure 8-11. Show Camera Coverage, Volume

8-17

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Calibration from Previously Collected Files

This section describes how to simulate a calibration using previously recorded data. A simulated collection of

EVaRT

square or wand data is done the same way that you can simulate tracking Raw VC data. Just follow these steps:

1.

2.

3.

4.

5.

6.

7.

8.

9.

Disconnect from your cameras.

Select

Raw Files

on the Real Time Dashboard.

Load the

CalSeed.vc1

file. At this point, you can mask out any extraneous data points if necessary.

Press the

Run

button on the Real Time Dashboard.

In the Calibrate sub-panel, press

Calibrate

in the Calibration with

Square box.

The cameras buttons on the Real Time Dashboard should turn Yellow which indicates calibration is square, but not wand calibrated. White means not square. Green means fully calibrated.

Select

Raw Files

on the Real Time Dashboard.

Load the

CalWand.vc1

file. At this point, you can mask out any extra data points that might be causing problems. This feature allows you to utilize wand data sets if there are extra markers in the field of view of one or more cameras.

In the Calibrate sub-panel, press

Calibrate

in the Calibration with

Wand box. You will see the wand results, see the cameras move to their final place, and see the measured focal distances.

Tips for Getting More Accurate Data

If the wand length is relatively close (e.g. 501.00 mm for a 500.00 mm wand), then select

Heavily Weighted Square

and

Run Again

. This keeps the Video origin very close to what the Calibration Square device defined as the origin. This is required by Biomechanics researchers where force plates are installed in the floor. If the number is much larger or much smaller, then proceed to

Extending the Seed Calibration

in the following section.

8-18

EVaRT 5.0 User’s Manual Chapter 8: Calibration Panel

Extending the Seed Calibration

If one or more cameras are not seeing the four-point Calibration Square device for any reason, they have not been properly calibrated and it shows up as a white camera button. The camera is shown on the floor at the origin point down, as shown in

Figure 8-6 on page 8-9

. This will result when you have a large capture volume and only some cameras see the Calibration Square or a camera threshold may be set incorrectly. This is not a problem. You can use the wand data to get the camera calibration seeded, then process the wand data again so the camera (or cameras) get both square and wand processing. The steps for this are as follows:

1.

2.

3.

4.

Calibrate using the Calibration Square as described in

“Calibration

Frame Tab” on page 8-5 . Cameras that are Yellow are seeded. Those

that remain White are unseeded and show up on the floor, at the origin pointing down.

Process the wand data. The cameras that saw the Calibration Square will show as Green camera buttons; the unseeded cameras remain

White.

Extend the calibration seed by clicking

Extend Seed Calibration.

This will then seed those cameras previously unseeded.

Click on

Run Again

. This runs the wand data again for all cameras.

After this, all cameras should be Green (calibrated).

Figure 8-12. Wand Processing Status Window

Heavily Weighted Seed

Cameras 5 and 7 are not seeded. Their cameras buttons stay white.

8-19

Chapter 8: Calibration Panel

Figure 8-13. Extend Seed—After First Pass

EVaRT 5.0 User’s Manual

5.

After the second pass, turn Heavily Weighted Seed

OFF

. Click

Run

Again

.

Figure 8-14. Extend Seed—After Second Pass

8-20

6.

Press

Extend Seed

, with Heavily Weighted Seed

OFF

.

EVaRT 5.0 User’s Manual

Figure 8-15. Extend Seed—After Extend Seed

Chapter 8: Calibration Panel

7.

Press

Run Again

.

Figure 8-16. Extend Seed—Third/Final Pass

Residual Values

Note:

After you have clicked on

Extend Seed

, check that the residual value for each camera is at a reasonable level. If all cameras do not eventually seed, you will need to check your Wand coverage.

8-21

Chapter 8: Calibration Panel EVaRT 5.0 User’s Manual

Post Processing Square and Wand Data

It is possible to process the square and wand data after the data has been collected. This is helpful if you did not have the time to process it during the capture or if you lost the project file that contains the calibration information. The steps are as follows:

Note:

You must have the

CalSeed.vcX

and the

CalWand.vcX

files from the data capture session.

1.

Select

Raw Files

in the Real Time Dashboard and then select

Cal-

Seed.vc1

. Select your CalSeed data set in the data capture folder.

Note:

You can create a mask to eliminate unwanted markers or reflections. Reselect the

CalSeed.vc1

file again and press the

Calibrate

button in the

Calibration with Square section.

2.

3.

In the Calibration with Square section of the Calibration sub-panel, click

Calibrate

.

Select

Raw Files

in the Real Time Dashboard and then select

Cal-

Wand.vc1

. Select your CalWand data set in the data capture folder.

Note:

You can create a mask to eliminate unwanted markers or reflections. Reselect the

CalSeed.vc1

file again and press the

Calibrate

button in the

Calibration with Square section.

4.

5.

In the Calibration with Wand section of the Calibration sub-panel, click

Calibrate.

Complete the wand calibration as described in

“Wand Calibration of

Cameras” on page 8-11 .

At this point you now have your cameras calibrated and you may proceed with your data collection.

8-22

Chapter 9

Motion Capture Panel

Topic

Overview

Tracking Sub-Panel

Building a Template from the Range of Motion Trial

Multiple Tracking Objects

Objects Sub-Panel

Output Sub-Panel

Recording Data

Tracking Strategies and Tips

Page

9-1

9-2

9-5

9-7

9-8

9-11

9-14

9-15

Overview

Motion Capture is the mode where you will spend most of your time during a recording session. In this mode you can:

Create and improve a template

Set the tracking parameters

Save data in a variety of file formats

These functions are described in this chapter.

There are a few preliminary steps that must be taken before starting a successful motion capture session. Tracking parameters tuned to your system must be set. Names must be assigned to the markers that will be used.

These names constitute a marker set and building this set is actually done using the Model Edit tools discussed in Chapter 11, Model Edit Panel . A template specific to the markers in use must be created. A template describes the minimum and maximum distances that separate linked markers, such as the distance between the right elbow and the right wrist. Templates are created using Motion Capture tools and Post Process tools described in Chapter 10, Post Processing Panel . Once these steps are completed, you are ready to begin a motion capture session.

9-1

Chapter 9: Motion Capture Panel

Tracking Sub-Panel

Figure 9-1. Tracking Parameters

EVaRT 5.0 User’s Manual

Centroid

Parameters

Min. Horizontal

Lines per Marker

The Tracking sub-panel allows you to edit the key parameters that are used when acquiring and tracking data.

Figure 9-1

is an example of the parameters entered for a typical setup. These settings are all saved in the

project

(

.prj

) file. These parameters fall into three categories:

Centroid Parameters control the minimum and maximum number of video lines that are permitted for marker images. If an image size falls outside these limits, no centroid will be calculated for it and it can never become a marker image.

Sets the minimum number of scan lines a marker must occupy on the camera’s sensor for it to qualify as a marker. The value of the parameter entered is dependent on the size of the markers and the distance the cam-

9-2

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Max Horizontal

Lines Per Marker

Shape Analysis

era is away from the markers. A typical value for a 1 inch marker would be

2

. For Eagle-4 cameras, these values will generally double to 4 lines per marker.

Sets the maximum number of scan lines a marker must occupy on the camera’s sensor for it to qualify as a marker. Again, the value of the parameter entered is dependent on the size of the markers and the distance the camera is away from the markers. A typical value for a 1 inch marker would be

50

.

Filters out the centroids of blobs that are not round (e.g. a marker is partially obscured, or two markers have merged).

Tracking

Parameters

Max Residual

Tracking parameters are used when correlating the images from several cameras to establish marker coordinates in three dimensions.

Max Target Speed

Marker Size

(millimeters)

Min. Cameras To

Use

(millimeters/frame)

Max Prediction Error

(millimeters)

Is the maximum average error when rays from several cameras are combined to establish the coordinates of one marker. If the residual exceeds this amount, it is assumed that these rays are not close enough together to be seeing the same marker. This parameter value should not be less than 4 times, and no greater than 8 times, the average residual value. The average residual value is found in the lower-left corner of the screen when the cameras are running. A typical parameter value is

5 mm

.

Sets a speed limit on the markers. A marker’s track is eliminated when it surpasses this value. When tracking the tip of a golf club or other object with fast moving markers, it is possible that this value will need to be increased. A typical parameter value is

100 mm/frame

.

Limits the size of the markers so that higher residual cameras do not see more than one centroid for the same marker. This parameter should be set to the physical size of the markers in use (25.4 millimeters = 1 inch). This parameter will also set the size of the markers that appear on the 3D Display.

Is is used to identify a marker in the next frame. While the software is tracking a marker, it is assumed that it will not deviate by more than this amount along its path. Otherwise the marker will not be identified in the frame. A typical parameter value is

30 mm

.

Tells the software what the minimum number of camera’s rays are required to triangulate (track) a marker during a frame. Some users will benefit by setting this value to

3

if spurious data points (ghost markers) are seen in the motion capture sequence.

Identifying

Parameters

Linkage Stretch Parameters, in Identifying Parameters, control the acceptable lengths of links in a template. Typically, body templates have rigid linkages, so the numbers should be low. Facial captures are considerably more elastic in nature to account for soft tissue deformation, so the numbers should be higher. These are used in the Rectify function, not Template Rectify.

9-3

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

To Reconsider (std. deviation)

Max Acceptable

(std. deviation)

Is the maximum amount of linkage stretch allowed for a linkage to be established. A typical parameter value is

15

.

Is the maximum acceptable amount of linkage stretch in standard deviations before the linkage is broken. This parameter value should be no less than

3

.

Correctly

Identifying

Markers

Automatically:

Motion Capture

Mode

Dynamic Template

Stretch Limits

To Reconsider

In the Motion Capture interface, markers are correctly identified if:

1.

2.

3.

The current "dynamic template" is good and the data fits into it.

All markers are present.

The

Identifying

check-box has been activated.

The software keeps a separate dynamic template, which is used during

Real Time or simulated Real Time tracking and only applies in the Motion Capture interface. It starts with the user created Template (from the

Create Template button) in the current project and dynamically adjusts itself as the markers are seen to stretch outside of the limits set in the current Template. This is useful when you are using the

EVaRT

system for live performances. If markers are all identified, right or wrong, the dynamic template is updated for that frame. Pressing the

Reset IDs

button in the Real Time Dashboard resets the Dynamic Template back to its original, user created Template. This is to be done if the identification gets mixed up for any reason. It is a likely result of the template being stretched too much, perhaps after a mis-identification of markers. This causes the dynamic template to perform poorly and it needs to be reset back to the original user created template.

For the Reset IDs function to be successful, all of the markers used when creating the template must be present. When the markers are un-identified, the software keeps looking in the current dynamic template to identify the markers. The software will also continue to identify the markers whose history it knows about, so you can see frames where some markers are correctly identified and other markers are shown as the black unnamed marker crosses. When about 1/3 or more of the markers become un-identified, the software tries to apply the dynamic template over the entire marker set to re-identify the markers.

These are parameters that affect how much stretch the dynamic template is allowed to change, which are set with in the

Motion Capture >Tracking > Identifying Parameters: Linkage Stretch Parameters: To Reconsider (std. dev.) and Max. Acceptable (std.dev).

This does not apply to the Post Processing Template Rectify feature, but only to the Motion Capture mode of tracking.

To Reconsider is a unit-less measure of linkage stretch checks which the current frame marker identifies against the current dynamic template. The dynamic template is a measure of the minimum and maximum of each of the linkages, which is updated as the person moves about (and the linkage lengths change in time). If any of the linkages are stretched beyond their limits, the identities of markers at both ends of those linkages are changed to “Unnamed”. The limiting factor is taken as multiples of the standard deviation of the linkage length. A typical number for a tight setting is 7

9-4

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Max Acceptable

and for a loose setting is 12. A bigger number will allow the template to grow more quickly, but can cause mis-identifications. A smaller number may keep the software from identifying the markers correctly.

This is a unit-less measure of linkage stretch that is applied to all unnamed markers in the current frame in an attempt to find proper identities for them. It is applied after any linkages were deemed to have stretched excessively. This number is usually 2 or 3 less than the To Reconsider value.

Building a Template from the Range of Motion Trial

A template tells the software what the minimum and maximum distances are that can exist between markers of a relatively fixed relation. It is necessary to allow the software to identify each marker in each frame. Template information is saved in the project (

.prj

) file.

Before a template can be created, a marker set that will apply to the subject being captured must exist. If such a marker set does not exist, it must be built using the Model Edit tools described Chapter 11, Model Edit

Panel . Once an appropriate marker set exists, follow these steps to create a template.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

Choose

Motion Capture

from the Mode Buttons.

Choose

Output

from the panel buttons.

Check the

Tracked ASCII (TRC) or Tracked Binary (TRB)

check box on the Output sub-panel.

Type a file name in the name box and press

Enter

.

Set the Duration (seconds) between

10

and

20

.

Start collecting the range of motion data of the subject by having the person stand in the middle of the capture volume with arms extended, palms parallel to the floor with thumbs facing forward, and all markers in full view.

Click

Record

on the Output sub-panel.

The subject must stay in an initial frozen position for two or three seconds.

After standing frozen in this initial position for up to five seconds, the person must move through a complete range of motion by moving and twisting, ensuring that each linkage exhibits the full extent of stretch that will be experienced during subsequent motion capture sessions. Exaggerated motion must be avoided and all markers should remain in full view. This step should not require more than fifteen seconds.

After ten seconds passes from the moment

Record

was clicked, the system will automatically stop collecting and tracking marker data.

At this point, a Tracked ASCII (TRC) or Tracked Binary (TRB) file has been generated in the current directory and is ready for editing. Next, the markers must be hand identified according to the marker list built for the subject’s marker set (assuming that the marker list was already defined prior to the motion capture).

9-5

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

11.

12.

13.

14.

15.

16.

17.

18.

Choose

Post Process

from the Mode Buttons.

Click

Quick ID

and identify the unnamed markers according to the conventions described in Appendix C, Marker Sets .

Click

Rectify

. This applies the naming convention across all the frames of data.

Manually cleanup and identify all tracks in this range of the motion file. The template should be defined as at least 75% of the visible frames selected.

Select

Template

.

Click

Create Template

.

Select the appropriate Frames Range:

Current

—the current displayed frame

Selected

—frames highlighted in blue, low to high in dashboard

Visible

—what is displayed across the screen, as a function of the time zoom

All Frames

—all frames

Click

Create Template

.

Figure 9-2. Create Template Dialog

Prop Definition

Selecting Prop Definition creates a

<projectname>.prop

file that is a rigid body measurement of the object. This .prop file can then be selected as one of the "Additional Tracking Objects" in the

Motion Capture > Objects

panel.

9-6

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Multiple Tracking Objects

The problem with tracking many people and props (all Tracking Objects) is that you need a unique project file and template for every combination of things you want to track. The project file has to have exactly the right number and names for the combination of markers that you want to use.

You then create a template for these combinations and load that project when that combination of objects is to be tracked.

With the Multiple Tracking Objects (MTO) architecture, you need to have one project and template for each object. You record the range of motion for each object separately and create a template for each object and save it in each object's

.prj

file.

For example, when Subject1 and Subject2 and the Subject3 are to be brought into the volume to track, you select the Tracking Objects

Subject1.prj

,

Subject2.prj

, and

Subject3.prj

files. The base project file, entered into the

Main Marker Set:

text box, that you load can be a "Calibration Only" project file (recommended for a higher number of objects that are entering and exiting the data set) or a project file with one marker set. You can then place projects with multiple, pre-defined subjects in the

Additional Tracking Objects:

text boxes. These text boxes can hold up

to 5 additional objects. For more information, see the Objects Sub-Panel

section on the following page.

Figure 9-3. Multiple Tracking Objects—Multiple Dancers

9-7

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

Objects Sub-Panel

The Objects sub-panel sets the main marker set and allows you to work with multiple tracking objects (MTOs) while tracking, and bring them into your motion capture data set. A good example of this would be bringing in a second dancer in a dance routine, or a prop such as a golf club when analyzing a golfer’s swing.

Under any condition, it is easier for the software to identify a marker set that has greater asymmetry. You may change the order of the additional tracking objects, as listed in the Objects sub-panel to vary the results of your data set. It has been found that the more asymmetrical of your multiple tracking objects should be put at the top of the Additional Tracking

Objects list.

Figure 9-4. Objects Sub-Panel

9-8

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Main Marker Set

Additional Tracking

Objects

Select Marker Set

Allows you to choose which object data set is your primary data set in the tracking session.

Allows you to add multiple tracking objects (i.e. people, props, etc.) to your tracking session.

Merge Marker Sets

System Objects

Lists and allows you to manage which marker sets are used in the project tracking session.

Merges the marker sets of the MTOs to become one merged project.

System Objects is a feature used to help track objects that are known to be rigid objects. The advantage in using this feature over using a standard objects with a Template is that rigid objects can be more robustly tracked and can have several markers define the 6 DOF Rigid Object coordinate system. This is useful for tracking and identifying props such as swords, baseball bats, and other sport accessories. To use this feature:

1.

2.

3.

4.

Create a project file with just the marker names of the prop. Each prop should have its own marker set and associated project file. There must be three or more markers for rigid bodies to work; five or more markers is recommended. Marker names do not matter when they become tracked as Rigid Objects. Linkages are not needed as Rigid

Objects have implicit linkages between all markers on the object. You can create linkages to help you visualize the prop when it is tracked if you wish but when a rigid object is identified as such, both the markers and the linkages are shown as a purple color.

Create a Prop file. Record some number of frames of the prop (maybe

2-5 seconds), and then load the Tracks into Post Processing. Identify the markers and click the

Create Template

button. Select

Rigid

Object

as the template type. The Create Template changes words to

“Create Prop File". A message pops up telling you that you created a prop file in your current project folder. A second message informs you that "The type will become rigid when selected on the Objects sub-panel and that a Rigid Object Template has been created.

Using the prop file in you local folder, the prop file can be selected as any Tracking Object in the

Motion Capture > Objects

sub-panel.

Props can be selected as normal tracking objects and they take up one of the tracking object slots. Currently the coordinate system of the prop is displayed in the Motion Capture 3D Display. When you record a trb or trc file, the XYZ coordinates of the prop are also recorded. Visually, you can see the coordinate system of the Rigid

Object defined. The coordinate system displayed has the origin at the center of mass of the defining markers and the directions of the XYZ axes are defined parallel to each of the calibration coordinate system axes as defined on frame 1 of the capture.

Create a global System Objects folder. If you want to build a library of props, you can create a folder under the launch folder of

EVaRT

(next to the Sounds folder) and name it SystemObjects (note: no space between System and Objects). Copy any props you want to use into the SystemObjects folder. You will need to quit and relaunch

EVaRT

and then you will see a new section at the bottom of the

9-9

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

5.

Motion Capture > Objects

sub-panel with the heading “System

Objects.” You can select prop files from that list to go into any of the

"Additional Tracking Objects" slots.

Order the objects in the Tracking Objects sub-panel. The priority for identifying objects is A-B-C-D-E and the Main Marker set is identified last. You will want to put the props at the beginning of the list into the top slots. Done this way, it is less likely for the props to be embedded into one of the other human objects.

Advanced

Example Data

Set

This example data set, found in

C:\Program Files\Motion Analysis\EVaRT50\Samples\24_Eagle 5 People

shows a five person capture of 1200 frames of data. Each person has their own project file which defines the marker set template for that person.

The TRB file creation process was done as follows:

1.

2.

The

FiveStars1.trb

was generated from the VC files. All markers are unnamed. This was done while using the

All_Five.prj

project file.

For each person:

a.

b.

c.

d.

e.

f.

g.

Load their project file (for example, load the

GreavesTemplate.prj

file).

Read in the Unnamed marker file (in this case it is the file

FivesStars.trb

).

Saved out to a new TRB file (in this case

FiveStarsGreaves.trb

).

Tracked the data by finding frames where the template identified the markers and used rectify to identify all the markers over all frames. Filled in data where necessary. In the Greaves example go to frame 1, select Template ID, select all frames, and then select Rectify. This will ID all the markers over all the frames.

When fully tracked, deleted all unnamed markers and saved the result to a TRC file.

Used Virtual Marker Join to reconstruct missing markers and saved this back out to the TRC file.

Used the final TRC file to create skeleton for import into Maya.

From the above procedure you can see how the TRB files were used to track visible markers as much as possible. Then all marker reconstruction was done using TRC files.

You can see the difference by loading, for example, the

JohnTemplate.prj

and then loading the

FiveStars1_John.trb

and comparing that to what you see when you load

FiveStars1_John.trc

.

Note:

It is important to keep the project files in a particular order for each time the MTO function is performed for a given data set.

9-10

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Output Sub-Panel

The Output sub-panel is used when the motion capture recording is initiated. It is set by the following procedure:

1.

2.

3.

4.

Choose the output file type (at least 1).

Enter a file name.

Set the file length (in time).

If necessary, set the external trigger mechanism.

Figure 9-5. Output Sub-Panel

9-11

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

Output Files

Settings

OK to Overwrite

Enable External

Trigger

Settings allows you to provide the specifics for your generated output files. This includes the name of your files, the trial number, and the duration of the trial. The duration lets you set the length of the motion capture, so there is little wasted time in the session.

Note:

VC, TRB, TRC, ANB, AVI, and TC files are all associated in the project by file name. If you rename the output files, they may not be recognized and they will lose their association to the project.

This allows you to redo an existing file, once it has been saved.

When using an external trigger mechanism, you will need to check this box for the software to recognize it. You can install an external trigger by plugging it directly into your COM1 port on your Host computer.

Post Trigger Mode

The Output files are the files generated during a motion capture session.

This section of the Output sub-panel allows you to choose which files are to be produced and saved in the project file directory.

When the Post Trigger Mode check box is activated, it enables the software to record the data from the end of the session, backwards based on the capture duration (X) that has been set (i.e. from the end of the data capture to –X seconds). This is useful for captures where there is no defined starting point or event, but the ending is well organized and smooth, and you would like to capture only the final moments.

9-12

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Batch

Processing

Batch Processing gives an overview of the current Tracking, Identifying, and Solving configuration in

EVaRT

.

Pressing the

Run

button will track and/or identify and or solve all of the

VC (raw video) data in a particular directory. This input and output directories are set using the

Setup > Misc

sub-panel.

This type of batch job has been designed to never overwrite any TRB,

TRC, HTR, and HTR2 files. If a batch job is terminated, partial files are detected and pressing run again will pick up where the previously terminated batch job left off.

Figure 9-6. Batch Process Interface

9-13

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

Recording Data

Figure 9-7. Recording Data

After setting up the template, always be sure to save the project file

(

File > Save Project

). Use the following procedure to record a motion capture session.

1.

2.

3.

4.

5.

6.

7.

8.

Select the Output sub-panel.

Select the output file types to be generated.

Enter a name for the file.

Enter a trial number. This is optional and will self increment if multiple trials of the same name are recorded.

Enter the estimated time length for the motion capture recording.

Select

External Trigger Mechanism

if you are using an external trigger.

Select

Post Trigger Mode

if you only want the last portion of your motion capture session. This function allows you to select from the end of the session, working backwards to a specified time point.

Click

Record

. You should see the word RECORDING in large red

letters on the lower right corner of the interface. See Figure 9-7

.

9-14

9.

When recording is complete, click

Load Last Capture

to replay the tracked data that was last recorded. This will automatically send you to the Post Processing mode.

EVaRT 5.0 User’s Manual Chapter 9: Motion Capture Panel

Tracking Strategies and Tips

Speeding Up

Tracking in

EVaRT

Max. Speed

Max. Predictor

Error

Changing the Max Speed to 30 mm/frame and the Max Prediction Error to 10 mm greatly enhances tracking for most normal speed data sets.

While this needs to be adjusted up for high-speed trials, keeping a 3:1 ratio works well. This can cut your CPU load by 50% or more.

Here is what is under the hood, see the

Motion Capture >Tracking

subpanel.

If a marker has NO track history (i.e. a new marker just found on this frame), how big of a sphere do we draw around its current location to look for its continuation in the next frame. This should affect only startup tracks. It can affect performance if it is TOO BIG by making the software check more points than are necessary. If it is TOO SMALL, it will not create contiguous tracks; the tracks will have many holes and lots of unnamed markers.

Max. Predictor Error determines a sphere around the projected (extrapolated) path trajectory into the next frame. If the marker is not found in that projected sphere, it is assumed to have disappeared. This should affect only continuing tracks with a history, which is the bulk of what is being tracked. This can have a big effect on performance if it is TOO BIG. If it is TOO SMALL, tracks will be broken up into smaller path fragments and there will be an excessive amount of unnamed markers.

Bring up the Task Manager and monitor the CPU usage. You can see this with trials you have already collected by selecting the Raw Video Files you collected, click the

Run

button. Change the above Tracking Parameters, press the

Run

button again and see the difference. Check that the

Setup > Misc

sub-panel is showing the

From Raw Video Files

and the

Real Time speed option is selected.

9-15

Chapter 9: Motion Capture Panel EVaRT 5.0 User’s Manual

9-16

Chapter 10

Post Processing Panel

Topic

Editing Tracked Data

Viewing Your Data

Selecting Frames

Unnamed Markers

Post Process Toolbar

Join Virtual

Data Painting

Time Lines

Analysis Graphs

Post Processing Strategies and Tips

Editing Tracked Data

The

EVaRT

Post Process mode allows you to play back and edit tracked data stored in TRB (binary) and TRC (ASCII) files. Markers can be identified if they have gone unnamed during the recording session. Gaps and aberrations in data can be filled or fixed by hand, frame by frame, or by employing mathematical functions across entire sections of a data set. Up to ten operations can be undone if you make a mistake, but it is recommended that you save your work frequently.

Typically, an editing session requires having both the 3D Display and the

XYZ Graphs open at the same time in two different Graphics Panes.

1.

2.

3.

4.

5.

6.

Click

Post Process

from the Mode Buttons.

From the Menu Bar, choose

Layouts > 2 Panes: Top/Bottom

.

Left-click in the Top Pane to select it.

Activate the 3D Display by pressing

F3

on the keyboard of by choosing

Data Views > 3D Display

from the Menu Bar.

Left-click on the bottom pane to select it.

Activate the XYZ Graphs by pressing

F4

on the keyboard of by choosing

Data Views > XYZ Graphs

from the Menu Bar. It will display X, Y, and Z tracked position data, and optionally, residuals, and the cameras that triangulated the markers.

Page

10-1

10-2

10-4

10-10

10-11

10-22

10-26

10-27

10-28

10-31

10-1

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Viewing Your Data

The XYZ Graphs will display none, any, or all markers you have selected in the Post Process panel. Selection of markers from a marker list is done with standard selection methods including

Shift

+ click,

Ctrl

+ click, and

Shift

+

Ctrl

+ click. In addition, the top row of the marker list acts as a special selecting button. Markers can also be selected by simply clicking and

Ctrl

+ clicking on markers in the 3D Display. Click on the back button to reset the previous list of selected markers.

The Post Process tools heavily utilize the Post Process Dashboard controls. These controls are itemized and described as follows.

Figure 10-1. Post Process Dashboard

Play

Play Speed Frame # Lowest/Highest Frame

Frame Selectors

Selected Frames—Low

Visible Frames—Low

Move 1 Frame

Time Code

Selected Frames—High

Visible Frames—High

Current Frame

Play Forward

Button

Play Backward

Button

Next Frame Button

Previous Frame

Button

Low Frame Button

High Frame Button

Time Zoom Slider

Current Frame is the frame that is currently seen in the 3D Display and is marked with a full height red line on the XYZ Graphs. The Current Frame number is found in the very center of the Post Process Dashboard.

The Play Forward button (default hot key is the

>

key.) plays forward through the data until the end and then repeat from the beginning. This also acts as a Stop button.

The Play Backward Button (default hot key is the

<

key) plays backward through the data until the beginning and then repeat from the end. This also acts as a Stop button.

The Next Frame Button (default hot key is the

F

key) moves the Current

Frame forward by one frame. This also acts as a Stop button.

The Previous Frame Button (default hot key is the

S

key) moves the Current Frame backward by one frame. This also acts as a Stop button.

The Low Frame Button sets Current Frame to the Low Visible Frame.

The High Frame Button sets Current Frame to the High Visible Frame.

The Time Zoom Slider sets/indicates the Low and High Visible Frames.

Double-clicking on this control expands the Visible frames to encompass all of the frames in the data set. If you only want to work with a specific range of frames, right-click on this slider to lock/unlock visible frames.

10-2

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Low and High

Visible Frames

Low and High

Selected Frames

Zoom In-Zoom Out

Target Marker

Scaling

The Low and High Visible Frames define the lower and upper limits of the visible frame range. The Current Frame is never outside of these limits. Absolutely no identifying or editing can occur on frames outside of the visible range, with the exception of the join tools. Their values are found in the white Visible Boxes and can be changed by typing numbers into these boxes and pressing

Enter

on the keyboard.

The Low and High Selected Frames are the lower and upper limits of the selected frame range. These values can be changed by typing numbers into these boxes and pressing

Enter

on the keyboard. See

“Selecting

Frames” on page 10-4

for details.

The XYZ Graphs right mouse pop-up menu includes these zooming features (default hot keys are the

I

and

O

keys). These features zoom in time

(frames) centering on the Current Frame. Zooming occurs more quickly by using the hot keys.

The best way to zoom into a particular set of frames is to select the frames in the XYZ Graphs by dragging with the middle mouse and then pressing the Zoom In Hot Key. Even finer control over zooming can be accomplished by pressing

Shift

+ middle-clicking to independently set the Low and High Selected Frames and then pressing the Zoom In hot key (

I

).

The XYZ Graphs also allows you to translate the data vertically and horizontally. This is accomplished by holding the

Alt

key while clicking and dragging the cursor inside the X, Y, or Z display.

It is often helpful to zoom into the data’s amplitude. Holding the

Alt

key and simultaneously pressing the left and middle mouse buttons zooms the data’s amplitude. The marker that was closest to the pixel on the display where zooming began becomes the Target Marker. Its data is centered either to the data in the Current Frame or optionally to the data in the frame that the cursor was on when zooming began. Data for this marker will remain centered on the screen at all times unless you forcibly translate it off the screen using

Alt

+ click and drag.

Unzoom is a means of resetting the display such that zoom and translate values are equal to zero. The default Hot Key is

U

on the keyboard and it is an XYZ Graphs right mouse pop-up menu item.

Picking a marker out of a crowd of data is done by double-clicking directly on a marker’s data line. This action will deselect all other markers leaving only the display of the Target Marker.

Other important view options that are general in nature are described here. These options can only be accessed as XYZ Graphs right mouse pop-up menu items.

Auto Scale

dynamically scales the display to accommodate data in the visible frame range.

Uniform Scale

the display such that X, Y, and Z conform to a uniform range.

Show Residuals and Cameras

shows residuals and cameras along with XYZ data.

10-3

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Selecting Frames

To the right of the Post Process Dashboard controls are five Frame Selector buttons. Refer to

Figure 10-2

.

1.

Select Backward

selects from the Current Frame to the Low Visible

Frame.

2.

Select Forward

selects from the Current Frame to the High Visible

Frame.

3.

Select Current Frame

selects only the Current Frame.

4.

Select Visible Frames

selects from the Low Visible Frame to the

High Visible Frame.

5.

Select None Safe Mode

—nothing is selected.

Figure 10-2. Selecting Frames Buttons

1 2 3 4 5

Joining Gaps in

Data

Filters

Clicking any one of these buttons sets the Frame Selector mode that you can return to at any time by pressing

Esc

on the keyboard. The Frame Selector mode is a User Preference. The default mode is Select Visible

Frames. This will also highlight the selected area in blue.

Join functions are not always confined to selected frames as a conveniencer. For example, if you select only one frame in a gap of marker data, tools intended to fill that gap will seek out appropriate endpoints to that gap, store all necessary data in an undo buffer, and effect a repair to that gap without requiring you to tediously hand select the appropriate endpoints.

The smooth function smooths data within the selected frames with a Butterworth Filter algorithm. This is a low pass (high block), two-pass, 4th order, zero phase shift filter. This data can be spikes—created by frames in which a marker has experienced an acceleration greater than or equal to a selected value, or gaps—missing data.

Manually selecting frames is done by dragging the mouse in the XYZ

Graphs with the middle mouse button pressed. Low and High Selected

Frames can be independently picked by pressing the

Shift

key and middle clicking on the XYZ Graphs.

Select All Frames

(default Hot Key is the

A

key.) displays and selects all frames in the data set. This is also a right mouse menu item on the XYZ

Graphs.

The Eagle and Hawk digital cameras generate extremely clean, noise free data. For the majority of data captures, it is never necessary to modify the data by filtering or smoothing. Occasionally, however, it is useful to remove artifacts in the motion capture data. This can happen in the case of

10-4

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

captures which contain a high number of marker occlusions or a large amount of marker merging (as it frequently happens with face tracking, for example).

For these purposes

EVaRT

provides 3 different smoothing filters that can be applied to tracking data. Each filter affects only the currently selected markers over the currently selected sample range. All three dimensions

(X, Y and Z) of each marker are smoothed. To access the options dialog, select the

Post Process > Options...

button.

Figure 10-3. The Post Processing Options Dialog Showing the Smoothing Options Tab

10-5

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Butterworth

Filter

The Butterworth filter is a low band-pass filter with excellent mathematical characteristics for biomechanical motion. The purpose is to remove high frequency motions (motions that are too fast for a person to actually perform) while leaving intact the frequencies of motion normal to human movement. The user has the choice of selecting how aggressively the filter will smooth the motion by choosing a frequency value. Lower values will cause a very smooth result while higher values will remain truer to the original data.

Figure 10-4. An Example of Original, Unfiltered Data with Some Unwanted Error

Figure 10-5. The Curve After an Application of the Butterworth Filter, with an Input Freq. of 3

10-6

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Figure 10-6. The Curve After an Application of the Butterworth Filter, with an Input Freq. of 6

Figure 10-7. The Curve After an Application of the Butterworth Filter, with an Input Freq. of 12

10-7

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Figure 10-8. The Curve After an Application of the Butterworth Filter, with an Input Freq. of 18

As the previous sequence of images shows, the input value to the Butterworth filter noticeably changes the result. The first example, a frequency input of 3, shows a lot of smoothing applied to the curve while the last example shows very little change to the curve. Depending on your needs, these might be appropriate levels of change. For most purposes, however, values between 6 and 12 work very well as can be seen in the middle two images. In these two examples, the noisy part of the data has been removed while the overall characteristics of motion have been retained.

10-8

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

3-Point Average and 5-Point

Average Filters

Two other filters, the 3 Point Average and the 5 Point Average filter are provided as an alternative to the Butterworth filter. In some circumstances

(particularly with facial data), these can provide better results. The 3 Point

Average smooths the data by taking a data point on either side of a given, original data point and averaging their values into the original one to create a new data value for that sample. This filter provides a moderate amount of smoothing as shown below:

Figure 10-9. The Curve After an Application of the 3 Point Average Filter

The 5 Point Average filter works just like the 3 Point Average filter except that it uses 2 data points on either side of the original data point to produce a new value. Since the width of the filter is wider the results are more aggressive which creates more smoothing, as seen here:

Figure 10-10. The Curve After an Application of the 5-Point Average Filter

The type of smoothing you choose depends on your needs and how much

(and in what way) you want to change your data. It is perfectly reasonable to make successive applications of the filter(s) to affect the data in various ways. The number of possible combinations are extremely high so some experimentation will be necessary to find the right one for you.

10-9

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Unnamed Markers

Unnamed markers are defined as unidentified markers that are either real or not real. Some unnamed markers represent good data, yet were unidentified during a motion capture session. Others are called ghost markers and should be deleted. Ghost markers can also be removed from future data captures by going to the Tracking sub-panel in Motion Capture, and setting the

Min. Cameras to Use

(minimum number of cameras to use) to

3

.

Note:

Caution should be taken here as this process may also eliminate good data.

10-10

EVaRT 5.0 User’s Manual

Post Process Toolbar

Figure 10-11. Post Process Toolbar

Chapter 10: Post Processing Panel

All Markers Radial

Button

Selected Markers

Radial Button

Template ID

Template Rectify

Create Template

Make Unnamed

Rectify Unnamed

All of the Identifying tools are accessible using hot keys, panel buttons, and right mouse menu items on the 3D Display and XYZ Graphs.

Allows you to select all markers at once for Identifying.

Allows you to select specific markers for Identifying.

Uses the template to ID all markers in the current frame.

Uses the template and continuous tracks to ID markers thorough time.

Refer to “Building a Template” on page 9-5 .

Make Unnamed will specifically move a marker’s data into the first unnamed marker slot. It is important to know that the data is not deleted by this operation.

Makes unnamed markers into contiguous paths to follow through the capture sequence. For more information on the Rectify functions, refer to

“Rectify Functions: What They Do and When To Use Them” on page

10-15 .

10-11

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Quick ID

Identifies the selected marker, identifying all markers one by one, according to the list. It will normally select with auto incrementation (Auto Increment).

Marker ID is the same as Quick ID, without the auto increment feature.

Marker ID

Exchange

Hide Markers

Exchanges the XYZ coordinates of two selected markers.

Hides selected 3D view markers.

Unhide Markers

Rectify

Unhides the hidden selected 3D view markers.

Re-identifies missing makers (gaps) in a determined frame range. For more information on the Rectify functions, refer to

“Rectify Functions:

What They Do and When To Use Them” on page 10-15 .

Rigid Body Rectify

Uses the selected markers to ID unnamed markers through the capture sequence. For more information on the Rectify functions, refer to

“Rectify

Functions: What They Do and When To Use Them” on page 10-15 .

Rigid Body Rectify and Template Rectify assume that all the current marker identifications are correct. They are intended for continuing the identification process without undoing previous work.

Rigid Body Rectify is a tool that could be considered a "stand-alone" tool.

It does not use anything from the marker set definition at all. When the tool is activated:

1.

2.

3.

The selected markers are dynamically turned into a "Rigid Body" definition and measured

The previous frame and the current frame are then used to predict the next frame

Identify the frame

This stops when less than three markers of the original selected markers is identified.

Note:

If one or more markers are already correctly identified, then that can help prevent errors.

This has been used to identify the entire body.

2.

3.

4.

5.

1.

Select ALL the markers (minus the obscured ones). The starting frame must be identified manually.

Press

Rigid Body Rectify

Go forward to the frame where the misidentification occurred

Make unnamed

Repeat steps 2 through 5.

Options

Sets the sliders, zoom, and search options.

10-12

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Rectify

Cut

Copy

Paste

Cut Outside

Smooth

Calculate Virtual

Markers

Join Cubic

Join Linear

Join Virtual

Exchange

Search

Used for cleaning up the Initial Pose for making a template when you have no template to start with. Takes ALL markers on the current frame

(regardless of the All vs. Selected radial button), measures the linkages on the current frame and uses those measures to automatically sort markers into the correct marker slots.

Characteristics of Rectify:

Uses all markers, Named and Un-named

Works only on the Highlighted XYZ Selected Time Range

Uses the Named marker linkages and XYZ path continuity

It will switch Named markers (Named markers are not automatically locked)

Adjusts Linkage lengths dynamically to fit the data (including mistakes)

Uses the

Motion Capture > Tracking > Identifying Parameters

function (typical)

Cuts the data within the selected frames inclusive of the endpoints.

Copies selected markers in selected frames.

Pastes data with the Current Frame being the first frame of the paste region.

Cuts the data outside of the selected frames exclusive of the endpoints.

Smooths data within the set frames with the selected filter type. The filter selection is found in the

Post Process > Options

form. For more infor-

mation, refer to “Filters” on page 10-4

.

This calculates the virtual markers based on the parameters set. For more information, refer to “Virtual Markers” on page 11-6 .

Calculates the values to place in the gaps with a cubic spline. If you manually select the endpoints of the gap before executing the join, the function will fill the gap with a linear interpolation because the second derivative at the endpoints equals zero.

Selecting this will automatically fill the gap with linearly interpolated data.

For all information regarding Join Virtual, refer to

See “Join Virtual” on page 10-22.

Exchange requires that exactly two markers are selected. The data is exchanged between the markers within the selected frames.

Finds gaps and/or spikes throughout the data set. The current frame will be set to the first gap or spike found in either the first selected marker on

10-13

Chapter 10: Post Processing Panel

Delete U_n

Options

RB Join

Undo

EVaRT 5.0 User’s Manual

the marker list or all of the markers. See the Post Options form for settings.

Deletes all unnamed markers.

The Options button opens a form that lets you set the sliders, zoom, undo, and search (for gaps and/or spikes) options.

The Acceleration at Spikes function will indicate the frames in which a marker has experienced an acceleration greater than or equal to the selected value. The indicator appears as a carat (V) at the top of the XYZ

Graphs.

The Memory Gauge lets you know when you computer is running out of memory to store edits in the undo buffer.

The rigid body join feature has been created for rigid objects with 4 or more markers per segment. For rigid or semi-rigid objects such as swords, spears, head markers, torso markers, multiple markers on a basketball, it is convenient to use this feature to join across missing marker data. You must select a starting frame where all markers that you select are all present and part of a rigid body. You then select a range of frames on which you wish this to operate. Select

RB Join

and it automatically joins across the missing marker data.

Undo retrieves data affected by the most recent Edit or ID function and places it back into the data set.

EVaRT

supports ten levels of undo. This feature can be disabled or cleared on the Post Options form. If you get the message that an Undo function may not execute, you may need to clean your Undo buffer. This can be found in the

Post Process > Options >

Undo

section.

10-14

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Rectify

Functions: What

They Do and

When To Use

Them

There are three main Rectify tools for naming markers and propagating the names through time (Rectify, Rectify Unnamed, and Rigid Body Rectify). They are not for generating XYZ data from the 2D camera views, which we call Tracking, but they are very useful for Identifying the tracks or markers. Rectify means to “make right” or “set right”. All of the Rectify functions start at the Current frame and go forward in time first, then backwards from the current frame.

Figure 10-12. Post Process Toolbar Buttons

Rectify

Rectify

Used for cleaning up the Initial Pose for making a template when you have no template to start with. Takes ALL markers on the current frame

(regardless of the All vs. Selected radial button), measures the linkages on the current frame and uses those measures to automatically sort markers into the correct marker slots.

Characteristics of Rectify:

Uses all markers, Named and Un-named

Works only on the Highlighted XYZ Selected Time Range

Uses the Named marker linkages and XYZ path continuity

It will switch Named markers (Named markers are not automatically locked)

Adjusts Linkage lengths dynamically to fit the data (including mistakes)

Uses the

Motion Capture > Tracking > Identifying Parameters

function (typical)

10-15

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Motion Capture >

Tracking >

Identifying

Parameters: Details

To Reconsider

If a link stretches more than the set amount, the path is snipped into two paths where the link stretches too much. This happens if markers come together and pull apart and the identity is not correct when they pull apart.

The software may not see it right away, but after a few frames, the linkage for the wrongly named markers get too long and the path is cut. The bigger the number, the more the link is allowed to stretch before it is cut.

Smaller means fewer errors, more cuts. Larger means more stretching is allowed before the cuts. It is measured in multiples of the standard deviation of the linkage length to make it accommodate linkages that normally change a lot (head to shoulder) and linkages that do not change much (elbow to wrist). Also used in Real Time streaming (Run) mode.

Figure 10-13. Linkage Stretch Parameters

10-16

Max Acceptable

If there is a missing marker and there is an Un-named marker within this distance of known linkages, the Un-named marker is accepted as a

Named marker. Also used in Real Time streaming (Run) mode.

Template Rectify

In Post Processing, use this when you have a reasonably good template.

Work from known good frames into unknown and difficult frame ranges.

Uses the template from the Create Template item. A template can be of one or more frames, should represent characteristic motions to be seen between markers, and is a measure of the min. and max. linkage lengths for

Named markers.

The characteristics of Template Rectify are as follows:

Uses only the Template information to move markers from the

Unnamed slots to the Named slots

Uses All Markers or Selected Markers according to the setting, as

shown in Figure 10-13 .

Works only on the highlighted XYZ selected time range, starting on the current frame going forward, then backward from the current frame

Protects all named markers, will not switch them

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Works only to move Un-named markers into the Named marker slots—here, all markers are locked

Does not use Linkage Stretch Parameters in

Motion Capture >

Tracking

Rigid Body Rectify

In Post Processing, it is used when you have bodies crashing into each other where linkage lengths can get very distorted and Template Rectify can give results which may require some editing. Rigid Body Rectify is for very tough trials where you can have the software look for a rigid body with markers on it and it tends not to make mistakes and hence does not require a lot of editing after the fact. Work from a frame with known marker IDs to difficult areas. A Rigid Body can be any set of 3 or more

Selected Markers. The software measures the markers with respect to each other and looks for this pattern in the Un-named marker to automatically assign them names. It will stop if either of two conditions are met:

1.

2.

It cannot find at least 3 of the markers of the Rigid body on a frame

(so use more than 3 marker if you can)

The measurements stretch too much

It can be re-started on a new frame if needed. The selected markers must be identified on the Starting Frame.

The characteristics of Rigid Body Rectify are as follows:

Measures all markers in the selected Rigid Body on the Starting

Frame

Uses All or Selected Markers (mostly used for Selected Markers)

Protects all named markers, will not switch them.

Works only to move Un-named markers into the Named marker slots—here, all the markers are locked

Does not use Linkage Stretch Parameters in

Motion Capture >

Tracking

Rectify Unnamed

Rectify Unnamed sorts the Un-named slots into continuous paths based on path continuity, similar to the tracking function. No templates or linkages are used. Path segments separated by 10 or more frames are considered to be separate paths and will not necessarily be continuous. This is used to clean up the Unnamed path segments and can make the Marker ID function work more smoothly. It means you may not need the Rectify Always On check box in Marker ID and Quick ID Items. First try the Rectify Unnamed function, then try the Marker ID function for the problem areas.

10-17

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Template ID

Uses the currently defined Template from Create Template to fit the linkages into the current frame’s marker cloud. If it succeeds, it tells you the number of tests on the Status Bar (lower left side of the screen) it took to complete. It may also fail or time-out, in which case you should make sure all of the markers are present on the current frame or perhaps remake your template. It only changes data on the current frame if it is successful. You can then use Template Rectify to get the correct IDs to other frames.

Figure 10-14. Template ID Details

How good is the Template?

You can tell how good your template is by how many tests it takes for the Template ID function to work. A low number means it is working quickly (low is say 500). A high number (5,000 or more) means the software has to try very hard to ID the current frame. It stops trying after about 50,000 tries in Connect (Live) mode and “times out” and gives the message Template ID: Timed out. In Post Processing mode, it will not time out until 500,000 tries. That means it did not get to try out all possible linkages. A large number indicates a potential problem. It might be because the current frame is stretched beyond the template or in a much different actor position, or it might be because the

Template is not very good. Try to use the Template ID on different frames and see if the number changes a lot. If you consistently get large numbers for the Template ID feature, try adding more links to your marker setup in the Model Edit panel.

Make triangles, especially triangles that will be fairly rigid during the movements. Lots of rigid triangles in your linkages make for solid Templates that have fast Template ID numbers. Triangles with equal sides can cause mis-IDs whereas triangles with unequal sides work much better.

This will help to determine where to place markers on a person.

You must have all of the markers present on the current frame or you will get the status bar message

Template ID failed

.

That means it did try out all possible linkages and could not get a match. Possibly the markers have moved or you need a better template.

10-18

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Summary of When

To Use What

Rectify

First, collect the trial from which you will make your template. It should be a simple trial with representative motions and not require any editing.

The goal is to represent the min. and max. of each linkage in your model.

Use Quick ID, then Rectify. If editing is required, consider taking another trial. Make sure there are no marker switches using the 3D and XYZ view, and then create a template using all of the (good) frames. Rectify may generate marker switches and you do not want those remembered in the template. It will haunt you later.

Template Rectify

For most motion trials, the use of Template Rectify will do most of the work in correctly identifying the marker tracks. Template Rectify is preferred method to use to rectify since it protects the named marker tracks: they are locked. It also keeps the template rigid and seems less likely to make mistakes. Correcting mistakes can take a lot of time. If there are some incorrectly ID-ed markers, it may be best to make all the markers

Unnamed for all except the starting frame (which can be frame 1 or any other frame). To make all unnamed except frame 1, go to frame 2, select

All Markers,

then select the time range: Select

Forward

, and press

Make

Unnamed

under the Identifying tab. Go back to frame 1 and press

Template Rectify

. On some complex trials where there is a crash or bang between people and or props, have the actors start in the T-pose, do the actions and return to the T-pose. You can then work the data from the start to the middle and also from the end to the middle. This working the data back and forth can save a lot of your time and not require that you hand

ID many frames.

More on Templates and Template ID

After a big crash of people with markers or extreme movement, the markers may have moved and the template may not be as good. You can Extend the template by ID-ing after the crash, extend it based on one or more frames that have no mistakes, then try Template ID to see if the template holds on the new frames. Template ID tells you how good the template is working and it is the same template the Template Rectify uses. Also, think that Template Rectify gets used as much to un-identify tracks as well as identify them. If any link stretches beyond the allowable range (plus a small amount of give), then Template Rectify will cause the offending marker or markers to become unnamed, while it seems perfectly obvious to you what is right. If Template Rectify does this, manually ID that marker on that frame and Extend the Template (an option under the Create Template button). You can see this when the subject bends over and causes stretches that might not have been recorded when creating the template.

10-19

Chapter 10: Post Processing Panel

Figure 10-15. Template ID and Rectify Used for Link Stretch

EVaRT 5.0 User’s Manual

Rigid Body Rectify

This takes the most time to use in that you must select a few markers and process a few frames, then repeat the process over the possibly many segments and frame sequences. If it stops, it means that the rigid body disappeared. You can re-start it again and it starts with new measurements on the starting frame. Where it is very useful is where it can slug through some tough sequences. Start with something simple, like the 3 or 4 or 5 markers on the head. As with Template Rectify, it may be best to Make

Unnamed all markers on all frames if there are any mistakes in the ID-ed data. If one of the head markers is incorrectly ID-ed as a neck marker, then Rigid Body Rectify will not see that as a candidate since it only looks in the Unnamed markers list. Process the data from both ends towards the difficult part, assuming there is a “crash” in the middle and clean data on both ends. After doing Rigid Body Rectify from the starting frames to the middle, then from the ending frames to the middle, use Template Rectify to go it again. For difficult data trials, this 1-2-3-4 combination will get you a lot of named markers for very little work.

Correctly

Identifying

Markers

Automatically:

Post Processing

Mode

Max Speed (mm/ frame)

In the

Post Process

panel, the data is identified (or re-identified) by pressing the

Template Identify

button. This affects the current frame only and it is successful if:

1.

2.

The template is good and the data fits.

All markers are present.

To continue the correct identification to successive frames, you need to have the Max. Speed and Max. Prediction Error settings correctly set for your data.

This applies to when a marker first appears and is identified with the current template. To keep the correct identity into the next frame, the software checks to see if it has moved too much to be the same marker. It can move in any direction. The Max Speed parameter tells how much movement is allowed. It is measured in mm from the first frame, hence the units of mm/frame. If no marker is seen within this search sphere, then the

10-20

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Max. Prediction

Error (mm)

target identify is not continued into the next frame. If a marker is found within the Max. Speed sphere, the target identity is continued into the second frame. If the number is set too small, tracking will slow down as the software tries unsuccessfully to find continuations of markers. This affects the first to second frame tracking time especially. If the number is set too big, you will see markers switch identities.

After a marker has a history of 2 or more frames of continuous identity in time, a track history allows the software to predict where the marker should be, based on a 2nd degree polynomial prediction. The software looks in a search radius of the Max Predictor Error about the predicted location for a continuation of each marker being tracked. The Max Prediction Error is usually set to about one-half of the Max. Speed parameter.

10-21

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Join Virtual

Join Virtual is an extremely powerful editing tool used to fill gaps in marker data with simulated data based on the relationship (positional interpolation) with other markers on or near the particular problem segment. This accurately simulated information is a result of making four passes over the data in both Real Time and Post-Processing modes.

To use the Join Virtual function:

1.

Find a gap within the position data of a marker. It is easiest to use the

Search function (right click in XYZ Graphs and select Search).

If you are not using the Search function, select the gap area in the

XYZ Graphs (middle-click and drag) for the problem marker.

2.

Select

Join Virtual

in the

Post Process

panel. Verify that the Marker to Join is set for the marker you want to edit.

Figure 10-16. Join Virtual (Virtual Marker) Definitions

Real Time

Streaming with

Join Virtual Fill

3.

Click on

Origin Marker

. It will be highlighted in blue and will allow you to choose which marker to use as the marker that is most rigidly

attached to the marker to join. See “Origin Marker” on page 10-23 .

Note:

Select markers using the 3D Display or the marker grids.

4.

5.

The function will then automatically jump to the Long Axis Marker input box. Continue selecting the proper markers for each remaining input box. Make sure they are all different. You cannot have two of the same markers in the Virtual Marker Join definitions.

Once you have defined the three definition markers for you Virtual

Marker, click on

Join Virtual

.

Note:

If data is missing for any definition marker (Origin Marker, Long Axis

Marker, Plane Marker) in the frame field, the gap in data will not be completely filled. You will need to select a different definition marker that has data for that frame field.

6.

7.

Repeat

step 1 through step 5

for all problem markers in your data set.

You may also setup the Virtual Marker definitions for as many markers as you feel will be needed, prior to capture.

Selecting File > Save Project will save all Virtual Marker definitions you have set into the project file.

If you will be continuing to capture motion using the same template, the

Virtual Marker Join definitions are now resident with the project files and template. The Join Virtual check box can now be activated (on the Real

10-22

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Time Dashboard) allowing for Virtual Markers to be created in Real Time where data is missing for the markers you have set definitions. Thus, streamlining the editing process or post processing tasks.

Figure 10-17. Join Virtual Check Box

Origin Marker

Long Axis

Marker

Long Axis Marker

Example

Plane Marker

The concept behind the Join Virtual and the Virtual Marker definitions are the same and are much more stable and more useful than the classic Rigid

Body data filling mechanisms. The reason is that you get to choose three markers in decreasing importance that determine the replacement data.

These three markers are:

1.

2.

3.

the Origin Marker the Long Axis Marker the Plane Marker

The Origin Marker should be the marker that is most rigidly attached to the marker to join. If there are two choices, pick the one that is more stable on the bone segment. For example, the elbow marker is a good Origin

Marker. It is usually attached close to a bone. The shoulder is also good for the upper arm segment, but not as good for the upper torso if the subject raises their arms.

For segments where you have multiple markers on a rigid segment, such as the head, it does not matter which marker is which. For example, if you have four markers on the head, each of the four can be defined by any order of the other three markers. But if you have only three markers on the head, the Top_Neck marker may well be used as the Plane marker for the Join Virtual definitions.

The Long Axis Marker defines a straight line from the Origin Marker and the Join Virtual is not sensitive to changes in the length of this line.

For the left Biceps, choose the L_Elbow as the Origin, the L_Shoulder as the Long Axis Marker, and the L_ Wrist as the Plane Marker.

The Plane Marker is the least strongly coupled marker to the problem marker to Join. It defines only the rotation of the coordinate system located at the origin. Join Virtual and Virtual Marker calculations use the

3D offsets from the problem marker to the coordinate system’s Origin

Marker and apply that throughout the Join Virtual.

The results are often astoundingly good and can be used directly to speed up your animation pipelines. More study is recommended before applying these results to Biomechanics research and medical applications.

10-23

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

In short:

The

Problem Marker

is the marker with the gap to be filled.

The

Origin Marker

maintains a fixed distance to the problem marker.

The

Long Axis Marker

defines a line to the origin marker.

The

Plane Marker

defines a plane with the Origin and Long Axis

Markers.

Figure 10-18. TRC File vs. TRB File With Join Virtual

TRC file—no Join Virtual

TRB file from same data set with Join Virtual

10-24

To test the operation of this feature, define a Virtual Marker Join for the

RHip marker based on 3 others that will remain visible. Cover the RHip marker (for example) and see if it appears in the 3D view. This can be done with a live person very easily, but if you do not have a setup available, go to the 2D view of a trial, mask out enough regions and cameras so that the RHip is no longer visible. It should appear in the 3D view if you turn on the Join Virtual feature.

Both the streaming and the post-process Join Virtual use a two-pass process to virtually join data across gaps. The data passes through the Virtual

Marker Join function twice, with the second pass using filled or partially filled gaps that were not filled the first time.

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Join Virtual

Guidelines

Head

Upper Arm or

Upper Leg

Hand or Foot

Note:

These guidelines are intended for an audience with a good knowledge of motion capture theory and practice. These are generalized guidelines only. Individuals may find that different definitions may work better for their particular applications.

For best results, it is recommended that you have at least three markers per effected segment. Ideally, for markers that have the possibility to become obscured (i.e. being covered up or lost between the ground and the subject’s body) you will need to place markers on the opposite side of the appendage or body. For example, if a subject is laying prone on the floor, the back markers become obscured. If you anticipate this, you can apply more markers to the chest or front torso area.

For defining virtual markers, when possible, define and use markers that are always seen on that segment or neighboring segments.

If any data is missing from other markers in that segment, the original data will improve, but only if the dependent markers are present.

A subject’s head usually will have four or five markers. Missing marker data for the head is joined using Join Virtual definitions with any three of the other markers.

Typically the upper arm segment is defined by three markers: Elbow, Biceps, and Shoulder. Ideally, a fourth marker on the Triceps would be present. If not, a marker on the forearm can be used as the plane marker for Join Virtual data.

If the shoulder data is missing, you may use the markers for the top of the neck, sternum, or mid-back to calculate the Virtual Marker data for the shoulder.

If you know of movements that are going to obscure the markers on a hand or foot, you may want to set redundant markers on that particular segment.

For example, if all lateral side markers on a foot are obscured from a subject laying down in a prone position, you may apply redundant markers to the medial side to provide the data for that segment.

The hand will use the same technique, with maybe a few less redundant markers on the opposite side.

For an example of a project with Join Virtual definitions for all markers, see the

6Eagle_VirtualJoin

directory in

C:\Program Files\Motion Analysis\EVaRT50\Samples.

10-25

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Data Painting

Data Painting allows you to directly manipulate the data in the XYZ

Graphs pane. Simply press

Ctrl

+

Shift

and left-click to modify or add data directly on the screen.

Figure 10-19. Data Painting

10-26

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Time Lines

Figure 10-20. Time Lines

In the Menu Bar, select

Tools > Time Lines

. Time Lines provide a general overview of the quality of the data in the marker slots, showing any breaks in the stream of data for all markers.

10-27

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Analysis Graphs

The Analysis graphs and their related control panel provide tools to analyze your realtime and post process data. This window has three tabs which calculate data for the following:

Position, velocity, and acceleration

Distance between markers

Included angles

The Analysis graphs are activated by pressing the

F7

key.

Position,

Velocity, and

Acceleration Tab

This tab creates graphs of the position data, calculated velocity data, or calculated acceleration data for up to 10 selected markers.

Note:

Any number of marker data can be exported.

The number of frames used to calculate the velocity and acceleration data is set by the user. The number of frames used can be either 3, 5, 7, or 9.

Using the higher number of frames to calculate the data will result in smoother output through noise reduction.

Figure 10-21. Position, Velocity, and Acceleration Tab

10-28

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Distance

Between Two

Markers Tab

The Distance Between Markers tab shows the distance between two selected markers for each frame throughout the tracked data. You may select any two markers in the tracked data to be analyzed by clicking those markers in the 3D Display or on the marker list grids.

To delete a pair of markers from the grid, click on the row and press the

Delete

key. You may select and delete several rows at once by pressing

shift + click

on the rows and pressing

Delete

.

Figure 10-22. Distance Between Two Markers Tab

10-29

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

Included Angles

Tab

The Included Angles tab allows you to select groups of four markers, which define two separate lines in space. Between these two lines, the angle is calculated for each frame through the tracked data. This information proves useful for detecting irregularities in movement, such as between two parts of a body. To delete a row, simply click on that row and press

Delete

. You may select and delete several rows at once by pressing

shift + click

on the rows and pressing

Delete

.

Figure 10-23. Included Angles Tab

Exporting

Analysis

Information

To export an ASCII text viewable file, select the

Export.ts (Time Series)

File...

button. This will write a file with the same file name as the tracked file you are working on. It will contain the information from all three

Analysis tabs.

10-30

EVaRT 5.0 User’s Manual Chapter 10: Post Processing Panel

Post Processing Strategies and Tips

Note:

The steps necessary to clean up data will vary significantly from one user to another. These tips are guidelines that outline a general approach to successful editing sessions.

It is best to start identifying from Frame 1 forward. Then, identify from the last frame backwards. This entails naming unnamed markers using the Marker ID and Quick ID tools.

Use of Rectify over small frame ranges may help in cleaning the data by taking unnamed data into named tracks.

After identifying, the Post Process tools can be used to fill gaps and cut out unwanted data sections, fix abnormalities, and smooth anomalies, and can be used to exchange switched markers.

It is recommended to save your files often, especially when performing heavy edits.

Many users will select all markers and all frames and execute a Join

Linear or Join Cubic, and possibly a Smooth as a very last editing step.

Learning and using Hot Keys is critical to high productivity.

10-31

Chapter 10: Post Processing Panel EVaRT 5.0 User’s Manual

10-32

Chapter 11

Model Edit Panel

Topic

Overview

Markers Sub-Panel

Tree View Sub-Panel

Virtual Markers

Virtual Marker Quick-Start Example

Page

11-1

11-2

11-5

11-16

11-20

Overview

The Model Edit panel provides tools to build and modify the model parameters that are mandatory for the project file. These parameters include markers, virtual markers, linkages, and segments.

Note:

It is important to save your project after building the model by selecting

File > Save Project

. For more information about project files (

*.prj

), refer to “PRJ—EVaRT Project File” on page G-4 .

11-1

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Markers Sub-Panel

The Markers sub-panel is intended for building and modifying marker sets.

Figure 11-1. Markers Sub-Panel

Clear Marker Set

Button

This button clears out the project’s marker set.

11-2

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Create Markers

This is done by double-clicking in the first box or by right-clicking and choosing

Insert

.

Select and Edit

Button

Create Linkages

Button

This allows you to select the names and colors of the markers in the marker set.

This button must be clicked prior to creating linkages on the 3D display.

Linkages can be built by connecting the dots. Linkages should reflect the rigid segments.

Draw 3D Points

Button

This allows you to draw points manually onto the 3D display.

Note:

You must be in Motion Capture before going into Model Edit to select

Draw 3D Points

.

Figure 11-2. Draw 3D Points

First select Motion Capture before going into Model Edit to select Draw 3D Points.

Step 1

Step 2

Step 3

Manually drawn 3D points

11-3

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Quick ID Button

Marker Names

Delete Key

This identifies the selected marker and steps through the listed markers one-by-one. It will normally select with auto incrementation (Auto Increment).

Marker names are accepted in the markers grid when you press

Enter

on the keyboard.

The Delete key will delete the current selected item from the marker set.

11-4

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Tree View Sub-Panel

The Tree View sub-panel provides an overview of the primary elements of the model and allows you to reorder the markers in the marker set by dragging and dropping. You can also insert and delete markers as well.

Figure 11-3. Tree View Sub-Panel

V-Marker

Definition

Button

This button opens the Virtual Marker Definition form. For more informa-

tion, refer to Virtual Markers on the following page.

11-5

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Delete All

Linkages Button

This deletes all linkages in the model.

Project

Name

Selecting the Project in the Treeview allows you to make changes to the current project file. The Project property values that can be changed include the following:

This is the marker set name you can display over the marker cloud. The toggle function for this is found 3D Display right-click menu. The default is set to the current project name.

Skeleton Engine

This provides a way to select which skeleton engine will be used to create the bone structure (if used). The choices are:

None

Skeleton Builder (SkB)

Calcium Type Skeleton—Note that this can have an optional MOD file associated to it.

SIMM OrthoTrak Skeleton—Note that this requires an associated

JNT file and an Init or T-pose trial.

Skin File

Skin files are rigid shells that do not scale with different sized subjects, and do not span/stretch across joints. See Talon Viewer for these capabilities.

The Skin File function allows you to select one of four pre-defined skins that work with

EVaRT

. These skins that are defined for two different skeleton types and are located in the

C:\Program Files\Motion Analysis\EVaRT50\User Files\Skins

directory. The four skin types are:

1.

2.

3.

4.

OrthoTrak Male (

MaleOTSkin.obj

)

OrthoTrak Female (

FemaleOTSkin.obj

)

OrthoTrak Polygons (

PolyBonesOT_Skin.obj

)

25 Bone Animation Skin (

25_Bones_Male.obj

)

If you are looking to develop an entirely new skin file, you will need to contact Motion Analysis Customer Support ([email protected]) for information.

Skin Transparency

This sets the transparency attribute of the skin—100% means the skin is invisible, 0% is solid.

Figure 11-4. Project Property Values

11-6

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

ShowIng the

Skin

An OBJ skin consists of two files:

1.

2.

A skin file:

<Skin>.obj

An associated base position:

<Skin>_Base.htr

The HTR and OBJ file must have "group" names that match the base position HTR’s segment names. The base position HTR file must have segment names that match the marker set. The order of the names doesn't matter. The matching is done with the currently selected skeleton engine.

To select it:

1.

2.

3.

4.

Go to

Model Edit >TreeView

.

Select the project name (first line of the tree).

Select the Skin File property (at the bottom of the sup-panel).

Select a file in the Open File dialog.

The filename is saved in the project file, so each marker set can include a skin.

Note:

The skin file and the associated HTR file is not saved in the project file.

Only their names and relative directory paths are saved.

11-7

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

OrthoTrak Example

The following is an example of how to get the OrthoTrak skins to show in the 3D display.

2.

3.

4.

5.

1.

6.

Select

File > Load Project...

and load the project file

Walk.prj

located in the

Samples\Helen Hayes Markers

directory.

Select

File > Load Tracks...

and load

Walk1.trb

.

Click on the

Bone Button

in the lower-left of the

EVaRT

interface.

Go to the Model Edit panel and select the project in the Treeview.

Select the Skin File in Property Value and select

PolyBonesOT_Skin.obj.

Right-click in the 3D display and select

Show Skins

and

Show Skeleton

.

This procedure will produce a subject with a skeleton and skin as shown in

Figure 11-5 .

Figure 11-5. OrthoTrak Skeleton and Skin Subject

11-8

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Animation Example

The following is an example of how to get the 25 Bone skin to show in the

3D display.

2.

3.

4.

5.

1.

6.

Select

File > Load Project...

and load the project file

Calcium

Solver.prj

located in the

Samples\Talon Viewer Calcium and SkB

directory.

Select

File > Load Tracks...

and load

DaveUmpOut.trb

.

Click on the

Bone Button

in the lower-left of the

EVaRT

interface.

Go to the Model Edit panel and select the project in the Treeview.

Click on the Skin File in Property Value and select

25_Bones_Male.obj.

Right-click in the 3D display and select

Show Skins

and

Show Skeleton

.

This procedure will produce a subject with a skeleton and skin as shown in

Figure 11-6 .

Figure 11-6. Animation Skeleton and Skin Subject

11-9

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Markers

Selecting Markers in the Treeview allows you to make changes to any of the markers associated with the current project file using the property value selections at the bottom of the sub-panel. You can also insert, delete and select a range of markers.The marker values that can be changed include the following:

Displays and edits the name of the selected marker.

Name

Index

Displays the marker number, in the order the marker appears in the marker list. This is not editable.

This is not used at this time.

Size

Color

Displays and edits the color associated to the maker in the 3D display. To change the color, click on the color property and select from the dropdown menu.

Weight

This is not used at this time.

X, Y, and Z Values

The 3D coordinates of the marker in calibration units at the frame number where the marker is selected. It is not updated with every frame change, but is updated when you select the marker.

Figure 11-7. Marker Property Values

11-10

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

VMarkers

Selecting VMarkers in the Treeview allows you to make changes to the virtual markers in the current project file. The VMarkers property values that can be changed include the following:

Displays and edits the name of the selected VMarker.

Name

Index

Type

Origin Marker

Long Axis (Y)

Displays the VMarker number, in the order it appears in the list of

VMarkers for the project. This is not editable.

This provides a way to select which type of virtual marker to be created:

1.

2.

3.

4.

Two-Point (Ratio)

Two-Point (Value)

Three-Point (Ratio)

Three-Point (Value)

Allows you to select and edit which marker is the Origin Marker of the

Virtual Marker definition. To edit, click on the property and select from the drop-down menu.

Allows you to select and edit which marker is the Long Axis (Y) of the

Virtual Marker definition. To edit, click on the property and select from the drop-down menu.

Plane Axis (XY)

Allows you to select and edit which marker is the Plane Axis (XY) of the

Virtual Marker definition. To edit, click on the property and select from the drop-down menu.

X Offset

Y Offset

Sets the X coordinate of a VMarker definition.

Sets the Y coordinate of a VMarker definition.

Z Offset

Sets the Z coordinate of a VMarker definition.

Figure 11-8. VMarker Property Values

11-11

Chapter 11: Model Edit Panel

Links

Index

Marker1

Marker2

Color

Extra Stretch

EVaRT 5.0 User’s Manual

Selecting any of the links in the Treeview allows you to make changes to the links in the current project file. The link property values that can be changed include the following: Right-click can only delete.

Displays the link number, in the order the link appears in the list of links for the project. This is not editable.

Allows you to select and edit which marker is the first end point of the link definition. To edit, click on the property and select from the dropdown menu.

Allows you to select and edit which marker is the 2nd (of two) end point of the link definition. To edit, click on the property and select from the drop-down menu.

Displays and edits the color associated to the link in the 3D display. To change the color, click on the color property and select from the dropdown menu.

The Extra Stretch factor is a statement about the quality of the motion that was used to create the template values. The Extra Stretch is a "Confidence

Factor" in the data used to make the template. What confuses the user most is that rigid body parts (the head is a good example) will not have their markers move around much no matter what data you use to create the template. So almost any template will work to give you values with high confidence. Conversely, if you have links that are stretchy, it's easy to not get a good template for it and so your confidence in the template will be less.

Because of this, many users think that the Extra Stretch factor is a statement about the stretchiness of the link, which it isn't. A link could be put between a hand marker and the toe marker but as long as it had sufficient data for the template Extra Stretch could be set to 1 (the lowest) and have it work just fine.

This is why there is a

CreateTemplate.sky

script. It shows how to use multiple TRB files as input to the template creation (and extension) process.

Assuming a standard Range of Motion (ROM) file, the Extra Stretch values are normally set as follows:

10 for the head

15 for the links on the hips/pelvis

15 for the feet

20 everywhere else

This is a starting point. As data is tracked, you can fine-tune the Extra

Stretch values.

In order to maintain compliance with legacy datasets, the Extra Stretch factor is not used in the same way in the Motion Capture panel as it is in

Post Process. In Post Process, the Extra Stretch is used in all operations

11-12

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

General Notes on

Extra Stretch

that use the template (TemplateID and Template Rectify). In Motion Capture, when processing data live from the cameras or from VC files, the

Extra Stretch is used as part of the TemplateID but not as part of the Template Rectify. For the rectifying process the Linkage Stretch parameters are still used to specify how to treat the data in the main marker set.

However, all additional tracking objects in slots A, B, C, etc... use the

Extra Stretch factor for all Template operations (just like in Post Process).

Only the main marker set is treated differently. The reason for the difference is the way in which the template has been used in the past during live capture (and still is for the main marker set). The template definition is used as a starting point for the rectifying process. The TemplateID identifies the markers using the Extra Stretch factor and then the template is modified on a frame by frame basis called the "running template". The running template is essentially the original template with an extend template operation performed on the template for each frame. This is why the

"Reset ID" button doesn't always work - if bad (swapped) data gets into the template then that information is encoded into the template. The only way to completely reset the running template is to Pause and then Run.

The running template continues to be modified for as long as the system is running without pause. Without any kind of check on this process the template could get extremely different than the original template. The

Linkage Stretch parameters put a brake on the process. They specify how different the running template is allowed to be. In future versions of

EVaRT

the Linkage Stretch parameters in the Tracking Parameters interface will become obsolete when the main marker set uses the Extra

Stretch factor just like it is used everywhere else. By then it is expected that users will have modified their marker sets to take advantage of the

Extra Stretch factor and so the transition to the new technique will be seamless.

Extending the template works exactly as if you took all the TRB data used to create the template and put it end to end and did a Create

Template with the whole works at once.

If the ES factors are too large you get misidentified markers.

If the ES factors are too small (and your template isn't complete enough) you will get unidentified markers.

The template is a pair of Min/Max values for each link. These values only get farther apart as you extend the template. If they are too far apart (as might happen if you had bad data to create the template) then you must start the template creation process over from the beginning because there's no way to tell the numbers to get close together

(again, this is why I made a Sky script for this - it makes it trivial to redo the template creation process).

11-13

Chapter 11: Model Edit Panel

Figure 11-9. Link Property Values

EVaRT 5.0 User’s Manual

SkB Segments

Name

Index

Parent

Origin Marker

Long Axis (Y)

Plane Axis (XY)

RX Offset

RY Offset

RZ Offset

Displays and edits the name of the selected SkB segment.

Displays the SkB segment number, in the order the segment appears in the list for the project. This is not editable.

Displays the parent segment of the selected SkB segment. To edit, click on the property and select from the drop-down menu.

Allows you to select and edit which marker is the Origin Marker of the

SkB segment definition. To edit, click on the property and select from the drop-down menu.

Allows you to select and edit which marker is the Long Axis (Y) of the

SkB segment definition. To edit, click on the property and select from the drop-down menu.

Allows you to select and edit which marker is the Plane Axis (XY) of the

SkB segment definition. To edit, click on the property and select from the drop-down menu.

RX is used to rotate the bone in the SkB segment along the X axis. RX is not used very often compared to RY. If you select a segment to rotate, it will bring up the rotation gizmo.

RY is used to rotate the bone along the Y axis. If you select a segment to rotate, it will bring up the rotation gizmo.

RZ is used to rotate the bone in the SkB segment along the Z axis. RZ is not used very often compared to RY. If you select a segment to rotate, it will bring up the rotation gizmo.

For more information on Skeleton Builder, refer to the Skeleton Builder

Quick-Start Guide found in the

C:\Program Files\Motion Analysis\EVaRT50\Samples\Skeleton Builder

directory.

11-14

EVaRT 5.0 User’s Manual

Figure 11-10. Skeleton Builder Segments Property Values

Chapter 11: Model Edit Panel

The various skeleton types are described in Chapter 13, Skeleton Types.

Calcium

Segments

Calcium is the graphical user interface to the Solver engine. Solver is the powerful numerical tool for calculating skeleton motion from marker data. The Calcium interface in

EVaRT

is what allows you to correlate the positions of a marker pose to the initial pose of a skeleton. The skeleton is usually created in an outside animation package, such as Maya, 3D Studio

Max or Kaydara and then exported to an HTR file by a Motion Analysis file IO plugin for that package. In this example we're using a skeleton from a Maya character.

Figure 11-11. Calcium Segments Property Values

Units

Global Scale

Matrix Method

Units of the Calcium segment lengths. Select from the drop-down menu in meters, centimeters, and millimeters.

Changes the scale of the entire hierarchy, multiplied by the number set

(e.g. a value of 10 would scale the hierarchy by 10 times the original size). This provides a quick method for scaling the HTR file to fit the marker cloud in the model pose.

There are two matrix methods to choose from:

Gauss Newton

—Which is faster at solving, but not as robust. This is generally used in Real Time applications.

11-15

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Accuracy

Levenberg-Marquart

—While is more robust, but not as fast. This should only be used in PP mode.

This is the accuracy parameter for the solve. It is generally set to .0001

and then left alone. It can be useful to debug and troubleshoot the Calcium segment.

Max Iterations

Use Joint Limits

Orient Body

This is the number of iterations the solver goes through to solve. When the solver gets "stuck", it can potentially iterate forever. Usually the solve happens in a very small number of iterations (1-5). Setting it to 100 is more than enough.

Enables or disables the use of joint limits in the Calcium model. Generally it is recommended to be turned off for animation applications. If the solution has joints flipping around, turn it off. Any model created from a joint file should have the limits enabled.

This is important on the first frame of any solve. First orient the root bone to the root bone markers, then do the solve. This helps to eliminate some first frame errors when bones get oriented incorrectly.

Note:

There is still a bug where joints get turned around on the first frame. The solve changes randomly whenever these last two flags are changed. User beware.

It is recommended that Orient Body is set to False when Real Time operation performance is a factor.

For complete information on Calcium software and Calcium segment definitions, please reference the Calcium for EVaRT Quick-Start Guide (p/n

651-1920-010).

Virtual Markers

Virtual Marker

Definitions

Virtual markers are markers that get their position from a combination of the position of two or three actual markers in the motion capture data.

Typically, a virtual marker is used to generate the actual joint center position of the performer (or subject) being motion captured. This is necessary since the actual markers lie on the outside of the performer. Joint center markers are desirable for use with analytical and skeleton reconstruction tools.

There are two methods for defining Virtual Markers (VM):

1.

2 Marker

- Two markers are used to define a line in space. A new virtual marker can be calculated anywhere on this line.

2.

3 Marker

- Three markers are used to define a plane in space. A new virtual marker can be calculated anywhere in space relative to the origin of this plane.

The placement of the virtual marker along a line or relative to a plane can be accomplished in real world measured values using the units of calibration as the units of measurement, or as a ratio. In the case of a line, the

11-16

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

ratio is based on the distance between the two markers defining the line.

In the case of the plane, the ratio is based on the distance between the two markers defining the Y axis of the plane.

Figure 11-12. Example Virtual Marker Setup

Types of Virtual

Hierarchical

Translation &

Rotation Data

Degrees of

Freedom

Currently, there are two primary methods of exporting the motion of a subject:

*.trc

(track row column) output files

*.htr2

(hierarchical translational rotation) output files

A

*.trc

file contains the X, Y, and Z translation values for the reflective markers relative to the capture volume’s coordinate system. To translate this data into a hierarchical segment model requires software having an

Inverse Kinematic (IK) Solver to create joint translations and rotations.

This I.K. approach works well when the proportions of the subject are similar to the animation model.

An

*.htr2

file contains hierarchical translation and rotation data representing the different identified segments (bones) of the subject’s body. In this approach, you must establish virtual markers at the estimated joint centers and create segment coordinate systems for each segment. The virtual markers and segment coordinate systems are defined once for a particular marker set and then stored and recalled from the project file.

A segment’s movement characteristics can be expressed as having various degrees-of-freedom. For example, if a single marker placed on the right shoulder is used to define the origin of the right upper arm, and we track this marker through space creating a trajectory, we will express the movement of the right upper arm origin as having 3 degrees-of-freedom (translations in X,Y, and Z).

If we add another marker to the right elbow and track it along with the marker on the right shoulder, we can now express movement of the right upper arm segment as having 5 degrees-of-freedom (translations in X, Y, and Z, and rotations in X and Z). This assumes that the Y axis extends from the right shoulder to the right elbow. If we add a third marker to the right wrist and track all three markers, we now have 6 degrees-of-freedom for the right upper arm segment.

11-17

Chapter 11: Model Edit Panel

Figure 11-13. Marker Number vs. Degrees of Freedom

Translation

X, Y, & Z

EVaRT 5.0 User’s Manual

One Marker = 3 Degrees-of-Freedom

A single marker can have a maximum of 3 Degrees-of-Freedom:

Translations in X, Y and Z over time

Translation

X, Y, & Z

Up & Down

Two Markers = 5 Degrees-of-Freedom

Side-to-side

Translation

X, Y, & Z

Up-Down

3 Markers = 6 Degrees-of-Freedom

Rotation about the bone axis

Side-to-side

11-18

EVaRT 5.0 User’s Manual

Figure 11-14. Examples of Hinge Joints

Knee Modified Hinge Joint

Secondary Rotation about the bone of the lower leg

(Limited to a few degrees.)

Note - No Rotation about the bone axis at the knee

Chapter 11: Model Edit Panel

Primary Rotation

Flexion/Extension

Secondary Rotation about the bone axis of the lower arm.

Ankle Modified Hinge Joint

Primary Rotation

Flexion/Extension

Primary Rotation

Flexion/Extension

Note - Hinge Joint assumed only to flex and extend-no other rotations

Elbow Hinge Joint

Secondary Rotation

Inversion/Eversion

11-19

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Calculating

Virtual Marker

Tracks

Calculate virtual marker trajectories based on the Virtual Marker definitions in the current project file. Virtual Markers are cleared when you track any new data.

Track and edit your trials before calculating the Virtual Marker Tracks.

1.

2.

3.

Open the Virtual Markers Definitions form by clicking the

V-Marker

Definitions

button in Model Edit or by selecting

Tools > Virtual

Marker Definitions

in the Menu Bar.

Fill out the Virtual Marker Definitions form for the desired markers.

Click

Calculate

to calculate the virtual marker positions.

Virtual Marker Quick-Start Example

Figure 11-15. Matching Data and Video Frame

Open the Sample

Project

The above images show a frame of motion capture data along with a matching frame of video showing the posture of the performer. Of particular interest are the positions of the motion capture markers relative to his body. In this example, virtual marker definitions will be created that estimate the actual joint locations of the performer.

It is always a good idea, when adding virtual markers to a marker set, to differentiate the virtual markers from the standard markers by adding a prefix to the name. The choice is up to you, but common ones are "JC_"

(for Joint Center), "VM_" or simply "V_". For this example, the last one will be chosen because, to help the user avoid name conflict problems,

EVaRT

inserts a "V_" at the beginning of the Name field of a new virtual marker definition.

To follow along with the example, please get started by loading the sample data and opening the program to the correct panel:

1.

2.

3.

Start

EVaRT

and load the project

Dave_Fresh.prj

located in the

EVaRT50/Samples/VirtualMarkers

directory.

Load the tracks file

Dave_ROM1.trb

.

Open the

Model Edit > TreeView

sub-panel. Figure 11-16

shows what you should see in the upper right of

EVaRT

11-20

EVaRT 5.0 User’s Manual

Figure 11-16. Initial Tree View for Dave_Fresh.prj

Chapter 11: Model Edit Panel

Ensure that the Show Virtual Markers option is turned on by doing a right mouse click in the 3D display and viewing the Show menu.

Center of the Pelvis

It is common practice to use the pelvis area as the "root" of a skeleton structure representing a person's motion (the process of constructing such a hierarchy of bones will be described in detail in the Skeleton Builder

Quick-Start Guide) it is normal to interleave the virtual marker creation process with the Skeleton Building process, therefore, as much as possible, the virtual marker definition process will match the Skeleton Building process even though these two guides are presented in two separate passes. Thus we start with the pelvis.

11-21

Chapter 11: Model Edit Panel

Figure 11-17. The Five Pelvis Markers

EVaRT 5.0 User’s Manual

The above image shows a close-up of the 5 pelvis markers. Any combination of 3 of the markers could be used to create the virtual marker, but it is always a good idea to choose markers that are the farthest from co-linear as possible. In this case, the top marker (M_Root) and the two, front, lower markers (M_FLHip and M_FRHip) almost form an equilateral triangle, this is a nearly ideal configuration for numerical stability of the virtual marker calculation.

7.

8.

9.

1.

2.

3.

4.

5.

6.

Start the marker definition by pressing the V-Marker Definition button.

The default setting of Three Marker Value (3-MV) is correct, do not change this.

Enter the name "V_Root" in the name field. When you press your keyboard Enter button the Origin Marker field will be highlighted.

With your mouse, do a single left mouse click on the M_Root marker in the 3D display. The highlighted field will switch to Long Axis

Marker (Y).

With your mouse, do a single left mouse click on the M_FRHip marker in the 3D display. The highlighted field will switch to Plane

Marker (XY).

With your mouse, do a single left mouse click on the M_FLHip marker in the 3D display. In the 3D display you will see a darkened triangle indicate the area of the virtual marker definition.

In the field Long Axis enter the value of 60.

In the field Plane enter the value of 35.

In the field Perpendicular, leave the value at 0.

11-22

EVaRT 5.0 User’s Manual

Figure 11-18. Virtual Marker Definition for V_Root

Chapter 11: Model Edit Panel

As you change the values in the numerical fields you will see the display of the virtual marker update. In this case, the option of Three Marker

Value was chosen instead of Three Marker Ratio. The "Value" types of definitions interpret the numerical fields as millimeters. The "Ratio" types of definitions interpret the numerical fields as percentages. For the

V_Root virtual marker, either three marker definition would be suitable.

One advantage of the Three Marker Value definition is that it has the optional Snap to this Marker field. By highlighting this field and then selecting a marker, the values necessary for the virtual marker to be at the location of the selected marker will be entered into the numerical fields for you.

Play around with the values in the numerical fields to get a sense of how the virtual markers location is calculated and how EVaRT visually displays the calculation parameters to help you place the virtual marker.

There is no need to press Calculate Virtual Markers to see updates to definition changes. This button is for when you would like to see an XYZ graph of your virtual markers. There are times when this is helpful for reconstructing data when markers were lost (and it wasn't noticed) before the data was captured. The normal procedure is for EVaRT to automatically (re)calculate the virtual markers any time there are changes to the original data or to the marker definitions.

To create a new virtual maker definition press the button New V-Marker

Definition, this will start you over with a fresh virtual marker definition.

To go back and edit previous virtual marker definitions you may do any one of these:

Close the Virtual Marker Definition Window; select the virtual marker from the treeview and then press V-Marker Definition. -OR-

11-23

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Press the button New V-Marker Definition and then select an existing definition from the treeview. - OR-

Select the virtual marker from the treeview and modify the parameters in the property editor in the lower right portion of the screen.

Each Leg

The process for creating virtual makers for each leg is symmetric so only the generation of the virtual markers for the left leg is shown here.

Hip

One of the goals of the virtual marker process is to create the markers as close to the person's actual joint centers as possible (in truth, there is no such thing as an actual joint center since real bones do not move in an idealized fashion - their joint center changes depending on the position of the bone - the knee joint's "center" may shift as much as 10 millimeters as the leg bends, for example). For the purposes of most applications, such as kinematic analysis and character animation, an approximation is good enough.

In the case of the hip joint, it is useful to note that the locations of the pelvis markers are usually above the actual location of the hip joint. So the position of the virtual marker with respect to the pelvis markers will be lowered as well as moved inwards.

Figure 11-19. Virtual Marker Definition for V_LHIp

11-24

Like with the definition of the V_Root marker, the V_LHip definition uses the Three Marker Value type of definition. Note the use of the Perpendicular field to position the virtual marker lower than the 3 definition markers. This is possible because the Three Marker types of definitions create a fully 3D reference frame for positioning the virtual marker.

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Knee

As with the definition of the V_LHip maker the goal for the definition of the V_LKnee marker is to estimate the location of the knee's joint center.

In this case it is a matter of going straight in from the location of the knee marker on the outside of the knee.

Figure 11-20. Virtual Marker Definition for V_LKnee

Note:

The yellow, M_BLHip marker is used instead of the red M_FLHip marker. The back hip marker is preferable because it is less co-linear than the front marker. This will help the V_LKnee calculations be more stable.

11-25

Chapter 11: Model Edit Panel

Ankle

Figure 11-21. Virtual Marker Definition for V_LAnkle

EVaRT 5.0 User’s Manual

Notice that the definition of the V_LAnkle marker uses the previously defined virtual marker for the knee. It is perfectly acceptable to cascade virtual marker definitions in this fashion. For the V_LAnkle marker this is necessary since there are otherwise too few markers available to create an offset from the outside, original ankle marker. Also note the negative value in the Long Axis field, this lowers the virtual marker to a location that more closely represents the center of the ankle joint.

11-26

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Heel

The heel virtual marker definition is optional, you may skip this definition and add just the V_LFoot marker. Like the V_LAnkle marker it relies on a previously defined virtual marker (the V_LAnkle marker, in fact) to generate its position. Once again, the Three Marker Value type is used:

Figure 11-22. Virtual Marker Definition for V_LHeel

11-27

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Foot

The principal goal of the definition of all the markers for the foot area (ankle, heel and foot) was to align the markers in a plane that matches the

V_LKnee and M_LToe markers. This will make for a more useful (as well as aesthetically pleasing) Skeleton Builder definition later.

Figure 11-23. Virtual Marker Definition for V_LFoot

11-28

Note that the V_LFoot marker uses nearly the same set of markers for its calculation as the V_LHeel. The order of the markers is different (different Origin Marker specification).

The virtual marker definitions for the right leg are done just like as shown for the left. The sample project

Dave_LegsDefined.prj

contains all of these definitions and can be used as a starting point for the next set of definitions for the torso.

Torso

The next major step in defining virtual markers is to work up the entire length of the torso along the spine. It is important (particularly in the

Skeleton Builder process) for the virtual markers of the spine NOT to be in a perfectly straight line. It's best to try and follow the natural curvature of the spine. For the skeleton later, this allows some "give" in the motion of the skeleton to allows it to better track the motion of the performer without undue stretching of the bones.

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Spinal Segments

Each of the spine virtual makers are defined in a similar fashion. The naming convention will be a little different, each of the spine virtual markers will be named in an alphabetical sequence (V_SpineA,

V_SpineB, etc...). This is similar to common practices used in skeleton definitions in animation systems.

Figure 11-24. Virtual Marker Definition for V_SpineA

The definition for V_SpineA is intended to represent the arch in the small of the back (relative to the pending definition of V_SpineB).

Figure 11-25. Virtual Marker Definition for V_SpineB

11-29

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Continuing on to V_SpineC and the "S" shape for the spine:

Figure 11-26. Virtual Marker Definition for V_SpineC

Note that a small left-right adjustment was made to V_SpineC to keep the spine aligned. Continuing on to V_SpineD and the "S" shape for the spine:

Figure 11-27. Virtual Marker Definition for V_SpineD

11-30

To keep a little distance from the upcoming V_Neck definition, the position of V_SpineD was moved down slightly.

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Neck

The neck is the end point of the spine definitions. As such it follows, and finishes, the "S" shape of the spine. Notice that it slopes almost as far forward as it does up. This is intentional and could be accentuated even more.

Figure 11-28. Virtual Marker Definition for V_Neck

11-31

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Head

The V_Head marker definition is designed so that while the performer's head is upright the position of the V_Head marker relative to the V_Neck marker will also be upright. This is different from earlier attempts to put a virtual marker at a joint location, here the V_Head marker rests on what would be roughly where the back of the performer's head is. The idea is that as an end effector, the orientation of the head segment (as will be defined by Skeleton Builder) is more important than the actual location of the joint so its relationship to the neck joint is more important.

Figure 11-29. Virtual Marker Definition for V_Head

11-32

Note that this is the first time the Three Marker Ratio type is used. When trying to get a central location for a virtual marker this type of definition can be very handy.

The sample project file

Dave_LegsTorsoDefined.prj

has all of the virtual marker definitions described so far. You can start with that project to move on to the next step.

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Each Arm

The definitions for each arm are symmetric so only the virtual markers for the left arm are described here.

Shoulder

Getting a good virtual marker definition for the shoulder joint is vital to getting natural looking motion for the arms. The difficulty here is that the

M_LShoulder and M_RShoulder markers, being located on the top of the shoulders of the performer, rotate far on the outside of the shoulder joint.

Therefore they have a large amount of translation data. You can see this by playing through frames 900 to 1050 of the tracks file

Dave_ROM1.trb

.

Note the distance of the shoulder markers from each other- when the arms are raised the markers are much closer to each other than when the arms are down. It's important to eliminate this disparity to get a good representation of clavicle motion in the Skeleton Builder definition.

Fortunately this is easy to obtain by a proper definition of the virtual markers. In this case the Origin marker is specified as the M_Neck marker and the type of definition is the Three Marker Value. This ensures that the location of the virtual marker are a set distance from the M_Neck marker. Small adjustments to the Plane and Perpendicular numerical inputs place the marker approximately at the actual shoulder joint location.

Figure 11-30. Virtual Marker Definition for V_LShoulder

Look again at frames 900-1050. You will see that the V_LShoulder joint stays a constant distance from the V_Neck marker.

11-33

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

Elbow

The definition of the V_LElbow marker is constructed to place it at the center of the actual elbow joint.

Figure 11-31. Virtual Marker Definition for V_LElbow

Wrist

The definition of the V_LWrist marker is constructed to place it at the center of the actual wrist joint.

Figure 11-32. Virtual Marker Definition for V_LWrist

11-34

EVaRT 5.0 User’s Manual Chapter 11: Model Edit Panel

Finger

The definition of the V_LFinger marker, despite the name, is really constructed to place it at the center of the knuckles on the left hand.

Figure 11-33. Virtual Marker Definition for V_LFinger

Lost Marker

Example

Note the use of the Three Marker Ratio type of definition. This makes it easy to place the marker right in the middle of the thumb and pinky markers.

After completing the right arm the construction of the virtual marker defin i t i o n s f o r t h i s p r o j e c t i s c o m p l e t e . T h e s a m p l e p r o j e c t f i l e

Dave_FullyDefined.prj

contains the definitions of all the virtual markers.

There are times, despite the best diligence of the motion capture personnel, when a marker falls off of a performer and the fact isn't noted until some time afterward. When this happens, a series of trials will be made with an incomplete data set. Trials that are, except for the missing marker, good. It is for this circumstance that a particular feature of virtual markers is very useful.

This example takes advantage of the fact that at some time during the motion capture session the missing marker was still attached to the performer. A trial containing the marker will be used to construct a virtual marker that will replace it in the trials where the marker is missing. It isn't possible to use Join Virtual to fix this problem because Join Virtual requires that the gapped marker exist in the same trial either before or after the gap that is being filled in.

In the sample data directory there are two TRB files:

Trial_12.trb

and

Trial_13.trb

. Trial_12 contains all the markers. Trial_13 has a missing front right hip marker. The following sequence of steps will use a virtual marker definition to fill in the missing marker in Trial_13 by constructing the virtual marker in Trial_12 using the Snap to this Marker feature.

11-35

Chapter 11: Model Edit Panel EVaRT 5.0 User’s Manual

1.

2.

3.

Start EVaRT and load the project

Dave_Fresh.prj

.

Load the tracks file

Trial_12.trb

.

Construct the virtual maker V_FRHip.

Figure 11-34. Virtual Marker Definition for V_FRHip

The virtual marker definition uses the Snap to this Marker to automatically calculate the numerical values necessary to place the virtual marker at the same location as the actual marker.

3.

4.

5.

1.

2.

Load the tracks file

Trial_13.trb

.

Select the

Calculate Virtual Markers

button in the

Post Process

sub-panel.

From the marker list, select the marker

M_FRhip

.

From the

Post Process

panel, press

Marker ID...

With the entire frame range selected, select the virtual marker

V_FRHip

in the 3D display.

The missing M_FRHip marker has now been filled in with data from the virtual marker and this can be saved to a new TRB file.

11-36

Chapter 12

User Apps Panel

Topic

Overview

X Sub-Panel

Sky Sub-Panel

Motion Composer Sub-Panel

BioFeedTrak Sub-Panel

QuickDB Sub-Panel

Page

12-1

12-2

12-5

12-6

12-7

12-9

Overview

The User Apps panel houses the plugins for the

EVaRT

user interface.

These plugins include:

X

Sky Writer

Motion Composer

BioFeedTrak

QuickDB

Figure 12-1. The User Apps Mode

12-1

Chapter 12: User Apps Panel EVaRT 5.0 User’s Manual

X Sub-Panel

Figure 12-2. X Sub-Panel

The X sub-panel provides a set of extra functions that are used in the post process mode.

Delete Outside

Volume

Snippets/Delete

Short Snippets

12-2

Eliminates all marker data outside of the volume defined in Calibration

Details.

Deletes data strings (in frames) that are shorter than a specified length.

EVaRT 5.0 User’s Manual Chapter 12: User Apps Panel

Refine Tracks

Figure 12-3. Refine Tracks

This feature will smooth data that has become jumpy due to camera on/off noise. When a camera is turned on and off, there is at times a small data spike in the frames before the camera is turned off and after it is on. This is useful on facial data where small increments (< 1 mm) will have a significant effect on the final results (animated character).

Missing data to be smoothed

To Refine Tracks you will need to:

1.

2.

3.

4.

5.

Load a

*.prj

file.

Load a

*.trb

file.

Select

File > Load VC for Post

.

Click-on

Refine Tracks

.

Select

File > Save Tracks

.

Global Marker Data

Adjustments

The Global Marker Data Adjustment section allows the user to modify the tracks data by translating, rotating, and/or scaling the data. This is an operation that applies to all marker data over all the tracks. It's especially handy for converting the overall orientation of the data (such as from a Zup coordinate system to a Y-up coordinate system).

Model Adjustments

The Model Adjustment section allows the user to update the Calcium

Solver model pose data and the template model pose data simultaneously

(this is the data displayed when the Show Model Pose flag is on). The marker data in the model pose is used for two different operations: as a starting pose for the template when doing a New Subject operation; and as the matching base pose for the skeleton in the Calcium Solver model.

Doing both adjustments at once is important to maintaining the integrity

12-3

Chapter 12: User Apps Panel EVaRT 5.0 User’s Manual

of the data. None the less, the user is allowed to change them independently if necessary.

Refit Identifying Template

—This option takes the current frame in Post

Process and compares the template linkage lengths of that frame with the stored model pose. The template Min/Max values are re-calculated based on the amount of change in the linkage lengths.

Update Model Pose Markers

—This checkbox indicates that the stored marker model pose is to be replaced with the marker data on the current frame.

Update Model Pose Skeleton

—This checkbox indicates that the stored skeleton model pose is to be replaced with the current skeleton data that has been calculated for the current frame.

These last two options are used to update the model pose of a performer between motion capture sessions (such as from one day to the next). This avoids having to spend time refitting the position of the skeleton to the new day's model pose data. The user should still verify that the fit is a good one, but if the markers on the performer have not moved by very much then the fit is likely to be good.

12-4

EVaRT 5.0 User’s Manual Chapter 12: User Apps Panel

Sky Sub-Panel

Sky Writer is the name of the scripting interface for

EVaRT

. It uses the

VB Script engine to provide the semantic structure of the language along with Visual Basic to provide the graphical user interface of the window pane. Sky (Writer) is intended as a tool for users to encapsulate elements of repetitive tasks such as file processing, data editing and parameter setting. This tool is intended for users who have some general knowledge of scripting and programming.

For more information refer to Chapter 14, Sky Writer and visit www.EVaRT-Forum.com <http://www.evart-forum.com/>.

Figure 12-4. Sky Scripting Interface

12-5

Chapter 12: User Apps Panel EVaRT 5.0 User’s Manual

Motion Composer Sub-Panel

Motion Composer is a suite of tools for collating, integrating, and presenting interactive motion capture data. Motion Composer is a collection of authoring tools, data structures, and visualization panes. These are integrated into

EVaRT

to help achieve a seamless workflow for the user to package and present a motion capture session. Some of the key features to be found in Motion Composer include:

Integrated Authoring

Interactive Player

Rich Media Support

Presentation Tools

Figure 12-5. Motion Composer Interface

12-6

EVaRT 5.0 User’s Manual Chapter 12: User Apps Panel

BioFeedTrak Sub-Panel

BioFeedTrak

is a general condition, evaluation, and response program, integrated into

EVaRT

, for designing and implementing biofeedback programs that can enable clinicians and patients to receive instantaneous audio feedback to kinematic movements.

BioFeedTrak

is able to give real-time feedback in the form of sounds based on kinetic variables that fall within certain bounds during the pre-defined performance of any type of physical task.

Kinematic variables include position, velocity and acceleration of individual markers (up to 45 markers) placed on key anatomical points of interest. Included angle between three markers as well as the angle of inclination of a segment defined by two markers can be used to provide feedback. Kinetic data include the following:

• horizontal and vertical forces

• moment about the vertical axis

• the coordinates of the center of pressure with respect to the forceplates

The program works in conjunction with the Motion Analysis Eagle Digital System or the Falcon Analog System. For a typical application procedure, the user will do the following:

1.

2.

3.

4.

Choose and set the variables to be monitored

Determine the starting and ending parameters for each variable to be assessed

Choose the volume and frequency of the audio feedback

Start the Real-Time system

With this, the patient and clinician are able to work side by side to retrain areas of the body that need further optimization.

More information may be found in the BioFeedTrak Quick-Start Guide.

12-7

Chapter 12: User Apps Panel

Figure 12-6. BioFeedTrak Interface

EVaRT 5.0 User’s Manual

12-8

EVaRT 5.0 User’s Manual Chapter 12: User Apps Panel

QuickDB Sub-Panel

QuickDB is an integrated database tool for

EVaRT

that allows the user to easily track all

EVaRT

project information. Microsoft Access databases are used to tabulate information about your projects (a copy of Microsoft

Access is not required to take advantage of this tool). A master database keeps a list of all the individual projects you make. Each individual project (called a "session list") keeps track of all the data associated with a specific project directory (project files, VC files, tracks files, etc...). It is fast and easy to create a session list of any data you already have.

QuickDB will scan your project directory for the data you have already collected and will create the session database for you.

QuickDB is all of the following:

Is very handy for keeping track of all your projects, you can scan for and then load project data with ease.

Will record your trials as you collect data.

Contains tracking information about a trial's Post Process status.

Allow's multiple user access to shared session databases.

Makes it easy to share databases with other users.

Allows the creation of capture lists ahead of time so that capture names can be loaded from QuickDB while recording a session.

Is an SQL database that can be used to generate reports on project status.

Figure 12-7. User Apps > QuickDB Sub-Panel

12-9

Chapter 12: User Apps Panel

Figure 12-8. QuickDB Interface

EVaRT 5.0 User’s Manual

12-10

Chapter 13

Skeleton Types

Topic

Overview

Skeleton Builder (SkB) Skeletons

Solver Type Skeletons

Which Skeleton Engine Should I Use?

EVaRT Skeleton Engine Selection

Skeleton Option Details

Exporting the Skeleton Data Into an HTR File

Multiple Characters and Multiple Skeletons

Overview

EVaRT

supports two kinds of skeleton calculations,

Skeleton Builder

(SkB) skeletons and

Calcium Solver

type skeletons. Either kind is calculated in the

EVaRT

software and either can be calculated from live camera data, simulated Real Time with VC files, or from XYZ data in Post Processing. Both the marker data and the skeleton data are available to the

Talon

streaming plugins, such as the Maya and Kaydara Talon streaming plugins. The user can write their own plugin with the

Talon SDK

(Software Development Kit) available from Motion Analysis.

Figure 13-1. Calcium Solver Integrated with EVaRT

Page

13-1

13-2

13-2

13-3

13-3

13-5

13-7

13-8

13-1

Chapter 13: Skeleton Types EVaRT 5.0 User’s Manual

Skeleton Builder (SkB) Skeletons

Skeleton Builder (SkB) skeletons are relatively simple, direct and fast calculations of segments (bones) that are defined and calculated from one marker center to another. The markers can be real or virtual (calculated) and are typically from one virtual joint center to a second virtual joint center. A 3D local coordinate system is defined with 3 markers:

1.

2.

3.

One marker defines the origin.

A second marker defines the bone (Y) axis.

The third marker defines the XY plane.

The advantage of the SkB type skeleton is that they compute very quickly and they represent the method of how most biomechanical models have been computed for many years. The disadvantage is seen when viewed on a skinned character in an animation package: The bones (segments) change length as a result of the calculation method. This is due primarily to the motion of the markers on the skin which changes from frame to frame. An animated character can be set up so that the character keeps a fixed length skeleton and the skeleton is driven only by the angles measured from the skeleton. This has the visually undesirable artifact that the character’s feet will appear to slide on the floor and possibly raise above or protrude below the floor.

Solver Type Skeletons

Solver type skeletons are calculated quite differently than SkB skeletons.

Solver uses the Global Optimization method of calculating the best fit of the marker data to the underlying fixed bone length skeleton. This technology was pioneered by Motion Analysis in 1990. The early version model setup was somewhat cumbersome and required physically measuring from a person’s joint centers to the marker locations before the skeleton could be used. Now, the skeletons are defined and edited within the

Calcium software interface within

EVaRT

. This provides a graphical way of either reading in an existing skeleton (or creating one) from an animation package such as Maya or 3DMAX. The typical way would be to create the character in the animation package and export the skeleton using an htr file using the MAC File IO plugins. The Global Optimization method is an iterative method of seeking the best fitting of the skeleton within the “marker cloud” of identified markers. The results are quite astounding: the animated characters motions derived from this method is very good. Final editing of the htr skeleton motion data can be done within the animation package or with a third party tool, such as Kaydara’s

Motionbuilder.

13-2

EVaRT 5.0 User’s Manual Chapter 13: Skeleton Types

Which Skeleton Engine Should I Use?

SkB skeletons are good for most biomechanical applications and have been the norm there for many years. They are also used for animation customers who want fast “Pre-vis” tools for quick pre-visualizations of your characters motions.

For the final cut and the big screen presentations, where details matter and looks are everything, you will be glad to have your Calcium Solver skeletons under the skin of your final characters.

EVaRT Skeleton Engine Selection

The Engine Selection sets the method of calculating the (optional) skeleton model that can be calculated in

EVaRT

and later versions. There are two main skeleton engines available from Motion Analysis: Skeleton

Builder, which was based on earlier SkB definitions, and Calcium Solver, based on Solver technologies. The

EVaRT

and later versions will run any skeleton engine that is previously defined without any additional licenses.

To define or edit the skeleton definitions, you need additional licenses.

In the

EVaRT

and later releases, the skeleton model is saved with your marker set information in the project file. The

File > Load Marker Set

menu item will load the skeleton type from a PRJ file once it has been stored there.

Real Time and

Simulated Real

Time Skeletons

The Motion Capture panel has a Skeleton check box to ask if you want the

Skeleton calculated with either mode: Connected (to cameras) or simulated real time (Disconnected-Use Raw Files). To see the skeleton, use the right-mouse click menu and select

Show Skeleton

.

Post Process

Skeletons

SkB Skeleton

Builder skeletons

Skeletons can also be calculated in the Post Processing panel from your current XYZ data that is visible in the 3D and XYZ data views.

Skeleton Builder (SkB) skeletons are always calculated if the

SkB Skeleton Builder

option is selected. You can see the Skeleton by selecting

Show Skeleton

from the right mouse menu in the 3D data view. If you are using one of the streaming plugins, like the Maya Talon plugin or the

Motionbuilder plugin, both the marker data and the calculated skeleton is available to drive your animated character from either the Real Time or

Post Processing mode.

13-3

Chapter 13: Skeleton Types EVaRT 5.0 User’s Manual

Calcium Solver

Skeletons

If you have a Calcium Solver Skeleton defined in Post Processing, the

Calculate Skeleton button (see Figure 13-2 ) appears on the lower left side

of your Post Processing dashboard. You can first select a starting frame from the current time slider, then press the

Calculate Skeleton Bone

button which activates the Solver engine to do the Global Optimization method and solve for the skeleton beneath the marker cloud.

Figure 13-2. Calculate Skeleton Button

13-4

EVaRT 5.0 User’s Manual Chapter 13: Skeleton Types

This can take several seconds to minutes, depending on the length and complexity of the skeleton(s). If you are using one of the streaming Talon plugins, the skeleton data is available after the Bone Button calculations are finished. You can scrub back and forth in the Post Processing mode or press the

Play

button and both the marker data and the skeleton (htr type) data are available to the streaming Talon plugins (like Maya).

Figure 13-3. Skeleton Engine Setup

Skeleton Option Details

No Skeleton

Calculation

Calcium Solver

1.1.2

Even if a skeleton model is defined, the

EVaRT

software will not use it if this button is selected. This is the default value and the value for all project (prj) files saved previous to the

EVaRT

release.

Skeleton Builder

(SkB) Engine

This calculates joint centers directly from real or virtual marker locations.

Bone segments are defined from one marker to another, typically virtual markers that represent joint centers. The advantage to this method is the fast and direct calculations from markers to joint centers to bone segments. The disadvantage is primarily for high resolution animation use:, since the joint centers are calculated directly from real and virtual marker locations. Bone lengths will vary slightly from frame to frame due to marker-skin motion which can cause the animated character’s skin to distort and not look as good as it expected.

This is a very different method of calculating the skeleton motion from marker locations. Typically, the skeleton is defined within one of the several animation packages and exported and saved in an HTR and a MOD

(model) file. This skeleton is not allowed to change size to fit the motion data, but the Solver engine software uses a best fit Global Optimization of the marker data to conform to the rigid underlying skeleton. This results in the very best way of animating characters from mocap data, but to use it in

EVaRT

, you need to save a .mod file with the same name as your prj file. The

Calcium

software allows you to export a .mod file. The Solver

Global Optimization method is resident in three of Motion Analysis Cor-

13-5

Chapter 13: Skeleton Types EVaRT 5.0 User’s Manual

poration's software products:

EVaRT, Calcium

, and the

SIMM

modeling package.

SIMM OrthoTrak

Model

This uses the Solver engine with the same advantages as the

Calcium

Solver method above, but with a known and fixed marker set that was developed for biomechanics use. To use it, you must use some variation of the

OrthoTrak

marker set, which has several required markers (such as the Knee, Ankle, Hip and Shoulder markers) and many optional markers that will introduce more detail and more bone segments into the solution.

The big advantage over the

Calcium

Solver model is that you do not need to create a MOD file, which means that you do not need to use or learn the

Calcium

software. The

EVaRT

software created an even more thorough

JNT (joint) file when you press the

Create OrthoTrak Model

button on the

Setup > Misc

sub-panel. You do need to have the person standing in a neutral pose, typically with the arms out in a T-pose, feet slightly apart and thumbs forward. That is the current TRC file that needs to be loaded when you press the

Create OrthoTrak Model

button.

Model Edit Tree

View

The Skeleton Engine type is also displayed and can be edited on the

Model Edit > Tree View

sub-panel when you select the prj name at the root of the Tree View. The Skeleton Engine type appears as the Value of the Skeleton Engine Property at the bottom of the Tree View sub-panel.

Figure 13-4. Model Edit Tree View

13-6

EVaRT 5.0 User’s Manual Chapter 13: Skeleton Types

Exporting the Skeleton Data Into an HTR File

The

SkB

or Solver type skeleton data can be saved to an HTR (hierarchical translations and rotations) file after you calculate it and view it in

EVaRT

. Select the

File > Export HTR file...

menu item. This is for use with animation packages. You will then select the default top Euler Angle

Order (ZYX) since that is how the plugins are built to receive the data.

The following are options on the Export HTR file menu:

Euler Angle Order

Use ZYX (which is the default) if you are going to import this with a Motion Analysis File IO plugin to the animation packages.Other Euler Angle orders will be decided by your local mathematicians. The numbers in the

EVaRT

software are stored internally in a certain way and exported to the

HTR file according to the method above.

Figure 13-5. HTR Export Options

Base Position

Options

Current Frame

Angles in the columns of HTR data are absolute angles according to the coordinate systems defined. This is typically used by biomechanics and research customers.

The absolute angles in the currently selected frame are written out in the file header of the HTR file. The angles in the columns are zero referenced to the angles in the file header. Use this method option if you are going to read the htr files into an animation package with the Motion Analysis File

IO plugins.

13-7

Chapter 13: Skeleton Types EVaRT 5.0 User’s Manual

Licensing Notes

SkB

skeletons are defined and edited within the

EVaRT

software and require a separate license item, but they can be run and the skeletons data generated without additional licenses in

EVaRT

and later versions. There is separate documentation provided with the Skeleton Builder that shows how to set up and edit

SkB

skeletons.

Calcium

Solver skeletons are imported or created in the

Calcium

software.

Calcium

software requires a separate license to edit or create the skeletons, but they too can be run and the skeleton data created from

EVaRT

or later versions without a separate license.

Calcium

also can create the HTR skeleton data using the same Solver engine as

EVaRT

and the

SIMM Motion Module.

Multiple Characters and Multiple Skeletons

When you specify additional marker sets, the skeleton engine needs to be in each of the project files that you select. The skeleton type is stored in the project file. For previously stored project files, open up each of the project files separately, go to the

Setup > Misc

sub-panel, and specify the appropriate skeleton engine, then save out the project file. Load one of the project files, then in the

Motion Capture > Objects

sub-panel, specify the second project file as an “Additional Tracking Object". You should then be able to load a .vc file, and both skeletons should become solved.

13-8

Chapter 14

Sky Scripting Interface

Overview

Graphical User Interface

The Functions

The Script

Script Examples

Installation Notes

Frequently Asked Questions

Topic Page

14-1

14-2

14-3

14-4

14-5

14-7

14-7

Overview

Sky is the name of the scripting interface for

EVaRT

. It uses the VB Script engine to provide the semantic structure of the language along with Visual

Basic to provide the graphical user interface of the window pane. Sky is intended as a tool for users to encapsulate elements of repetitive tasks such as file processing, data editing and parameter setting. This tool is intended for users who have some general knowledge of scripting and programming.

Most of the Sky functions are direct, simple wrappers for the corresponding

EVaRT

calls. Some exceptions have to do with sending messages to

EVaRT

and in re-arranging arrays of data that get passed back and forth.

Sky does have an OCX file, which is named ScriptPlugin.ocx. When this file is loaded into the

EVaRT50\Plugins

directory, the Sky interface tab in the User Apps panels becomes active. The panel has one button that brings up the floating interface.

The Plugin Kit mostly consists of calling the

EVaRT.dll

functions, but it has some additional functions that must be filled as information to

EVaRT

. (e.g. name, type of plugin, etc.).

14-1

Chapter 14: Sky Scripting Interface EVaRT 5.0 User’s Manual

Graphical User Interface

The user interface for Sky is found by going to the

User Apps > Sky

subpanel.

Figure 14-1. Sky Sub-Panel

Click on the

Open Sky Interface

button to bring up the floating window for the Sky interface.

Figure 14-2. Sky User Interface

Toolbar

14-2

New

—Clears the Script text editor and start editing a new script.

EVaRT 5.0 User’s Manual Chapter 14: Sky Scripting Interface

Open

—Brings up a file browser to find a sky file not in the current directory.

Save

—Saves the Script text to a new sky file.

Run

—Executes the text in the Script editor. This automatically saves the current script text to the current file.

Cancel

—Cancels the currently executing script. This works only if the script was written to use the swCancelled function.

Info

—Prints out basic information about Sky.

Refresh

—Causes Sky to refresh the Local Sky Files list (or the Global Sky Files list, which ever one is currently shown) by re-reading the appropriate directory.

Initialize

—Copies Sky files from the folder

MAC_DIR/UserFiles/

SkyFiles/CopyPerProject

to the current working project directory.

This makes it easy to initialize a new project with your favorite scripts. Will not overwrite existing scripts of the same name.

Local Sky Files

This contains a list of all the Sky files in the same directory as the current project. Load a file into the script interface by single-clicking on the filename. When you click a new file name, any changes you made to the currently loaded file are saved automatically. The currently loaded file name will continue to be highlighted for as long as that file is current.

Note:

If you select the Local Sky Files tab when it is already selected then this will save out the current file. The current file is saved when you switch to the other tab.

Global Sky Files

This contains a list of all the Sky files in a common, global directory. This directory is a sub-folder of the folder pointed to by the environment variable MAC_DIR Load a file into the script interface by single-clicking on the filename. When you click a new file name, any changes you made to the currently loaded file are saved automatically. The currently loaded file name will continue to be highlighted for as long as that file is current.

Note:

If you select the Global Sky Files tab when it is already selected then this will save out the current file.The current file is saved when you switch to the other tab.

Script

The currently loaded script. This is what will be run when the Run button is selected.

Output

Displays output from the scripting engine. Error messages from the scripting engine will be displayed as well as any text messages from the script. The output text area is refreshed when a new file is loaded.

The Functions

The scripting functions (for the most part) in Sky are simple wrappers for functions found in the

EVaRT

dll. Any call you make is immediately passed on to the corresponding core function in

EVaRT

. These are the same functions that the

EVaRT

GUI uses at runtime so (again, for the

14-3

Chapter 14: Sky Scripting Interface EVaRT 5.0 User’s Manual

most part) you can do just about anything through the scripting interface that

EVaRT

does interactively.

All the script functions have the same name as the corresponding dll function call except a "sw" has been added to the beginning ("sw" stands for

"Sky Writing").

A Functions list can be found in the file

Functions.html

located in the

C:\ P r ogr a m Fi le s\ M oti on An al y si s\ E Va RT 50 \ Us er Fi le s \Sk y -

Files\SkyDocumentation_Mar23_05

directory.

The Script

The scripting language is VB Script. The only requirement for the script is that it has a Main() function in it, like so:

Sub Main

' commands go here

End Sub

Sky calls the script and executes the main function. You can create and use other subroutines and functions from Main.

14-4

EVaRT 5.0 User’s Manual Chapter 14: Sky Scripting Interface

Script Examples

Data files for all the script examples, along with copies of the scripts and the data they use, are found in the DataExamples subfolder. Note that some scripts might require a change in file paths to point to where you have placed this data on your computer system.

Resample.sky

—Invokes an external command line program to resample a TRC file (changes the frame rate).

MesssageTest.sky

—Logs messages to a file. How to delete a file.

How to create a folder.

MsgBox.sky

—Example of using the built-in VB Script functions

MsgBox and InputBox.

CreateTemplate.sky

—Create a template from a ROM (Range of

Motion) track and then extend the template with a series of other TRB files.

Record.sky

—Load a VC file and record to a TRB file.

DeSpikeAll.sky

—Find and remove spikes from all markers in the tracks file.

DeSpikeSelected.sky

—Find and remove spikes from selected markers in the tracks file over the selected region.

SingleFrameDelete.sky

—Find and remove single frame trajectories.

VMJoinAll.sky

—Perform a VM Join (virtual marker join) on all markers.

FilterAll.sky

—Filter all markers over all frames.

FullProcess.sky

—A complete data example showing a master script calling a sequence of other scripts to perform a complete track, clean, and gap fill of a long motion with many occluded markers.

Trb2HTR.sky

—load a sequence of tracks file and create a set of corresponding HTR files.

Trb2Trc.sky

—Export a sequence of Ascii tracks files from binary tracks file.

FindAutoIDFrame.sky

—Find a valid for an AutoID and do the

AutoID (Template ID)

AutoIDAllFrames.sky

—Perform an Auto ID on all frames.

RecordFive.sky

—Record five different files from the same VC capture file that had five performers.

ProcessBill.sky

—Process one of the five people in the Process-

Five.sky example.

ProcessFive.sky

—Process a data file containing five performers.

FindMaxSpeed.sky

—Find a value for the Max Speed tracking parameter by find the largest distance a marker moves in one frame.

A complete listing of these Sky example files is located in the

Examples.html

file in the Sky folder of the

EVaRT

release CD.

14-5

Chapter 14: Sky Scripting Interface EVaRT 5.0 User’s Manual

Example of

Loading a Sky

File

This example shows a simple demonstration of a Sky file and what it can typically do for a data set. This example will list all the marker names used in the current project file.

1.

2.

3.

In the Sky sub-panel, click on the

Open Sky Interface

button.

In the Global Sky Files tab, select the

MarkerNames.sky

file from the listing.

Click on the

Run Script

button (see

Figure 14-3 ). You will notice that

the script runs, but there are no markers listed (as you have no project file loaded).

Figure 14-3. Sky Interface - Run Script Button

Run Script Button

14-6

4.

5.

6.

Load a project file and then a tracks file.

Click on the

Run Script

button again.

At the bottom of the Sky interface, you will see the model name and the markers of the project file listed.

EVaRT 5.0 User’s Manual Chapter 14: Sky Scripting Interface

Installation Notes

Sky Writer is automatically installed into

EVaRT

. If you are using an earlier version of

EVaRT

and it is not installed, you can utilize the information in this section.

Installing

ScriptPlugin.ocx

The file "ScriptPlugin.ocx" is the Sky plugin file for

EVaRT

. It needs to be copied to the "Plugins" folder where

EVaRT

is installed.

Installing msscript.ocx

The file

msscript.ocx

is a Windows file that must be registered with

Windows to activate the use of the VB Script engine. To install it (which must be done only once) copy the

msscript.ocx

file to a permanent folder location, such as the Motion Analysis folder (usually

C:\Program

Files\Motion Analysis\

) and then execute the following Windows sequence:

1.

2.

Under the Start menu in Windows, select

Run

.

Execute the following (assuming the

msscript.ocx

was copied to the location listed above):

regsvr32 "C:\Program Files\Motion Analysis\msscript.ocx"

This can also be done from the command prompt.

Installing

Example Sky

Files and Data

In order for the example scripts associated with the example project data to work, certain Sky scripts need to be copied to the Global Sky Files location. The contents of the SkyFiles folder in the Example Data directory should be copied to the SkyFiles folder that resides in the folder pointed to by the MAC_DIR environment variable. If the folder doesn't exist, you need to create one (when Sky starts up it will create the folder if it doesn't exist).

Execution of

Command Line

Programs

There are Sky commands (swTrcResample, swExecuteProcess) that allow for the invocation of command line applications. The swTrcResample command requires that there be a bin directory right below the Motion

Analysis root directory (pointed to by the MAC_DIR environment variable). The application file

TrcResample.exe

must be placed in the bin folder for the swTrcResample command to work. TrcResample is part of the mocap toolkit available from the FTP site. Contact Motion Analysis

Customer Support for information on getting the toolkit ([email protected]).

Frequently Asked Questions

1.

When I press "Run" it doesn't run the current script, it appears to run another script or it gives me an error at line 0.

This is likely due to not having a "Sub Main" before any actual commands (and a matching "Sub End" at the end of the script). The scripter requires a Main routine to call.

2.

I get a "Script Syntax Error" when I try to call a function.

14-7

Chapter 14: Sky Scripting Interface EVaRT 5.0 User’s Manual

If you're certain that you've spelled the function's name correctly then it's likely the specific function hasn't been implemented yet. Check with Motion Analysis' Customer support.

Also be sure that all the function arguments are spelled correctly and are the correct type. This is especially important for String arguments.

You might have to force the correct type by putting CStr() around the argument.

3.

Is there a "Listener" that can track my interactive work with

EVaRT

and list out the corresponding Sky commands to perform the same task?

No, unfortunately not at this time. Since Sky was developed as a plugin to

EVaRT

it isn't possible to incorporate this type of feature.

4.

When do my Sky file changes get saved?

At every opportunity to do so the Sky interface will automatically save out the contents of the Input Text window to the currently selected file. Some examples: when the file list is changed from Local to Global; when you run the script; when you load another script and when you close the Sky window. The only times when a script isn't saved is when

EVaRT

is closed down without first closing the Sky window or if

EVaRT

crashes.

14-8

Chapter 15

QuickDB

Overview

Installation Instructions

Quick-Start Steps

Usage Notes

QuickDB Terminology

User Interface

Frequently Asked Questions

Topic

Overview

QuickDB is an integrated database tool for

EVaRT

that allows the user to easily track all

EVaRT

project information. Microsoft Access databases are used to tabulate information about your projects (a copy of Microsoft

Access is not required to take advantage of this tool). A master database keeps a list of all the individual projects you make. Each individual project (called a "session list") keeps track of all the data associated with a specific project directory (project files, VC files, tracks files, etc...). It is fast and easy to create a session list of any data you already have.

QuickDB will scan your project directory for the data you have already collected and will create the session database for you.

QuickDB is all of the following:

Is very handy for keeping track of all your projects, you can scan for and then load project data with ease.

Will record your trials as you collect data.

Contains tracking information about a trial's Post Process status.

Allow's multiple user access to shared session databases.

Makes it easy to share databases with other users.

Allows the creation of capture lists ahead of time so that capture names can be loaded from QuickDB while recording a session.

Is an SQL database that can be used to generate reports on project status.

Page

15-1

15-2

15-2

15-2

15-3

15-3

15-10

15-1

Chapter 15: QuickDB EVaRT 5.0 User’s Manual

Installation Instructions

The main projects database file is called

MasterProjectsList.mdb

. This file must be stored in the

Userfiles\Databases

folder under where the main

EVaRT

directory is found. For a standard installation, this will be

C:\Progr am File s\ Motion Anal ysi s\EVaR T47\ Use rfil es\ Databases\MasterProjectsList.mdb.

Be sure to copy the same databases folder to this location. Another required database file is

SessionTemplate.mdb

. This is also found in the

Userfiles\Databases

folder. This is a blank-session database that is used to make new databases and add them to the

MasterProjectsList.mdb

file.

Copy the QuickDB.ocx file to the plugins folder under the specific

EVaRT installation. For example, this might be

C:\Program Files\Motion

Analysis\EVaRT47\EVaRT4.7.5\plugins\QuickDB.ocx

.

If you do not have development tools (Visual Studio 6 or later) then you will have to install some system OCX files that are required by QuickDB.

The installer for this is found in the

Package

directory and is called

Setup.exe

. Run this installer to install and register the Windows ocx files for handling database interaction. Again, Microsoft Access is not required to be installed to run QuickDB.

Quick-Start Steps

1.

2.

3.

4.

5.

6.

Open the QuickDB interface found in the

User Apps

panel.

Select the

New

button on the right to create a new database.

Click on the

Scan Folder...

button and then select a folder full of trials that you have already created.

Select a files from the TrialName list

Select

Load into EVaRT

.

Select

Sync with EVaRT

.

You should now have a new session database and the trial that you selected is now completely loaded into

EVaRT

(project file, VC file and any tracks files). Furthermore, all the information about this trial is saved in the new session database that you just created.

Usage Notes

All motion capture data files are assumed to be of the form "Base##.ext".

Where "ext" is the extension type ("vc1", "trb" or "trc"). The ## portion refers to the trial number that is automatically appended to the base name of the trial. The QuickDB interface will not work with files that do not have this form.

15-2

EVaRT 5.0 User’s Manual Chapter 15: QuickDB

QuickDB Terminology

The following are general phrase definitions associated with QuickDB.

Master Project List

Session List

This is the master list of projects from the

MasterProjectList.mdb

file.

This is the database file which contains the list of trials. Each item in the

Master Project List refers to one of these.

Trial

One motion capture data set from

EVaRT

. Sometimes called a "shot". This is one set of VC files and the corresponding TRB or TRC file.

Shot

Record

Same meaning as "Trial".

In database terminology, this is a line of data in the database table.

User Interface

Figure 15-1. User Apps > QuickDB Panel

Bring up the main QuickDB interface by selecting the

Open Access Database Interface

button in the User Apps > QuickDB panel.

15-3

Chapter 15: QuickDB

Figure 15-2. QuickDB Interface

EVaRT 5.0 User’s Manual

Project

Databases

This is the list of databases from the

MasterProjectsList.mdb

file. Select the session you wish to view by selecting any of the fields in the projects list. You can edit any of the fields by selecting them and typing in your changes. These changes take effect immediately and are stored in the database as soon as you make the change. To delete any item from the list, select the whole row and press the

Delete

key on your keyboard. This will remove the entry from the master list but it does NOT delete the corresponding session database file. This file is left alone.

When you start up the QuickDB interface, it will automatically move to the end of the projects list to put you on the most recently created entry.

15-4

EVaRT 5.0 User’s Manual

Add

Figure 15-3. Add DB Panel

Chapter 15: QuickDB

Folder/File

Selection

Client Name

This is to add a session database entry to the Master Project List for an existing session database. For example, if another

EVaRT

user has created a list for you to use, you can copy this .mdb file to a new location or you can add a reference to the original file created by the other user.

Select the database file you wish to add by browsing to the correct folder and then selecting the .mdb file.

This is a string field that allows you specify the client for the data. It is really nothing more than another comment field, but since this field is displayed first in the Project Databases list, it should be something that is unique and descriptive.

15-5

Chapter 15: QuickDB EVaRT 5.0 User’s Manual

Creation Date

Comments

Add

Cancel

New

Figure 15-4. New DB Panel

The creation date of the mdb file. The current date is entered by default, but you can change it if necessary.

A comment field limited to 255 characters. This is used to help you remember what the session database contains.

When all the above fields and sections are made, this button is enabled.

When pressed, a new record is added to the Project Databases list (which is saved in the

MasterProjectsList.mdb

file).

Cancel the Add operation at any time. You can also close the dialog window and cancel the operation using the Windows Close button in the upper right corner of the dialog.

15-6

This creates a new session database and places the new .mdb file in a location specified by the user. The default location is to use the

Userfiles\Databases

folder but the location of the database can be anywhere. An option to make the database in the same folder as the current project is provided. This makes the database easily relocated along with all the project and mocap data. This operation works by copying an existing mdb file (

SessionTemplate.mdb

) from the

Userfiles\Databases

EVaRT 5.0 User’s Manual Chapter 15: QuickDB

Database Name

Place New

Database In

Current Project

Folder

Client Name

Creation Date

Comments

Ok

Cancel

Trial List

Scan Folder

folder to the name of the new database. If the

SessionTemplate.mdb

file doesn't exist, then the operation will fail.

The name of the new database to be created.

A flag indicating that instead of placing the new database in the

Userfiles\Databases

folder, the new database should be located in the current project directory.

This is a string field that allows you specify the client for the data. It is really nothing more than another comment field, but since this field is displayed first in the Project Databases list, it should be something that is unique and descriptive.

The creation date of the mdb file. The current date is entered by default, but you can change it if necessary

A comment field limited to 255 characters. A reminder to help you remember what the session database contains.

When all the above fields and sections are made this button is enabled.

When pressed, a new record is added to the Project Databases list (which is saved in the MasterProjectsList.mdb file) and a new database file is created for the session database.

Cancel the new operation at any time. You can also close the dialog window and cancel the operation selecting the

Close

button in the upper right corner of the dialog.

The trial list contains the names of all the trials in the session database.

All the values for the currently selected item are shown in the fields to the right of the Trial List. Any changes take effect immediately and are immediately stored in the session database.

This brings up a dialog that allows the user to specify a folder to be scanned for pre-existing motion capture data. The folder can be scanned for one of VC, TRB or TRC files. One project file must be specified for all trials to be created (there isn't any way to determine which project should be associated with each trial by examining the trial data). The user is expected to go back later and update the project file for each trial.

When searching for a folder to scan, the dialog will look for files of each of the types and will disable the corresponding option if files of that type are not found.

A new trial will be created for each motion capture file found. The file names must adhere to the form described in the Usage Notes section. This is to allow the scanner to find the trial number for each file.

15-7

Chapter 15: QuickDB EVaRT 5.0 User’s Manual

You can scan as many folders as many times as you wish to add items to the session database. There is no requirement that all the trials be located in the same folder.

New (From

EVaRT)

This creates a new trial and adds it to the session database. Fields are filled in by inspecting

EVaRT

for information. This is just like selecting

New

(Blank) and then selecting

Sync with EVaRT

.

New (Copy

Current)

This creates a new trial and adds it to the session database. All the fields are copied from the currently selected trial. This is useful for creating a lot

of trials to be used later in a motion capture session (see Load Into EVaRT

for more information).

New (Blank)

Sync with

EVaRT

This creates a new trial and adds it to the database. All the fields are made to be blank.

Load Into EVaRT

This loads the information from the currently selected trial into

EVaRT.

The project file is loaded first. Then, if the VC file and any of the tracks files exists, they are loaded into the Post Process mode. If you have made changes to the currently loaded data, then this will be lost with no ability to recover.

If

EVaRT

is currently streaming live data (or in Record mode from a VC file) when you select this button, a warning message will be displayed indicating that no data was loaded.

If

EVaRT

is currently connected to cameras when you select this button, then ONLY the trial name, trial number, and any of the MTO projects will be updated in the Output panel of

EVaRT

. This is so that the connection to the cameras is not interrupted (as they would be if the project were loaded) and the trial list now becomes a template for a capture session.

Note:

QuickDB is not currently able to notify

EVaRT

that the Trial Name and

Trial Numbers have changed. If you select a panel other than the Output panel and then back to the Output panel you will see that the names have changed. Any subsequent records will go to the name specified by

QuickDB even if you do not refresh the Output panel.

This inspects settings in

EVaRT

and updates the current record to match.

This is a handy way of having

EVaRT

fill in the FPS, Duration and Timecode fields (only if a VC file has been loaded will this information be there). A useful trick after scanning a folder for trial names is to select each of the files, load into

EVaRT

and then select

Sync with EVaRT

. This information is immediately used to update the session database.

Record Trials

When selected, QuickDB will listen for Start Record and End Record events from

EVaRT

and will create a new Trial Record for each. The name of the trial will be taken from the Name field in the Settings area of the

Motion Capture > Output

panel.

15-8

EVaRT 5.0 User’s Manual Chapter 15: QuickDB

Trial Fields

FPS

Duration

Timecode

Props

Actors

Track Item

Tracker

Frame In

Project Directory

Project File

Trial Name

Trial Number

Trial Date/Time

File Types

Frame Out

These are all the fields to the right of the Trial List. If you modify these values then they will have immediate effect. The check boxes to the left of the fields indicate which fields are to be displayed in the trials list.

This is the directory where all the project and motion capture data is contained. A special name of "." indicates that the project directory is the same folder as the location of the session database.

The project file associated with the main markerset.

The name of the trial.

The number of the trial.

The time when the trial was captured. This is generally taken from the time stamp put on the file by Windows.

The types of files that exist for the trial. This is also used to indicate what kind of files are to be loaded into

EVaRT

. So it might be possible that a

TRC files exists, but if it is left unchecked then it will not be loaded into

EVaRT

when the

Load Into EVaRT

button is selected.

Frames Per Second—This information is provided by

EVaRT

to QuickDB when

Sync With EVaRT

is selected.

The length of the trial in frames. This information is provided by

EVaRT

to QuickDB when

Sync With EVaRT

is selected.

The Timecode value of the first frame of the trial. This information is provided by

EVaRT

to QuickDB when

Sync With EVaRT

is selected.

A comment field used to indicate what props were present during the capture of the trial.

A comment field used to indicate what actors were present during the capture of the trial.

The specific item (associated with a particular markerset) to be tracked and delivered by the user.

The person responsible for tracking the data.

The starting frame of the data to be delivered. Timecode is not used here because it isn't specific enough (more than one frame of data might fall on the same Timecode value).

The ending frame of the data to be delivered. Timecode is not used here because it isn't specific enough (more than one frame of data might fall on the same Timecode value).

15-9

Chapter 15: QuickDB EVaRT 5.0 User’s Manual

Tracking Status

Multiple Tracking

Objects

The current status of the trial.

Markerset names that are used to fill in the Additional Tracking Objects fields of the

Motion Capture > Objects

panel.

Frequently Asked Questions

I have pressed

Load Into EVaRT

but I get no response. What is happening?

If

EVaRT

is currently connected to cameras, then very little in

EVaRT

will change except the Name and Trial Number fields of the Settings area in the

Motion Capture > Output

panel and the Additional Tracking Objects list.

15-10

Appendix A

System Hardware

Interconnections

Overview

Topic

Standard Eagle and Hawk System Configuration

Power Consumption

Video Processor (MIDAS) Connections

Analog Camera Connections

Page

A-1

A-2

A-15

A-16

A-17

Configuring Ringlight Changes for Eagle and Hawk Cameras A-23

Overview

This chapter provides information and illustrations on how to set up the hardware to be used with

EVaRT

. It is broken up into sections for setting up the Eagle, Hawk, Falcon, Pulnix, and Cohu camera systems.

When using

EVaRT

in the Motion Analysis motion capture system, hardware connections are straight-forward. The connection of the cameras to the EagleHub1, EagleHub2, or EagleHub3 (Eagle and Hawk digital camera system) or the Midas box (Falcon, Pulnix, and Cohu) all have unique labeled connectors.

Note:

Frame rate, shutter speeds, and ring light brightness for the Eagle and

Hawk digital cameras are set using the

EVaRT

user interface.

A-1

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Standard Eagle and Hawk System Configuration

Figure A-2 shows a standard

EVaRT

system set up for use with Eagle and

Hawk digital cameras. It includes:

• a set of Eagle or Hawk digital cameras with Eagle network and power cables for each camera

• an EagleHub1 or EagleHub2 (1 for every 8 cameras) or

EagleHub3 (1 for every 12 cameras)

Figure A-1. EagleHub Will Power 8 Eagle or Hawk Cameras with 17 Ethernet Connections

• a Tracking Computer (host) with monitor, keyboard, and mouse

Figure A-2. Standard Eagle or Hawk System Configuration

A-2

EagleHub

EVaRT Host Computer

Eagle or Hawk Cameras

For more detailed diagrams, please refer to Figures A-3 through A-10.

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Power

Consumption

The maximum power consumption you can expect for some typical Eagle

and Hawk system configurations is shown in Table A-2

. The actual power consumed depends on the video frame rate and the intensity of the ring lights and is usually less than that indicated.

For the most reliable system operation, it is recommended that all camera assemblies and computers be powered by an uninterrupted power supply

(UPS). If you want to save your data when power is lost altogether, you will also need to power the VGA monitor from the UPS.

Table A-1. Power Consumption of Typical Eagle and Hawk Setups

EagleHubs with Cameras

21” SVGA Monitor

Dual Processor Computer

TOTAL

8 Cameras

265 W

125 W

200 W

580 W

12 Cameras 16 Cameras

400 W 530 W

125 W

200 W

725 W

125 W

200 W

855 W

Basics of

Ethernet

Switches and

Hubs

There is a difference between the older Ethernet hubs and the newer

Ethernet switches, even though they look alike and are functionally similar. The difference is in the performance. Ethernet switches guarantee the full rated Ethernet bandwidth between all ports simultaneously, whereas the older Ethernet hubs share the bandwidth for all ports. We call the EagleHubs “hubs” to indicate that they are the center connection point for a block of cameras, but inside the EagleHub resides a switch. This performance difference is important and necessary for the Eagle and Hawk systems to function properly.

A 100 Mbps Ethernet switch works well for an Eagle or Hawk system with up to 16 cameras (low to moderate frame rate). For larger numbers of cameras (above 16), it is important to use the 1 Gbps Ethernet NIC (network interface card) inside the computer and the 1 Gbps Ethernet switch that collects and concentrates the camera traffic to the

EVaRT

Host computer.

A-3

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Ethernet Tutorial and

Troubleshooting

Guide

Troubleshooting

There are two types of female Ethernet connections that use the same 8pin Ethernet connector:

1.

2.

NIC-Type—The Ethernet connector that is found on the Network

Interface Card on computers and on your Eagle or Hawk cameras.

Hub-Type—The standard Ethernet plug that is found on Ethernet switches and hubs.

There are two types of Ethernet cables:

1.

2.

Standard Ethernet patch cables

Ethernet cross-over cables

The standard patch cable is used to connect computers to Ethernet hubs and for connecting Eagle or Hawk cameras to the EagleHubs.

The Ethernet cross-over cable is used for connecting Ethernet hubs to other hubs, unless you use the Uplink port on either hub. In this case, you can use a standard patch cable. The cross-over cable would also be necessary if you were to bypass the Eagle hub and plug the Eagle or Hawk camera directly into your computer’s NIC.

You can tell if you have a live Ethernet connection if the indicator light goes on when you plug the cable into the hub. This is also the best way to figure out whether or not your Ethernet cables are plugged in correctly. It will not damage anything if you plug in the wrong type of cable (patch or cross-over) into an Ethernet jack. For the indicator light to go on, there has to be a live Ethernet connection on both ends of the cables.

A-4

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

EagleHub3

Connections

When first setting up your Eagle or Hawk system, you will notice that the both the power and Ethernet connections for the cameras are integrated through the EagleHub3(s). Remember, the Eagle or Hawk system allows for 12 cameras per EagleHub3.

All camera power connectors are plugged into the power connectors of the EagleHub3. Order is not imperative as long as each power connector is close to an open Ethernet connector.

Mixing EagleHubs

There is no problem if you mix EagleHub1, EagleHub2, and EagleHub3 hubs in the same system. The EagleHub provides power and Ethernet connections for both Eagle and Hawk cameras and can be mixed.

1 Gbps Switch

A 1 Gbps Ethernet switch is required for best performance if you have more than 16 cameras and plan to use the higher camera speeds available.

Figure A-3. Standard 1-12 Eagle Camera, Single EagleHub3 Configuration

EVaRT Host Computer

EagleHub

Power, Cameras 1 - 12

Integral A-D

(optional)

Camera Network, Cameras 1 - 12

EVaDV Computer

(optional)

Patch Cable

Eagle

Ethernet

NIC

Ethernet NIC for

Customer LAN

Eagle A/D Computer

(optional-older method)

A-D Interconnect Box

(optional)

A-5

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Figure A-4. Eagle or Hawk Camera, Multiple EagleHub3 Configuration (12+ cameras)

EagleHub3

Power, Cameras 1 - 12

EVaRT Host Computer

Camera Network, Cameras 1 - 12

EagleHub3

Power, Cameras 1 - 12

Eagle

Ethernet

NIC

Ethernet NIC for

Customer LAN

Camera Network, Cameras 12 - 24

1 Gbps Hub

EVaDV Computer

(optional)

Additional EagleHub3(s)

A-6

Note:

The standard multi-EagleHub3 setup is designed for the hubs to be distributed around the motion capture volume. If you desire to have the EagleHubs in a central location, you may need to use different camera cable lengths than those originally shipped with your system.

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

EagleHub1 and

EagleHub2

Connections

When first setting up your Eagle or Hawk system, you will notice that the both the power and Ethernet connections for the cameras are integrated through the EagleHub(s). The Eagle system allows for 8 cameras per each

EagleHub1 or EagleHub2 (EagleHub1|2).

All camera power connectors are plugged into the power connectors of the EagleHub1|2. Order is not imperative as long as each power connector is close to an open Ethernet connector.

Note:

The first seven cameras for each EagleHub1|2 are connected directly to the Ethernet connectors. The eighth camera must be connected to one of four ports of the Network Interface Card (NIC) on the rear of the host

computer or to a separate Ethernet switch as shown in Figure A-5

. This is done using a female to female (F-F) Ethernet adapter, and an extension

Ethernet cable.

Important

If using a frame rate greater than 120 Hz, you may only connect up to six cameras to any EagleHub1|2. Due to power considerations, using more than six cameras at high frame rates will dim the ring lights.

Figure A-5. Standard 8 Eagle or Hawk Camera, Single EagleHub1|2 Configuration

EVaRT Host Computer

EagleHub1|2

Power, Cameras 1 - 8

5 to 8 Port

Ethernet Switch

Patch Cable

Camera Network, Cameras 1 - 7

F - F

Camera Network, Camera 8—Direct to NIC of Host

Computer with F-F Adapter and Extension LAN Cable

Patch Cable

Eagle

Ethernet

NIC

Ethernet NIC for

Customer LAN

Note: If your Host Computer has a 4-Port NIC, you may connect the Patch Cable and the Camera 8 cable directly to it.

A-7

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Note:

If using the Uplink connector on the EagleHub to link to the Ethernet

Switch, you must use a crossover cable.

Figure A-6. Uplink Hub Connector with Cross-Over Cable

Crossover Cable to NIC on

Host Computer

Figure A-7. 8 Camera, 2 EagleHub1|2 Configuration

EagleHub1|2

Power

EVaRT Host Computer

Cameras 1 - 4

EagleHub1|2

Power

Patch Cable

5 to 8 Port

Ethernet Switch

Patch Cable

Eagle

Ethernet

NIC

Ethernet NIC for

Customer LAN

Cameras 5 - 8

Patch Cable

Note: If your Host Computer has a 4-Port NIC, you may connect the Patch Cables directly to it.

A-8

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Using More than 8

Cameras

If using more than eight Eagle or Hawk digital cameras, you must use an additional EagleHub1|2 (i.e one EagleHub1|2 for every eight cameras).

For example, if your system has 32 Eagle cameras, you will need four EagleHubs.

Each EagleHub1|2 configuration is the same for each 8 camera block, which then plugs into the NIC. For example, cameras 1 through 8 use Hub

#1 and cameras 9 through 16 use Hub #2.

Note:

If you are using two EagleHubs(1|2) for 16 cameras, you will need more than four NIC ports.

Note:

If you are using an Eagle Analog Computer and two EagleHubs(1|2), the following setup (with one NIC) limits you to using 15 Eagle cameras. The fourth port on the NIC must be used for the Eagle Analog Computer, not the 16th camera.

Figure A-8. 16 Camera, 2 EagleHub1|2 Configuration

EagleHub1|2

EVaRT Host Computer

Power

Cameras 1 - 7

F - F

Camera 8, Direct to NIC with F-F Adapter

EagleHub1|2

Patch Cable

5 to 8 Port

Ethernet Switch

Patch Cable

Eagle

Ethernet

NIC

NIC for

Customer LAN

Power

Customer Supplied Hub

Cameras 9 - 15

F - F

Patch Cable

Camera 16, Direct to NIC with F-F Adapter

EVaDV

Computer

(optional)

Animation

Plugins

Computer

(optional)

Note: If your Host Computer has a 4-Port NIC, you may connect the Patch Cables and 8th and 16th camera cables directly to it.

A-9

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Figure A-9. 16 Camera, 3 EagleHub1|2 Configuration

EagleHub1|2

Eagle Analog Computer

(optional)

Power

EVaRT Host Computer

Cameras 1 - 5

EagleHub1|2

Power

5 to 8 Port

Ethernet Switch

(1 GBit)

Eagle

Ethernet

GBit NIC

NIC for

Customer LAN

Cameras 6 - 10

EagleHub1|2

Customer Supplied Hub

Power

Cameras 11 - 16

EVaDV

Computer

(optional)

Animation

Plugins

Computer

(optional)

Note: If your Host Computer has a 4-Port NIC, you may connect the Patch Cables directly to it.

A-10

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Figure A-10. 24 Camera, 4 EagleHub1|2 Configuration

EagleHub1|2

Power

Eagle Analog

Computer

(optional)

1 Gbps Switch

Cameras 1 - 6

EagleHub1|2

Power

Host Computer

Cameras 7 - 12

EagleHub1|2

Power

Cameras 13 - 18

Cameras 19 - 24

EagleHub1|2

Power

Eagle

Ethernet

NIC

NIC for

Customer LAN

Customer Supplied Hub

EVaDV

Computer

(optional)

Animation

Plugins

Computer

(optional)

A-11

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Eagle and Hawk

Digital Camera

Connections

Eagle or Hawk digital cameras are connected to the EagleHubs using camera cables with both power and Ethernet connectors. When fitting connectors together, be sure the connections are secure and snap firmly into place.

Figure A-11. Eagle Rear Panel Connectors

Future Ethernet

Aux

Power

Power Connector

The power connector powers the camera with a 48 Vdc source from the

EagleHubs, through a CAT5 cable.

Ethernet Connector

The Ethernet connector is set for a 4-wire, full duplex 100 Mbps Ethernet.

Aux Connector

Future Connector

The Aux connector can be used for testing VGA and diagnostics and is generally not needed for normal customer use. An Aux cable is supplied with each Eagle and Hawk system which has three connectors on one end

(VGA, COM 1, and BNC) for use in various applications and diagnostic testing.

The Future connector is reserved for future use.

A-12

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Network Interface Cards and IP Addresses

Unless you are using a 1 Gbps hub (see

Figure A-10 ) for a large number

of cameras, your computer must have two Network Interface Cards

(NIC):

1.

2.

Single port for your own LAN connection

4-Port dedicated to the Eagle or Hawk system

Configuring a

Network with

Your Eagle

Cameras

Single Port

—For use with your LAN. If you are using the same

10.1.1.xxx addressing, you must use a fixed IP address. If you are using a different addressing scheme from what the Eagle and Hawk cameras are using, you can use DHCP (dynamic IP addressing).

In a command prompt window, you should be able to ping your Eagle and

Hawk camera with the following:

> ping [IP address of camera]

For example, for a camera with an IP address of 10.1.1.205:

> ping 10.1.1.205

Entering the command

> IPCONFG

will tell you the status of your IP connections.

Dedicated Interface: Part of

Setup > Cameras > Eagle

. Use the IP address of the 4-port NIC that is connected to your Eagle cameras

(10.1.1.199 is recommended).

A-13

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Analog Camera System Configuration

Figure A-12 shows a standard

EVaRT

system using a Falcon, Pulnix, or

Cohu digital camera. It includes:

• a Video Processor (MIDAS) computer

• a Tracking Computer (host)

• an SVGA Monitor shared by the Video Processor and Tracking

Computer

A MultiSync (Threshold) Monitor for viewing raw video data

Figure A-12. Standard Falcon, Pulnix, or Cohu System Configuration

Threshold Monitor

SVGA Monitor for

Tracking Computer and Video Processor

EVaRT Tracking

Computer

Monitor

A / B

Switch

Midas Box

A-14

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Power

Consumption

The maximum power consumption you can expect for some typical sys-

tem configurations is shown in Table A-2 . The actual power consumed

depends on the video frame rate and the intensity of the ring lights and is usually less than that indicated.

For the most reliable system operation, it is recommended that all camera assemblies and computers be powered by an uninterrupted power supply

(UPS). If you want to save your data when power is lost altogether, you will also need to power the VGA monitor from the UPS.

Table A-2. Power Consumption of Some Typical Systems

Falcon Camera Assy.

50 W max. each

6 Cameras

300 W

Midas Computer with VPAT cards

200 W

21” VGA Monitor

125 W

21” Threshold Monitor

Dual Processor Computer

TOTAL

140 W

200 W

965 W

10 Cameras 16 Cameras

500 W 800 W

250 W

125 W

140 W

200 W

1215 W

300 W

125 W

140 W

200 W

1565 W

A-15

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Video Processor

(MIDAS)

Connections

Back panel connections for a standard configuration of the Video Processor (

MIDAS

for

M

otion

I

ntegrated

D

ata

A

cquisition

S

ystem) are shown in

figure A-13

.

Note:

Other configurations may have slightly different arrangements or have the position of the video processor cards and computer interface reversed.

Figure A-13. Video Processor (MIDAS) Connections

Computer interface Video processor cards

A-16 keyboard connector mouse connector

Ethernet connector

SVGA monitor connector parallel port (printer) serial port (COM2)

not used

Video output to individual video monitors

Camera connectors to individual cameras

(video, sync)

Daisy chained video output to threshold video monitor.

Camera # selected in software

(grey scale & binary)

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Analog Camera

Connections

Video cameras are connected to the Video Processor using camera cables.

There are two available cable configurations.

1.

2.

A camera cable with a 25-pin connector is connected directly to the

Video Processor. Then, a long extension cable, with BNC connectors, is installed between this cable and the camera.

The long extension cable is connected to a back panel assembly rather than the camera cable described in Step 1.

Figure A-14. Standard Pig-Tail Cable

25 pin connector

(connects to video processor - MIDAS)

The long extension cable to the camera connects here horiz. sync.

red

vert. sync.

green blue

video audio in audio out

These only exist on the cable for

Channel 1

9 pin event trigger connector

(to trigger push-button)

RED event trigger wire

(connects to optional analog input board, pin 38)

Figure A-15. Back Panel Camera Connector Assembly

THRESH AUDIO OUT

12 11 10 9 8 7 6

HD

5 4 3 2 1

EVENT TRIG

TIME CODE AUDIO IN

VD

VID

A-17

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Connections to

Specific Analog

Cameras

The connections to a specific camera depends on the manufacturer. The individual red, green, and blue cables from the long extension cable are combined in short-stub cable specific to a particular type of camera. The following pages show diagrams for connecting to a variety of cameras.

Figure A-16. FALCON HR 240 Camera and Strobed LED Connections

SEL Switch set at 1

VIDEO A

STROBE VIDEO B

DEFAULT

RATE

60

60 D

120

120 D

180

240

4000

2000

1000

500

250

OFF

SHUTTER

VIDEO A & VIDEO B

Stand-alone Video

Connectors

AUX

,

when connected, overrides the RATE switch, allowing software control of frame rate.

DEFAULT RATE SWITCH

Used Only with Stand-alone Mode

(AUX cable disconnected)

AUX PWR

To VPAT Board

DEFAULT RATE POSITIONS

60

60D

60 Hz, 240 lines

60 Hz, 480 lines

120 120 Hz, 240 lines

120D 120 Hz, 480 lines

180

240

180 Hz, 240 lines

240 Hz, 240 lines

s s fp

0

fp

24

60

-

60

To Power Supply

± 13Vdc

+ 6 Vdc

BNC Sync Cable

Strobe Duration

60 - 240 fps

(green scale)

1 = 1/8000

(sec)

2 = 1/4000

3 = 1/2000

4 = 1/1000

60 fps only

(blue scale)

5 = 1/500

A-18

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Figure A-17. Pulnix Camera Switches and Connections

Set at

1 or 2

4

5

2

3

0

1

8

9

6

7

Shutter Control Switch

Manual shutter mode

Note: Shutter mode

#4 is recommended no shutter

1/125

1/250

1/500

1/1000

1/2000

1/4000

1/9000

1/16000

1/32000

8

9

7

6

Shutter

Speed

0 1

2

5

4

3

Video

Gain

ASY N O

NFM P T

Power

Switch 1 - MAN Switch 2 - N

7

8

9

6

Shutter

Speed

0 1

2

3

4

5

Video

Gain

ASY

NFM

N

P

O

T

240

60

60 -

60 fps, 480 lines

120 fps, 240 lines

}

Switch 3

Power

Normal mode: N O (60 Hz progressive scan)

Double scan: N T (120 Hz two row scan)

A-19

Appendix A: System Hardware Interconnections

Figure A-18. Cohu 4915 60 Hz Camera Connections

(Inside Select Switches) before Revision E

EVaRT 5.0 User’s Manual

Camera & Strobe Power

(from 12 V, 800 mA

DC Power Adapter)

Cohu 4915 optional

NI = Non interlaced

3-D-with AUX External Driver

“Normal” position --- DOWN

12 Volt DC from camera/strobe power supply-polarity does not matter (either wire to either connector)

LENS

NI

AUX

Sync

VIDEO

Norm

40 60 - 2

60

12V AC/DC

Strobe Duration

1 = 1/8000

2 = 1/4000

3 = 1/2000

4 = 1/1000

5 = 1/500 nominal values

LED On Duration is synced to the camera 60 Hz V sync.

The numbers indicate how long during that 60 Hz interval the LEDs stay on.

To Camera connector

A-20

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Figure A-19. Summary of Camera and Strobe Settings

1

60

-

24

0 fp s

6

0

f p s

PULSE

SELECT

2

3

5

4

SYNC IN

1

2

3

SEL

DC IN

Revision B & C

(Earlier version - Does not have external Select Switch)

Dip Switches located inside the Strobe Control Box.

Remove Cover For Access To Dip Switches.

Revision E Strobe Select Switch

(Latest version available - effective May 19, 1994)

Slide switch to desired position

Type of Camera

FALCON HR 240

Pulnix TM-6701AN-534

COHU 4915 W/SW

Type of Strobe with Switch Setting per

Camera Type

Strobe Revision B & C

Dip Switch Settings

N/A

1 & 2 OFF, 3 & 4 ON

1 & 2 ON, 3 & 4 OFF

Strobe Revision E

SEL Switch Position

Position 1 (top)

Position 2 (middle)

Position 1 (top)

A-21

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

Figure A-20. Push-Button Switch For Eagle or Hawk System

Event Push-button Switch

(Normally Open)

Ground

Event 1 IN

Schematic

Black

White

10k

5

4

6

DE9P Plug

COM1 9-pin Female Connector

Event Push-Button Switch

DE9P Plug

To COM 1 Port on EVaRT

Host Computer

Figure A-21. Push-Button Switch For Midas System

Event Push-button Switch

(Normally Open)

Event 1 IN

Ground

Schematic

8

1

DE9P Plug

For computer emulation of this switch closure, use an open collector or TTL driver that goes low

(=0v) when data capture starts.

Midas Event Push-Button Switch

DE9P Plug

To VPAT Card #1 on Midas Computer

A-22

EVaRT 5.0 User’s Manual Appendix A: System Hardware Interconnections

Configuring Ringlight Changes for Eagle and Hawk Cameras

If you change the ringlight on your Eagle or Hawk camera, you must tell the camera what kind of ringlight it currently has attached. If you do not, it could send too much power to the ringlight and damage the electronic circuitry or LEDs.

Parts You Will

Need

1.

2.

Computer with COM port and HyperTerminal software (standard with Windows).

Eagle/Hawk Test Cable (shown below) that came with your Eagle or

Hawk system. This cable is about 6 feet long, has special connector one end to plug into the back of the camera. The other end has three connectors: a 9 pin COM port, a 15 pin VGA, and single BNC.

Figure A-22. Eagle/Hawk Test Cable

Steps

1.

2.

3.

Put the camera on a table, connect the single connector end of the

Eagle/Hawk Test cable to the back of the camera.

Connect the Eagle/Hawk Test cable’s COM port to the COM1 port of your computer. COM2 may also be used.

Launch HyperTerminal. If you have installed

EVaRT

, there should be a file named

EagleCOM1.ht

in the folder containing the

EVaRT

executable (for example:

C:\Program Files\Motion Analysis\EVaRT50\

EagleCOM1.ht

. Double click on the HT file name and HyperTerminal should launch. If you are on a different computer, look from the

Start Menu under

Programs > Accessories > Communications >

HyperTerminal

. This allows you to view and type messages to the

A-23

Appendix A: System Hardware Interconnections EVaRT 5.0 User’s Manual

4.

software in the Eagle or Hawk camera. You must quit

EVaRT

(or uncheck

Motion Capture > Output: > Enable External Trigger

to free up the COM1 port. COM port settings: 9600 8-N-1.

Boot the camera by plugging it into the EagleHub or turning the power off and on to your EagleHub. After a few seconds you should see messages similar to the following example:

* * * * * * * * * * * * * * * * * * * * * * * *

* MAC Camera Control Program

* Camera Configuration version 1.4

copyright

(c) 2003, Motion Analysis, Inc.

* NET+WORKS Version 3.00

copyright (c) 2000,

NETsilicon, Inc.

* * * * * * * * * * * * * * * * * * * * * * * *

Serial channel used for diagnostics will use a baud rate of 9600.

After the camera board is reset, the camera will wait 5 seconds for the user to signal any changes on the keyboard.

- - - - - - - - - - - - - - - - - - - - - - - -

Press any key within 5 seconds to change these settings.

5.

6.

7.

Press the Space bar (or any other key) to change the camera configuration.

Press

M

to modify the camera settings

Press

Enter

about 5 times to leave the other settings unchanged. Wait for the message that tells which Ringlight is currently configured. If the ringlight type is infrared, then the display should be Ringlight

type = infrared. (7). If the ringlight type is red, then the display should be Ringlight type = red. (1). For example, to change from: near Infrared (4) to red (1), type in the 1 character as:

Ringlight type = near infrared. (4) 1

(You type in the single digit 1 followed by Return)

8.

9.

Press

Enter

to leave the other items unchanged.

Wait for the camera to boot again and check that Ringlight type was successfully changed.

A-24

Appendix B

Analog Input Hardware and

Software

Topic

Overview

Installing NIDAQ Software on an EVaRT Computer

Analog Signal Naming Conventions

32-Channel, 16-Bit NI USB-6218 Configuration

64-Channel SCB-100 and NI PCI-6071E Configuration

Page

B-1

B-3

B-4

B-5

B-14

Overview

The

EVaRT

system can accept analog data from external devices and synchronize it with video motion data.

Analog cards known to work with

EVaRT

software include the following

A-D configurations and the necessary NIDAQ software.

EVaRT

will support one or two of the devices listed in

Table B-1

. The devices must have the same resolution (12-Bit or 16-Bit). Other NI Analog A-D input configurations should work but have not been tested.:

Table B-1. A-D Configurations Used with EVaRT Software

A-D Configuration

NI USB-6218, 32-Channel, 16-Bit (up to 6

USB devices)

NI PCI-6071E, 64-Channel, 12-Bit

NI PCI-6071E, 64-Channel, 12-Bit (higher performance)

NI DAQ Card-6024E, 16-Channel, 12-Bit

NI PCI 6254, 32-Channel, 16-Bit

(up to 2 cards)

NI USB-6259, 32-Channel, 16-Bit

(up to 2 USB devices)

NIDAQ Software

NIDAQ 8.3 or later, EVaRT 5.0.3 or later

Traditional NIDAQ 7.0 > 7.4, EVaRT 4.4

NIDAQ MX 8.0 or later, EVaRT 5.0

Traditional NIDAQ 7.4, EVaRT 4.6

NIDAQ MX 8.0 or later, EVaRT 5.0

NIDAQ 8.1 or later, EVaRT 5.0.2

Note:

If you are using NIDAQ MX 8.0 or 8.1 versions and if you are collecting data for only one channel in

EVaRT

, you will need to install a jumper wire from screw terminals PFI 7 to PFI 0 on the A-D interconnect box. This

B-1

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Performance

Specifications

Separate A-D

Computer

may be changed in future versions of the NI software drivers. If you have two analog acquisition devices installed, the same is applied for each device: If only one channel is sending data on that device, the PFI7 to PFI0 jumper must be installed.

This appendix documents both of the 32-channel (16-bit) setups and the

64-channel setups of analog input available, accessed by means of the

Analog Terminal Box which contains 100 screw terminals. The 32-channel setups are also broken down to USB and PCI connections. A list of the location of channel numbers in the Analog Terminal Box for digital camera systems is given for each type of A-D setups. You may find it useful to make a copy of this chart and use the column titled Setup Name to record the connections for your installation. The connections for a typical application using two AMTI Force Plates and ten EMG channels for digital cameras is shown for each setup.

Although connecting analog inputs is not particularly difficult, it is important that certain naming conventions be followed for the external data to work smoothly with supplementary Motion Analysis software such as

KinTrak

and

OrthoTrak

. This is described in

“Analog Signal Naming

Conventions” on page B-4

.

The Motion Analysis Digital Camera system is capable of collecting up to

192 channels of analog data at any frequency between 60 and 5000 Hz. In newer systems, using the NIDAQ MX 8.0 or later software, analog rates can be much higher. The maximum rate can be up to 255 times the video capture rate, but performance may vary with different computers. The master digital camera provides the clocking signals to the A-D card in the

A-D computer, which provides the phase-locking mechanism. You must connect the A-D cable from the master camera to the A-D interconnect box. Data can be collected in the pause mode or the run (live) mode, without any delay or drift between analog and video signals.

This configuration is no longer supported in the

EVaRT 5.0

software.

Please consult [email protected] for further details.

B-2

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Installing NIDAQ Software on an EVaRT Computer

For a new installation, start with Step 4 below. For

EVaRT 5.0

and later software, we recommend installing NIDAQ version 8.0 or later as it gives better analog performance and allows higher analog sample rates. With

NIDAQ 7.1 through 7.4, the maximum analog sample rate is 5000 samples/sec for all channels. With NIDAQ 8 and above, you can go to higher rates (typically 20,000 samples per second) for all channels.

Note:

NIDAQ version 7.5 DOES NOT WORK with any version of

EVaRT

. Versions of NIDAQ software are available for downloading from www.

ni.com.

If NIDAQ

Software is

Already Installed

1.

2.

3.

Shutdown/power OFF the computer and remove the A-D card.

Power ON the computer, go to

START/Settings/Control Panel, Add/

Remove Programs,

select

NI_DAQ

and then select

Remove

.

When complete, shutdown/power OFF the computer, wait 10 seconds, power ON the computer, let the system boot up, and then log-in when prompted. It is necessary to have the computer boot without the

A-D card or software. Proceed to the next step.

New Installation of NIDAQ

Software

1.

2.

3.

4.

5.

6.

7.

Install the

NI-DAQ software, version 8.0 or later

. Install all of the default entities that are checked, then shutdown/power OFF the computer. Note that

EvaRT 5.0

+ users can use either NIDAQ 7.1 through

7.4 or NIDAQ 8.0 or later. NIDAQ 8 allows higher sample rates and better overall analog performance.

EvaRT 5.0

uses the newer NIDAQ

MX libraries whereas earlier versions of

EVaRT

use the “Traditional”

(Legacy) NIDAQ libraries. Also, note that NIDAQ 7.5 does not work at all with

EVaRT

applications.

Power ON the computer and let it boot completely without the A-D card installed. This will complete the National Instruments software installation. Once the computer is completely booted, Shutdown/ power OFF the computer once again.

Install the A-D card and power ON the computer. It should come up with the Hardware Wizard and the “Found new hardware” pop-up window. At this point, the computer will automatically install the NI-

DAQ drivers correctly.

Shutdown/power OFF the computer one last time and then power it back ON.

Go to the National Instruments Test and Measurements software and select Traditional NIDAQ devices and then right click and select test panel.

Run through some of the channels to verify that the board is seeing the data. (test by having someone step on the forceplate).

Close and launch

EVaRT

and connect to cameras. You should see that all cameras are found as well as the A/D.

Note:

If you have NIDAQ 7.0 + drivers already installed onto your system, it is not necessary to un-install the software when upgrading.

B-3

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Analog Signal Naming Conventions

Kistler Forceplates

8 channels per plate:

Analog Channel 1 connects to the F1X1 signal.

Analog Channel 2 connects to the F1X3 signal, etc.

If there are two plates:

Analog Channel 9 connects to the F2X1 signal.

Analog Channel 16 connects to the F2Z4 signal.

The reserved names for

OrthoTrak

and

KinTrak

are:

PLATE #1: F1X1 F1X3 F1Y1 F1Y2 F1Z1 F1Z2 F1Z3 F1Z4

PLATE #2: F2X1 F2X3 F2Y1 F2Y2 F2Z1 F2Z2 F2Z3 F2Z4

Note:

These naming conventions are already set up in the Analog sub-panel.

For AMTI or Bertec

Forceplates

When connecting force plates and EMG equipment to the

EVaRT

system, certain requirements must be met and conventions followed.

Typically, forceplates are connected to the first channels of the A/D system and then the EMG channels. Specific Analog signal names for the forceplates must be used if

KinTrak

and

OrthoTrak

are used. These names depend on the forceplate manufacturer.

6 channels per plate:

Channel 1 connects to the F1X signal.

Channel 2 connects to the F1Y signal.

Channel 6 connects to the M1Z signal.

If there are two plates, it connects to Channels 7 through 12. The reserved names for

OrthoTrak

and

KinTrak

are:

PLATE #1: F1X F1Y F1Z M1X M1Y M1Z

PLATE #2: F2X F2Y F2Z M2X M2Y M2Z

EMG Signal Name

Conventions

For

KinTrak

, you must specify the channel names in the

KinTrak

project definition as well as in the

EVaRT

analog setup screen, which is saved in the

EVaRT

project file.

See the

OrthoTrak Reference Manual

for

OrthoTrak

muscle name conventions.

B-4

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

32-Channel, 16-Bit NI USB-6218 Configuration

Note:

The following has been tested on Windows XP Pro and Windows 2000

Pro with the latest web-based updates from Microsoft. The

EVaRT

software will support up to 6 USB devices and up to 192 channels of analog input. USB 2.0 ports work best. USB 1.0 ports will also work but with reduced data rates.

Note:

Up to 2 cards can be installed into the host computer. The Sync Cable must be connected to both cards.

Figure B-1. 32-Channel, 16-Bit A-D Hardware Setup for the USB-6218 A-D Card

First: Install the

Software

Before you plug the NI USB device into the host computer, you must first install the NI Acquisition software. You need to have installed the

NIDAQ software version 8.3

(or later) and you must be using

EVaRT 5.0.3 or later

for using the NI-USB 6218 A-D device. Install the National Instruments software, accept all the defaults, then let it finish and re-boot your computer. This may take 10-15 minutes. The remainder of the process takes less time. For more information, please refer back to

“Installing

NIDAQ Software on an EVaRT Computer” on page B-3 .

B-5

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Second: Install the Hardware

Note:

1.

2.

Plug in the A-D device (NI USB-6218 will be used as the example below).

You will then automatically go through the “Found New Hardware

Wizard” operation for a USB-device.

You will do this two times. The first round is for the 621x Loader. The next round is then for the 6218 Device.

3.

Select

Yes, this time only

, then

Next > .

Figure B-2. Found New Hardware Wizard Interface—First Round

4.

Click

Next >

then

Finish

.

Figure B-3. Completing the Found New Hardware Wizard Interface

B-6

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

5.

Now it repeats for the USB Device USB-6218.

Figure B-4. Found New Hardware Wizard Interface Repeated

6.

Click

Next >, Next >,

then

Finish

. A message pops up that says the new hardware is installed and ready to use. Select

Take No Action

, and then check

Always Take This Action

.

Figure B-5. New Data Acquisition Device Interface

B-7

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Notes

You will get two message sequences for EACH USB A-D device that you plug in and you will may get it again if you plug it into a different

USB port, so it is simpler to plug into the same port each time. It seems to work fine either way.

Run Test Panels….

Setting the NI software to “Referenced Single

Ended”: While it is not necessary for the

EVaRT

software to work, to get the correct looking signals on the Test Panels display, you need to set the

Analog Input > Input Configuration

to

RSE

(Referenced Single Ended). The

EVaRT

software sets this mode as part of its analog setup procedures, which is why it is not needed for

EVaRT

to work properly. Also, to have more than 8 channels displayed in the Test

Panels (0-7), you must change this setting to

Referenced Single

Ended

from its Differential Mode default.

EVaRT

starts numbering the analog channels starting at 1 whereas the

Test Panels and chart (

Table B-2 on page B-13

) starts numbers at channel 0.

Figure B-6. NI USB-6218 Configuration Interface

B-8

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Installing the Clock

Wiring from the

Master Camera

Figure B-7. Clock Wiring on Rear of NI USB-6218

Master Camera Red (or white): Connects to Pin 1 (PFI0)

Master Camera Black: Connects to Pin 11 (D GND)

2.7 k

Ω Resistor connects from Pin 1 (PFI0) to Pin 10 (+5 V)

Note:

NI USB-6259 uses PFI7 for clocking.

B-9

Appendix B: Analog Input Hardware and Software

Figure B-8. NI USB-6218 Pinouts

EVaRT 5.0 User’s Manual

B-10

The USB device pinouts are available online after you install the device.

To find them:

1.

2.

Launch

Measurement & Automation

which was installed when you installed the NIDAQ software.

Select

Devices and Interfaces

, then

NI-DAQmx Devices

and the page should appear as shown above.

Note:

NI Numbering starts at Channel 0 where

EVaRT

channel numbering starts at 1. For example: AI 0 (above) corresponds to

EVaRT

Channel 1.

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Which USB Device is Channels 1-32, which is 33-64?

The first device you plug in should be channels 1-32 in your

EVaRT

software. The second device should be channels 33-64 and so on for more devices. If you are not sure: When you connect to the Cameras in

EVaRT

5.0.3 or later, the channel numbers (1-32) and the Serial Number of the

USB device are reported in the dialog box. The Serial Number for the

USB-6218 device is located on the bottom of the USB device.

Clock Wiring for Single and Multiple USB Devices

Eagle and Hawk cameras require a 2.7 k

Ω pull-up resistor from +5 Volts to the PFI-0 pin as shown

Figure B-9. Clock Wiring for Mutiple Devices

A Single Pull-up resistor with a 2.7 k

Ω (or close) value will work for connecting up to 6 USB devices together as shown below. You need only to connect PFI0 and D GND signals in parallel for clocking the multiple

USB Devices. A single pull-up resistor works for all devices.

USB 2.0 Cable Lengths

The USB-6218 comes with a one meter USB cable. It has been tested with longer USB cable lengths up to 16 ft. (5 meters) and it works well with either and any mix of cables.

B-11

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

USB Expansion Ports

Internal or External USB Expansion devices also have been tested and appear to work well without problems. This allows you use a single USB port on the main capture computer and plug in as many USB devices as needed.

B-12

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

40

41

42

43

44

45

46

47

48

33

34

35

36

37

38

39

EVaRT

Channel #

CH 1

CH 9

AI GROUND

CH 2

CH 10

AI GROUND

CH 3

CH 11

AI GROUND

CH 4

CH 12

AI GROUND

NOT USED

AI GROUND

NOT USED

NOT USED

CH 5

CH 13

AI GROUND

CH 6

CH 14

AI GROUND

CH 7

CH 15

AI GROUND

CH 8

CH 16

NOT USED

NOT USED

AI GROUND

NOT USED

NOT USED

Screw

Terminal

#

20

21

22

23

24

25

14

15

16

17

18

19

29

30

31

32

26

27

28

8

9

10

11

12

13

1

2

3

4

5

6

7

CH 17

CH 25

AI GROUND

CH 18

CH 26

AI GROUND

CH 19

CH 27

AI GROUND

CH 20

CH 28

AI GROUND

NOT USED

AI GROUND

NOT USED

AI GROUND

Table B-2. Analog Input Channel Connections and Master Camera Clocking (NI USB-6218)

Setup

Name

88

89

90

91

92

93

81

82

83

84

85

86

87

94

95

96

EVaRT

Channel #

PFI 1

PFI 2

PFI 3

PFI 4

PFI 5

PFI 6

PFI 7

CH 21

CH 29

AI GROUND

CH 22

CH 30

AI GROUND

CH 23

CH 31

AI GROUND

CH 24

CH 32

AI GROUND

NOT USED

AI GROUND

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

PFI 0

Screw

Terminal

#

68

69

70

71

72

73

62

63

64

65

66

67

77

78

79

80

74

75

76

56

57

58

59

60

61

49

50

51

52

53

54

55

Jumper Cable

Setup

Name

4.7 k

Ω Resistor & A/D Sync Cable &

Jumper Cable

PFI 8

D GROUND

PFI 9

D GROUND

PFI 10

D GROUND

PFI 11

D GROUND

PFI 12

D GROUND

PFI 13

D GROUND

PFI 14

D GROUND

PFI 15

+5 V

A/D Sync Ground

4.7 k

Ω Resistor

B-13

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

64-Channel SCB-100 and NI PCI-6071E Configuration

Figure B-10. Digital Camera System and SCB-100/NI PCI-6071 Hardware Setup

Master Camera

Host Computer

Eagle Hub

Ethernet Cable

Eagle/Hawk Camera Cable

Master A-D

Sync Cable

USB Cable

Ethernet Cable

Ethernet

Switch

SIMM Computer (optional)

Ethernet Cable

NI USB-6218 or

NI USB-6259

Important

1.

2.

3.

Channels are numbered 1 through 64 in the

EVaRT Setup > Analog

panel, whereas the manufacturer of the analog board,

National

Instruments

, uses channel numbers 0-63.

Channels which are marked as

NOT USED

should not be connected to any external circuitry or damage or malfunction may result.

If you are using an optional second PCI-6071E (for up to 128 channels), you need to have the Master A-D Sync cable connected to both cards. Only one 4.7 k-ohm resistor is required to be installed into both

PCI-6071E units.

B-14

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Figure B-11. Master Camera Connections (SCB-100 and PCI-6071E A-D Card)

A/D Sync Ground to Pin 33

4.7 k-ohm resistor from Pin 34 to

Pin 46

A/D Sync Cable to Pin 46

Typical Forceplate or

EMG Cable

A/D Sync Cable to

Master Camera

B-15

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Table B-3. Analog Input Channel Connections for SCB-100 Used with PIC-6071E A-D Card

Screw

Terminal

#

42

43

44

38

39

40

41

45

46

47

48

49

50

32

33

34

35

36

37

26

27

28

29

30

31

20

41

22

23

24

25

14

15

16

17

18

19

8

9

10

11

12

13

1

2

3

4

5

6

7

Channel #

Setup

Name

Screw

Terminal

#

GROUND

GROUND

CH 1

CH 9

CH 2

CH 10

CH 3

CH 11

CH 4

CH 12

CH 5

CH 13

CH 6

CH 14

CH 7

CH 15

CH 8

CH 16

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

GROUND

+5 V

NOT USED

NOT USED

NOT USED

A/D Sync Ground wire

4.7 k

Ω Resistor, End A

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

92

93

NOT USED 94

NOT USED 95

PF17 StarScan 4.7 k

Ω Resistor and A/D Sync Cable

96

NOT USED 97

NOT USED

NOT USED

NOT USED

88

89

90

91

98

99

100

82

83

84

85

86

87

76

77

78

79

80

81

70

71

72

73

74

75

64

65

66

67

68

69

58

59

60

61

62

63

51

52

53

54

55

56

57

Channel

#

CH 58

CH 51

CH 59

CH 52

CH 60

CH 53

CH 61

CH 54

CH 62

CH 55

CH 63

CH 56

CH 64

GROUND

CH 37

CH 45

CH 38

CH 46

CH 39

CH 47

CH 40

CH 48

CH 49

CH 57

CH 50

CH 17

CH 25

CH 18

CH 26

CH 19

CH 27

CH 20

CH 28

CH 21

CH 29

CH 22

CH 30

CH 23

CH 31

CH 24

CH 32

CH 33

CH 41

CH 34

CH 42

CH 35

CH 43

CH 36

CH 44

NOT USED

Setup

Name

B-16

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Table B-4. Typical Connections For 2 AMTI Forceplates and 10 EMG Channels for SCB-100

Screw

Terminal #

44

45

46

47

48

49

50

41

42

43

37

38

39

40

31

32

33

34

35

36

25

26

27

28

29

30

19

20

41

22

23

24

13

14

15

16

17

18

7

8

9

10

11

12

1

2

3

4

5

6

Channel #

Setup

Name

Screw

Terminal #

GROUND

GROUND

CH 1

CH 9

CH 2

CH 10

CH 3

CH 11

CH 4

CH 12

CH 5

CH 13

CH 6

CH 14

CH 7

CH 15

CH 8

CH 16

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

Forceplate Grounds

EMG Grounds

F1X

F2Z

F1Y

M2X

F1Z

M2Y

M1X

M2Z

M1Y

EMG01

M1Z

EMG02

F2X

EMG03

F2Y

EMG04

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

GROUND

+5 V

NOT USED

NOT USED

A/D Sync Ground wire

4.7 k

Ω Resistor, End A

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED 94

NOT USED 95

PF17 StarScan 4.7 k

Ω Resistor and A/D Sync Cable

96

NOT USED 97

NOT USED 98

NOT USED

NOT USED

99

100

91

92

93

87

88

89

90

81

82

83

84

85

86

75

76

77

78

79

80

69

70

71

72

73

74

63

64

65

66

67

68

57

58

59

60

61

62

51

52

53

54

55

56

Note: A-D interconnect schemes will be similar for different analog configurations.

Channel

#

CH 50

CH 58

CH 51

CH 59

CH 52

CH 60

CH 53

CH 61

CH 54

CH 62

CH 55

CH 63

CH 56

CH 64

NOT USED

GROUND

CH 37

CH 45

CH 38

CH 46

CH 39

CH 47

CH 40

CH 48

CH 49

CH 57

CH 23

CH 31

CH 24

CH 32

CH 33

CH 41

CH 34

CH 42

CH 35

CH 43

CH 36

CH 44

CH 17

CH 25

CH 18

CH 26

CH 19

CH 27

CH 20

CH 28

CH 21

CH 29

CH 22

CH 30

Setup

Name

EMG05

EMG06

EMG07

EMG08

EMG09

EMG10

B-17

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

Table B-5. Analog Input Channel Connections for Midas-Based Camera Systems (with the

SCB-100 A-D Interconnect Box and PCI-6071E A-D Card)

Channel #

Setup

Name

GROUND

GROUND

CH 1

CH 9

CH 2

CH 10

CH 3

CH 11

CH 4

CH 12

CH 5

CH 13

CH 6

CH 14

CH 7

CH 15

CH 8

CH 16

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

PFI0/TRIG1 Red trigger wire

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

44

45

46

40

41

42

43

47

48

49

50

34

35

36

37

38

39

28

29

30

31

32

33

22

23

24

25

26

27

16

17

18

19

20

41

10

11

12

13

14

15

4

5

6

1

2

3

7

8

9

Screw

Terminal

#

Screw

Terminal

#

94

95

96

90

91

92

93

97

98

99

100

84

85

86

87

88

89

78

79

80

81

82

83

72

73

74

75

76

77

66

67

68

69

70

71

60

61

62

63

64

65

51

52

53

54

55

56

57

58

59

Setup

Name

Channel

#

CH 59

CH 52

CH 60

CH 53

CH 61

CH 54

CH 62

CH 55

CH 63

CH 56

CH 64

CH 45

CH 38

CH 46

CH 39

CH 47

CH 40

CH 48

CH 49

CH 57

CH 50

CH 58

CH 51

CH 17

CH 25

CH 18

CH 26

CH 19

CH 27

CH 20

CH 28

CH 21

CH 29

CH 22

CH 30

CH 23

CH 31

CH 24

CH 32

CH 33

CH 41

CH 34

CH 42

CH 35

CH 43

CH 36

CH 44

NOT USED

GROUND

CH 37

B-18

EVaRT 5.0 User’s Manual Appendix B: Analog Input Hardware and Software

Table B-6. Typical Connections For 2 AMTI Forceplates and 10 EMG Channels for Midas-Based

Camera Systems (with SCB-100 and PCI-6071E A-D Card)

Screw

Terminal # Channel #

38

39

40

41

35

36

37

42

43

44

45

46

47

48

49

50

29

30

31

32

33

34

23

24

25

26

27

28

17

18

19

20

41

22

11

12

13

14

15

16

5

6

7

8

9

10

1

2

3

4

Setup

Name

GROUND

GROUND

CH 1

CH 9

CH 2

CH 10

CH 3

CH 11

CH 4

CH 12

CH 5

CH 13

CH 6

CH 14

CH 7

CH 15

CH 8

CH 16

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

PFI0/TRIG1 Red trigger wire

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

NOT USED

Forceplate Grounds

EMG Grounds

F1X

F2Z

F1Y

M2X

F1Z

M2Y

M1X

M2Z

M1Y

EMG01

M1Z

EMG02

F2X

EMG03

F2Y

EMG04

Channel

#

CH 49

CH 57

CH 50

CH 58

CH 51

CH 59

CH 52

CH 60

CH 53

CH 61

CH 54

CH 62

CH 55

CH 63

CH 56

CH 64

CH 36

CH 44

NOT USED

GROUND

CH 37

CH 45

CH 38

CH 46

CH 39

CH 47

CH 40

CH 48

CH 22

CH 30

CH 23

CH 31

CH 24

CH 32

CH 33

CH 41

CH 34

CH 42

CH 35

CH 43

CH 17

CH 25

CH 18

CH 26

CH 19

CH 27

CH 20

CH 28

CH 21

CH 29

EMG05

EMG06

EMG07

EMG08

EMG09

EMG10

Setup

Name

Screw

Terminal #

88

89

90

91

85

86

87

92

93

94

95

96

97

98

99

100

79

80

81

82

83

84

73

74

75

76

77

78

67

68

69

70

71

72

61

62

63

64

65

66

55

56

57

58

59

60

51

52

53

54

B-19

Appendix B: Analog Input Hardware and Software EVaRT 5.0 User’s Manual

B-20

Appendix C

Marker Sets

Topic

Overview

Animation

Biomechanics

Examples

Developing Optimum Markers Sets

Page

C-1

C-1

C-3

C-4

C-5

Overview

When deciding how to place markers for

EVaRT,

it is important to realize that asymmetry is used by the software to distinguish left from right on the subject. Therefore, thigh markers may not be placed symmetrically, left to right, and a single marker might be placed on one shoulder to distinguish left from right.

Also, asymmetry is used to distinguish 3 markers linked together in a triangle. Therefore, the hand and thumb marker should not be the same distance from the wrist marker and should be well separated.

Another limit is the actual number of markers used. For a very detailed skeleton, you may be tempted to use a large number of markers. However, since each marker requires computation time, there is a practical limit to the number of markers used before the speed of real-time tracking is impacted.

Specific examples of marker sets suited for both animation and biomechanics are given in the following figures.

Animation

Figure C-1

is an example of a typical marker set using 35 markers. This example also shows suggested naming conventions. However, naming conventions that best suit your needs should be used.

Note:

Biceps and thigh markers are intentionally placed asymmetrically to help the template distinguish left and right more easily.

C-1

Appendix C: Marker Sets EVaRT 5.0 User’s Manual

Figure C-1. Typical Animation Marker Set

4

1

2

11 9

7

6

19

20

12

13

15 17

Note-When placing markers on end segments, the markers should not form a line and should not have mirror symmetry. Thus, thumb and hand markers should never be the same distance from the wrist marker and should be well separated.

28

29

31

30

34

35

3

16

15

14

13

12

22

18

21

23 24

6

7

8

9 10

27

25

26

37

36

37

5

40

41

31

1 - TopHead

2 - LFrontHead

3 - LRearHead

4 - RFrontHead

5 - RRearHead

6 - RShoulder

7 - RBicep

8 - RElbow

9 - RWrist

10 - RPinky

11 - RThumb

12 - LShoulder

13 - LBicep

14 - LElbow

15 - LWrist

16 - LPinky

17 - LThumb

18 - TopSpine

19 - RFrontShoulder

20 - LFrontShoulder

21 - MidBack

22 - LShoulderOffset

23 - LowBack

24 - RRootOffset

25 - Root

26 - RRearHip

27 - LRearHip

28 - RFrontHip

29 - LFrontHip

30 - RThigh

31 - RKnee

32 - RAnkle

33 - RHeel

34 - RMidFoot

35 - RToe

36 - LThigh

37 - LKnee

38 - LAnkle

39 - LHeel

40 - LMidFoot

41 - LToe

40

38

39

33

32

34

C-2

EVaRT 5.0 User’s Manual Appendix C: Marker Sets

Biomechanics

Within EVaRT

When using

EVaRT

in biomechanics applications such as

OrthoTrak

, the standard

Helen Hayes

marker set must be modified by adding one additional marker to either the left or right scapula. Also, new linkages must be added. This will give the asymmetry required so that the dynamic template can distinguish left from right.

In addition, the order of markers is important in real-time since the order in the list determines how quickly the software can establish marker identity using the dynamic template. In general, the marker list should start with the topmost marker. Proceed down the trunk of the figure, and then down each extremity.

For example, if head markers are used, they should be at the top of the list. If no head markers are used, the shoulder and pelvis markers should be at the top of the list.

The recommended procedure is as follows:

1.

2.

Launch

EVaRT

and click on the

Connect

button.

Create a template with the modified marker set and save the results in the project file. For building a template, refer to “Building a Template” on page 9-5 .

You now have a template that can be used to automatically identify markers in real-time with this subject. When you click on the

Run

button and the subject enters the capture volume and all markers are visible, the linkages to the markers will appear automatically, indicating that the markers are properly identified.

C-3

Appendix C: Marker Sets EVaRT 5.0 User’s Manual

Figure C-2. Helen Hayes Marker Set Marker Placement

Top.Head

Rear.Head

Front.Head

Offset

V.Sacral

L.Heel

L.Shoulder

R.Shoulder

L.Toe

R.Heel

L.Elbow

R.Elbow

R.Asis

R.Wrist

L.Asis

L.Wrist

R.Thigh

R.Knee

R.Knee.Medial

R.Shank

L.Thigh

L.Knee.Medial

L.Shank

R.Ankle

R.Toe

L.Heel

L.Toe

L.Ankle.Medial

R.Ankle.Medial

Examples

Refer to the sample project folders in the

C:\ProgramFiles\MotionAnalysis\EVaRT50\Samples

directory, which includes complete marker sets.

C-4

EVaRT 5.0 User’s Manual Appendix C: Marker Sets

Developing Optimum Markers Sets

It has always been very tempting to anyone in the world of motion capture to get one “optimum” marker set. But typically, even the best animators and researchers use flexible marker setups, altering the marker sets to fit their desired capture goals.

If you are set on developing an optimal marker set, there are several things to take into consideration when you are trying to develop this.

1.

2.

3.

4.

5.

What kind of movements are you doing?

Is there going to be a lot of bending at the waist? If so, then front hip markers are probably not good choice to use.

Are you going to be doing a lot of movements like rolling on the floor or sitting in a chair or laying down? If so, consider how you have markers on the back, since they will be obscured a lot.

Are you crouching a lot? If so, markers on the front of the body

(chest) might be a bad idea.

Camera placement: Are you using a single camera placement scenario

(that is will you move your cameras around)?

If yes, then you have more flexibility with a single, “optimal” marker set.

If no, then you have to take the issues listed in (1) above into consideration. Especially in regards to the movements where the markers are blocked by your subject’s body.

Optimal number of markers: A general rule is that if you want a full

6-DOF set of information for each segment, you must have a minimum of 3 markers per segment. Currently we can shortcut that by allowing markers to be shared across joints (like the knees, ankles and elbows). Also, consider the Join Virtual definitions, to get good quality Join Virtual definitions, you want to have enough markers on the segment so that you can reconstruct missing markers using markers on that same segment. Typically in this case 4 markers per segment is advantageous.An example would be to place markers on the upper arm in the following positions: Shoulder, bicep, tricep and elbow.

Landmarking: Markers should be positioned, when possible, on bony landmarks. A bony landmark is an area like your ankle malleolus, elbow, knee condyles, wrist bones etc. This avoid the undo influence of soft tissue movements which can lead to noise in the marker positions.

Between people, the markers don't have to be in exactly in the same position. But a close approximation to the different sizes of people is required, especially if you are planning on using the PoseID-autofit option.

If you don't care about using the PoseID-autofit option, then you can place the markers anyway you like.

C-5

Appendix C: Marker Sets EVaRT 5.0 User’s Manual

C-6

Appendix D

Capturing Facial Motion

Topic

Overview

System Configuration and Setup

Marker Placement

Building a Face Template

Examples

Facial Animation Techniques for Motion Capture

Page

D-1

D-1

D-4

D-5

D-7

D-9

Overview

The

EVaRT

system can also be used to capture the fine nuances of human facial motions. Three to six cameras, positioned up to 30 degrees apart around a relatively stationary subject, will provide sufficient coverage.

The motion of 4 mm reflective markers, strategically placed about the subject’s face, is captured and 3D translation data provides manipulation to an animated character’s face model.

Animation programs like

SoftImage

,

Maya (Alias)

,

Motionbuilder

, and

3D Studio Max

will currently accept this data.

S ystem Configuration and Setup

In order to use the

EVaRT

system for facial motion capture, some additional equipment is required. This is known as the Facial Motion Capture

Accessory Kit, and it contains a Calibration Square, a facial marker set with glue, a mirror, and tweezers. The Facial Calibration Square is included with this User’s Manual.

The longer focal length lenses allow positioning of the cameras an appropriate distance from the subject, resulting in ring light illumination that is evenly distributed across the field of view. The cameras should be positioned on a subject wearing the reflective markers.

The goals of camera placement are:

Have at least 3 cameras see as many markers as possible. When 3 or more cameras see a marker, the chance of a ghost marker occurring is minimal.

Minimize the merging of markers and marker dropout in the camera views. Both are undesirable.

D-1

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

Maximize and balance angular displacement between cameras by having at least 30 degrees of angular displacement between the cameras. The exception is the lower camera (see

Figure D-1

).

Optimize what each camera sees by ensuring that each field-ofview is filled with as many markers as possible.

1.

2.

3.

4.

5.

6.

7.

8.

9.

Start by setting up the cameras as shown in

Figure D-2 on page D-4

.

Have the subject sit comfortably on a stool or chair facing the camera array.

Optimize the subject-to-camera distance by ensuring that markers fill the field of view, but are not outside the field of view.

Have the subject open their mouth wide, and make sure the head and chin markers stay in view. Look for potential merging between markers. This can especially be a problem around the lips.

With the camera positions optimized, place tape on the floor marking where the legs of the stool or chair are located.

Attach the Facial Calibration Square to the light stand and position it next to the subject. With the subject still sitting, adjust the height of the light stand until the square is positioned at the same height as the subject’s head.

Remove the subject and stool from the capture zone and position the

Calibration Square within this zone. The square is now located where the subject’s head was located.

Adjust the square’s position so each camera sees as many calibration markers as possible. Remember that the subjects’s face will be within this calibrated space during motion capture. Mark the floor with tape where the feet of the light stand are positioned. This will facilitate quick recalibration if it becomes necessary.

As an alternative, you may put the Facial Calibration Square against a wall, calibrate, and then capture as long as the subjects face is within

1 foot of the Facial Calibration Square.

D-2

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

Figure D-1. Three and Four Camera Facial Motion Capture Setup

Note—This setup is given as a minimum for the required setup.

2

1

30° 30°

3

Overhead View

Head

1

*4-Camera Facial Motion Capture Setup

2

Upper Camera

(aiming downward)

3

Lower Camera

(aiming upward)

4

30°

20°

30°

*See Figure D-2 for another look at this setup.

Head

D-3

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

Marker Placement

The number and placement of markers for facial motion capture is dependent on the animation character’s face model, and the animation software used to apply motion to the model. If a human face is to be animated, the markers should be placed at the major motion points on the face. If the face of a non-human character is to be animated, markers will be placed where the facial characteristics unique to that character will be accentuated.

In most cases, general areas of the face will need to be marked and captured. The following are suggested marker placements for facial motion capture:

Head

Three markers are used to identify head movements. If possible, the markers should be placed on areas with little or no skin movement. A tight fitting skull-cap may be used for attaching markers to the head. One marker should be placed on top of the head and one on each side of the head. These three markers are used to calculate the center of the head, which is the point from where all other marker translations are calculated.

Figure D-2. Marker Set for Facial Motion Capture

Eyebrows

One to three markers per eyebrow are used to track eyebrow movements.

The exact position of markers on or around the eyebrows depends upon the subject’s face.

D-4

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

Nose Bridge

Eyelids

Nose

Cheeks

Lips

Chin

Jaw

Place one marker between the eyes, on the upper bridge of the nose. This area tends to be a junction point between the different regions of the face.

Both the top and bottom eyelids may be marked; however, you can expect some optical interference from the eyelashes, which can add more time to tracking and editing. Also, if the bottom eyelids are marked, these markers should be offset from the position of top eyelid markers to minimize marker merging.

The nose has relatively little motion except for the nostril. If nostril flaring is of interest, attach a marker to each nostril.

At least one marker should be placed on each cheek. Exact location will depend on the animation character model and the facial features of the subject.

The lips usually have the greatest amount of movement on the face. From

4 to 9 markers can be used to capture lip movement. Markers on the top lip should be offset from markers on the bottom lip to minimize merging.

Also, areas around the lips can be marked to provide motion transitions.

One to three markers can be attached to capture chin motion.

Attach one or more markers along the jawbone for jaw motion. This is very important for lip syncing.

Building a Face Template

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

Choose

Motion Capture

from the Mode Buttons.

Choose

Output

from the sub-panel buttons.

Check the Tracked Binary (TRB or TRC) check box on the Output sub-panel.

Type a file name in the name box and press

Enter

.

Set the Duration (seconds) to

10

.

Collect motion data of the subject by having the person stand in the middle of the capture volume.

Click

Record

on the Output sub-panel.

The subject must stay in an initial frozen position for three to five seconds.

After staying frozen in this initial position for up to five seconds, the person must move through a complete range of facial motion that exhibits the full extent of stretch that will be experienced during subsequent motion capture sessions. Exaggerated motion must be avoided and all markers should remain in full view. This step should not require more than fifteen seconds.

After fifteen seconds passes from the moment

Record

was clicked, the system will automatically stop collecting and tracking marker data.

D-5

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

At this point, a Tracked Binary (TRB or TRC) file has been generated in the current directory and is ready for editing. Next, the markers must be hand identified according to the marker list built for the subject’s marker set.

11.

12.

13.

14.

15.

16.

17.

18.

Choose

Post Process

from the Mode Buttons.

Click

Quick ID

and identify the unnamed markers according to the conventions described in Appendix C, Marker Sets .

Click

Rectify

.

Manually cleanup and identify all tracks in this range of the motion file.

Click

Create Template

.

Select

Face Template

.

Select the appropriate Frames Range:

Current

—the current displayed frame

Selected

—frames highlighted in blue, low to high in dashboard

Visible

—what is displayed across the screen, as a function of the time zoom

All Frames

—all frames

Click

Create Template

.

Note:

You must use the Objects sub-panel to select the marker sets.

Face Template

Considerations

1.

2.

3.

Face templates link all markers invisibly to other markers for the

Template ID and Streaming ID functions. With all markers linked to other markers, the Template ID works much more quickly.

Explicit linkages in the Marker Set are used for the Streaming (Real

Time) Rectify and Post-Processing Rectify along with the two Linkage Stretch parameters. Generally the face template with many linkages works better than one with fewer linkages for Rectify to work.

Use only relatively rigid links and asymmetric markers if possible.

Keep the Linkage Stretch Parameters in the

Motion Capture > Tracking

sub-panel at 7 and 5 (or close) to prevent mis-IDs and allow high enough performance.

D-6

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

Examples

A Face Close-Up

Tutorial

For an example of facial motion capture data, refer to the

C:\Program

Files\Motion Analysis\EVaRT50\Samples\Dave Face Stabilization

directory and open up the project file

FaceOnly.prj

.

With this project, you can review how face marker data is tracked with body marker data. The

FaceOnly.prj

project file defines a 17 marker face capture template.

To play the motion capture data, load the VC file

DaveFaceCloseup1.vc

.

This was a capture done with a close-up view of the face taken by the video camera.

To see the video data:

1.

2.

3.

Open another window and select a Data View type of "Full Color

Video".

Right-click in the video window to bring up the AVI Frame Offset input dialog.

Set the value to –48.

This properly aligns the video data with the motion capture data.

This directory also contains a set of example files demonstrating how to use the marker stabilization tool in

EVaRT

. The file

Head.prj

defines a marker set of just the head markers for the performer (these markers are separate from the face markers). These head markers were tracked and exported to a TRC file—

Head.trc

. This TRC file was then used with Calcium and a single segment skeleton was created with the only segment being called "Head". This creates a 6DOF segment which exactly tracks the motion of the head of the performer.

Figure 15-5. “Head” Segment and its Driving Markers

D-7

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

The project file is called

Head.prj

and it contains the Calcium setup information.

In

EVaRT

, the

Head.prj

file is loaded and the skeleton generating tools are turned on. In addition, the selection of the Streaming Option

Make object A relative to segment named "Head" of Main object

is turned on.

For each TRC file that you capture, you calculate then export the HTR skeleton to an HTR file. The TRC file and HTR file are used by a standalone command line program called "Stabilizer" (from the Mocap Toolkit) to generate a stabilized TRC file from the original TRC file. The sta-

bilizer command would look as shown in Figure 15-6

.

Figure 15-6. Stabilizer Command Dialog

This indicates that the

DaveFaceCloseup1.trc

file is to be stabilized by the

Head.htr

file using the segment named "Head" and the output is to be

StabilizedHead.trc

.

The face markers will be repositioned such that the motion of the head segment is removed. This effectively places the face markers at the origin of the data space, as shown in

Figure 15-7 .

Figure 15-7. Face Markers at Origin of the Data Space

D-8

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

You can view this TRC file by loading the

FaceOnly.prj

file into

EVaRT

and loading the tracks file

StabilizedHead.trc

.

This data is now ready for use in a facial animation system.

Facial Animation Techniques for Motion Capture

Types of Facial

Animation

The goal of any facial animation technique is to move the geometry of the face around in a meaningful way. The way the mesh is modified must look very convincing to the eye since people are very attuned to facial motion and any anomaly will be quickly picked up. The two basic types of mesh modification used for facial animation are morphing and direct mesh deformation.

Mesh Deformation

Mesh deformation is a direct manipulation of the facial mesh. Markers are placed on the mesh and connected, such that, as the marker moves around so does the mesh. Each marker is given an area of influence on the mesh

(areas of influence may overlap) that fades away the farther away the mesh is from the marker. Within any particular animation system, this technique is often identically the same as what is used to do full body skinning. This is not to be confused with clustering, where groups of markers are lumped together under one control handle (the cluster). For example, as the handle moves around so do the markers as a single group.

Clustering is frequently used in facial animation but usually as a way of creating faces used for morphing.

Figure D-3. Face Model in Base Position with a Set of Markers

D-9

Appendix D: Capturing Facial Motion

Figure D-4. Face with Motion Capture Data Applied to the Markers

EVaRT 5.0 User’s Manual

Morphing

As the markers move around they pull the mesh with it. Care has to be taken on the areas of influence, especially around the mouth, so that markers affect only what they should affect. The upper lip markers, for example should not influence the lower lip.

Mesh morphing is by far the most commonly used facial animation technique. It is an extremely powerful and easy to use technique. Like with the mesh deformation technique an animator starts with a base face. Copies of the base face are made, each copy is modified into a different facial expression (open mouth, smile, eye blink, etc.). To create a blend of these expressions, the animator specifies how much of each expression is used to compose the result. For example, 50% open mouth, 10% smile, 100% eye blink. Note that the percentages do not have to add up to 100. The animator only specifies how much of each expression goes into the face.

Often times the expressions are called sub-expressions, morph targets, or simply targets.

D-10

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

Figure D-5. Base Face with the Eye-blink Expression to the Left

This is the same face as used in the mesh deformation example. This demonstrates how it is possible in some animation systems to combine the techniques for even more powerful results.

A careful inspection of the base head image with the markers should reveal that there is only one eye-lid marker. It is on the right eye of the character. This particular facial motion capture data set had only one eye-lid marker so it would not be possible to use the mesh deformation technique to animate the eye-blinks of both eyes. Only the right eye could be used.

However, using the morphing technique, the up-down motion of the right eye-lid marker can be used to control the contribution of the eye-blink morph target.

Figure D-6. Result of Combining Techniques

The right eye-lid marker has moved down a small amount to indicate that the actor blinked. Using the morph target on the left, both eyes appear closed when the target expression is applied to the final result on the right.

Unlike the mesh deformation technique, the morph target technique uses multiple meshes to do the facial animation. One requirement for all the morph target meshes is that they have exactly the same topology (the exact same number of vertices and polygons with exactly the same con-

D-11

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

Using Motion

Capture Data with Facial

Animation

Facial Retargeting with Offsets for

Mesh Deformation

Gesture

Recognition for

Morph Targets

nections between them). Therefore, it is the usual practice to model the base face first, then make copies of it for modifying into other expressions.

This section describes how motion capture data is used with each of the facial animation techniques. The direct mesh deformation technique is extremely well suited to using motion capture data while morphing is very badly suited to using motion capture data. By themselves, each technique has its advantages and disadvantages. The correct answer for an animation system lies in the ability to use both simultaneously which allows the animator to have the best of both worlds. Each is described in the following subsections.

The marker placement on the actor rarely coincides exactly with a corresponding marker placement on the character. A character's face is almost always exaggerated in some fashion that makes it impossible to find an actor to exactly match it. However, without careful placement of the markers on the mesh of the character, the deformation of the mesh simply will not work. The solution which satisfies both of these problems is to use the motion of the actor as an offset from the base position of the character.

To use this solution, you need to create a marker set on the mesh of the character that has the same number of markers with the same names as the marker set of motion capture data from the animator. The only difference between the two sets of markers is their starting positions in the base

(neutral) pose of the faces. Rather than using the absolute position of the makers from the actor, you calculate the offset of motion of a marker from its base position. That is, how far it has moved from its starting point.

This offset is what you apply to the character's markers. This way it doesn't matter if the character's face is really wide or really long or otherwise oddly proportioned. The starting points of the markers will always make sense and their motion from the actor will almost always work.

You can even scale the magnitude of the offset motion to exaggerate the motion or to dampen it.

The advantage to this technique is that it is technically easy to understand and implement. The drawback is the lack of high level expression control for the animators. If there are expressions that couldn't be captured (or weren't captured) it is hard to use keyframe animation on the mesh deformation markers to create totally new expressions.

The advantages and disadvantages for using morph targets are somewhat reversed with respect to the mesh deformation technique. Morph targets are much easier to use by animators to control expressions at a high level.

Standard keyframe animation techniques work well with morph targets.

On the other hand it is extremely difficult to create a general purpose ability to use motion capture data within such a system. Some simple and useful exceptions are found for parts of the face (such as the eye-lid example used above, or perhaps the jaw) but for some parts of the face, most notably around the mouth, it is very hard.

D-12

EVaRT 5.0 User’s Manual

Other Facial

Animation

Inputs

Keyframe

Animation

Phoneme

Recognition

Appendix D: Capturing Facial Motion

Most facial expressions do not limit themselves to a single spot on the face (such as the location of a single marker). Each facial expression moves many markers at once. A smile, for example, not only moves all of the markers around the mouth but it also moves markers around the eyes, temples and forehead (most people squint when they smile). So there isn't an easy way of linking a smile morph target to one or two motion capture markers. A smile is a true smile if and only if a whole set of markers move in just the right way. Systems capable of doing this kind of analysis have been made for doing realtime facial animation but this technique has not yet found its way into most animation systems.

Currently, the most advanced facial animation systems (the ones used to make some of the popular feature length animated features) use knowledge of skeletal and muscular anatomy to understand how the underlying tissue affects the skin movement of a character. These complex animation systems are all morph targets based in the sense that the animator still works with a blend of high level expressions to achieve their final result.

They still say "I want half a smile" and "part of a smirk". The animation engine accounts for muscle and bone movement while composing the final result for the skin.

Motion capture data isn't the only kind of input used for doing facial animation. A summary of other techniques is given here.

All animation systems have keyframe animation at their core. The ability to use keyframe animation in conjunction with motion capture data is vital to getting the best overall results. It is important to use the keyframe tools without damaging the motion capture data. Some keyframe animation systems require that the motion capture data be simplified in order to control the data with keyframes. This is a mistake of large proportions as motion capture data should never be decimated. Motion capture data can't quite get everything that an animator will need from the motion so extra motion has to be layered on top. This is true of all motion capture types

(face, body, hands) and devices.

Lip-synching is an important sub-problem of facial animation. The motion of the face (particularly the lips) must be synchronized to the audio track of the voice talent. The classic technique is for the animator to listen to the audio track and keyframe animate the facial expressions to match.

The first part is to get the mouth in the right position to match the syllable.

If the animator is using a morph target technique, it is very common to have a series of morph targets, each representing a common phoneme used in speech. This makes it straightforward, but time consuming to animate.

There are a number of automatic phoneme recognition technologies available for evaluating audio input and generating phonemes. This information can be used as a source of animation data for facial animation. Not only is phoneme recognition hard to get right, but the general approach has inherent limitations. Phoneme recognition can be helpful, but will never entirely solve the problem of facial animation. A summary of the drawbacks is as follows:

D-13

Appendix D: Capturing Facial Motion EVaRT 5.0 User’s Manual

Eye Movement

Waldos

Other Motion

Capture Issues

Marker Size and

Capture Volumes

Too precise

—At any given moment the phoneme generator gives you only one phoneme (this will improve in the future). There is no notion of blending between them. All people slur their syllables to one degree or another, this kind of information is missed.

Anticipation

—Almost all sounds require a setup motion for the face.

You open your mouth before you actually say anything, the audio track doesn't have that kind of information.

Other facial movement

—When people talk, their necks, ears, and other parts of the face move around significantly. This information is not conveyed. Lack of eye blinking information is perhaps the most important.

Non-audible facial expressions

—Most people intersperse their conversations with a variety of facial expressions to punctuate the conversation.

Despite these limitations an automatic phoneme generator can provide an excellent first pass at facial animation with the intent of going back over it to augment it with more facial animation information.

Eye motion is a subtle, but a vital part of facial animation that must be present. A variety of techniques exist for obtaining eye motion, but it is difficult to get without hampering the acting talent. The most common approach is to take a video image of the eye and track the eye movement from the video footage. Some techniques track the whites of the eyes, others the pupils. Some use visible light, others use infrared.

In all cases, a 2D image is used to generate information about the translational movement of the eye in the image. This is then turned into rotation information to rotate the eyeball of the character.

Waldos are physical input devices use as puppeteering controls for characters. Each input type is given a high level meaning such as head rotation, eye-blink, or a particular facial expression. For this reason, waldos and morph target facial animation systems work well together.

A miscellaneous collection of issues which affect, or are related to, facial motion capture are detailed in the following sub-sections.

Marker size, camera resolution, and capture volume calibration are all key elements to determining how large the capture volume can be and how much of the performer can be captured at once. Camera refresh rates and the number of markers can affect this too, although they're not as important. The latest camera systems have higher resolutions and the latest software has easy to use techniques for handling lens distortion and capture volume calibration. This adds up to the ability to have smaller markers and more of them which allows for full body and face motion capture simultaneously.

Having face and body data at the same time is an important technological hurdle. It simplifies a great number of face-body coordination issues and allows for real-time processing of the data so that live motion capture sets can be created (a live set is when the data is acquired, processed, applied to a virtual character and rendered at 30 frames a second).

D-14

EVaRT 5.0 User’s Manual Appendix D: Capturing Facial Motion

Marker Stabilization

Sometimes the facial animation techniques require working with the data in a simple reference frame as though the head were an object by itself sitting on a table top. If the facial motion capture data is captured as part of a full body and face capture, then the facial markers have to be segregated and recalculated relative to the motion of the head segment of the body.

This process is known as stabilization. This is a vital tool for keeping the facial data under control.

It is particularly important to have good, solid head motion in your character skeleton if you use this technique. Any amount of slippage in the motion of the head relative to the facial markers will result in jittery, noisy face data.

Sync to Body

Capture

Motion Capture of

Hands

Ideally, the face and body are captured simultaneously so that the data is automatically synchronized because it's all part of the same data set. The good news is the latest motion capture systems allow for this. Global time information needs to be encoded in all motion files so that they can be later synchronized if they're not captured simultaneously. Time stamp information is always useful to have in any case.

The motion capture of hand motion, and the fingers to be more specific, presents many of the same issues as facial animation. The capture volume limitations are about the same since the markers are about the same size.

It is even common to use morphing techniques on the hands like on the face rather than treating the hand as a mini-skeleton and animating it like you would animate the full body. Certain hand gestures are very common so it's effective to model those few (fist, flat hand, pointing a finger) and morph between them. If motion capture data is used, however, skeleton animation techniques are easier to use and apply to the hands.

Limitations

Facial motion capture does have some limitations, not all the information that an animator might want can be captured directly from the face of the actor. Some examples are:

Tongue

—You can't put markers on the tongue. There is no effective way of capturing the full motion of the tongue with any kind of technique.

Neck

—Various regions of the neck move while talking. The motion of the tongue causes the underside of the jaw to move while swallowing and breathing causes other areas to move. Extra markers could be placed under the jaw and around the neck but then visibility issues become a concern.

Eyes

—Markers can't be used to track the eyes. Other techniques might be used for this.

Curl and other twisting motions

—The skin of the face doesn't just travel in straight lines. Many parts of the face scrunch and curl around. Pursing one's lips or pouting motions cause the lips to bend around in a variety of ways. Markers do not directly transmit this information, only careful placement of groups of markers can effectively sense this.

Number of markers

—Ideally it would be best to use as many markers as you can put on the face. Capture limitations prevent this so much information that might theoretically be measured will have to wait for higher resolution cameras and even smaller markers.

D-15

Appendix D: Capturing Facial Motion

Production

Issues

EVaRT 5.0 User’s Manual

Despite this list, facial motion capture is, by far, the best overall technique for generating facial animation from a performer. No other kind of system is as versatile or productive.

Many practical issues creep into the animation process that do not have much to do with techniques for animation but rather with the process of animation itself. That is to say, it has to do with the relationship between the animator and the production tools. This is true for all tools, especially facial animation tools. Here is a list of issues that anyone doing facial animation should keep in mind:

Requirements do change

—Real productions do not march relentlessly from front to back, animators have to go back over the data many times to get the result they need. For example, a common mistake in morph target facial animation is the realization that you need more flexibility in a certain part of the mesh so you can make a new target expression that you suddenly discovered you needed. Since all the expression meshes must exactly match, you have to update all the existing meshes to incorporate your new change. This can be tedious and error prone. Some morph target systems do automatic updates for you. Be ready for this when setting up your workflow. Consider the possibility that you might, at times, have to work backwards through problems.

Facial animation is just one element

—Facial animation would be somewhat easier if all you had to worry about is the face. The problem is that you almost always have to attach the face to a head (and therefore to a body). The relationship between the face and body elements must be considered when setting up a character for animation.

Interactive versions versus full render versions

—The high resolution final images (if that is the final output) require high resolution facial meshes. This sometimes hampers the inter activeness of the animation system. The ability to use low resolution meshes for interactive work and then replace them before the final product is a valuable production tool.

Synchronization

—All input data needs to be synchronized. Global time information needs to be present in all the data so that as the animator works with different kinds of data (face, body, voice, video), they can be matched up in time.

Output

—The final output of motion data is not always a final image.

The final output might be animation data sent to a game engine (or some other kind of interactive environment). In which case, the algorithms which underlay the facial animation need to be present in the game engine so it can reproduce the facial animation interactively.

D-16

Appendix E

Forcepla.cal File Format

Topic

General Information

Forceplate File Data

Forceplate Scaling Factor, X-Width and Y-Length

Using AMTI and Bertec Forceplates

Using Kistler Forceplates

Using Kyowa Dengyo Forceplates

Page

E-1

E-3

E-5

E-6

E-8

E-9

General Information

Up to eight forceplates may be placed within the video capture space to measure gait forces. While

EVaRT

gathers video data, it simultaneously acquires forceplate data.

To accomplish this, the forceplate output is connected to an analog input card in the

EVaRT

system. The forceplate data is interpreted using a file called

forcepla.cal

. When

EVaRT

reads in the trial data, it first searches the current directory where the project resides for the

forcepla.cal

file. If none is found, it then searches in the directory:

C:\Program Files\Motion Analysis\EVaRT5\Samples\Example Forcepla.cal Files.

The

forcepla.cal

file contains information describing the location, orientation, and calibration of each forceplate used. The exact form of the file will depend on the forceplate manufacturer.

Figure E-1 shows the file

form for Bertec and AMTI forceplates.

Figure E-2

shows the file form for

Kistler forceplates.

The

forcepla.cal

file contains no text, only numbers. For multiple forceplates, the data for each forceplate in the system is included in one

forcepla.cal

file (see

Figure E-4 ).

Note:

The

forcepla.cal

file must be in the same directory as either the

EVaRT50.exe

file or the

*.prj

file. Otherwise, the forceplate outlines will not appear in the 3D collection view.

Note:

Forcepla.cal

files in the past have been named with a “t”, as forceplate.

Be sure to check that there is no “t” in

forcepla.cal

.

Example

Forcepla.cal

files for each type of forceplate (AMTI, Bertec, and Kistler) can be found in the directory:

C:\Program Files\Motion

Analysis\EVaRT5\Samples\Example Forcepla.cal Files.

E-1

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

Figure E-1. Forcepla.cal File Structure for Bertec and AMTI Forceplates

Line# Description

1 Forceplate number (1 through 8)

2 Forceplate scaling factor and [optional length and width of forceplate]

(25 for AMTI setup with amplifier gain switches set to 4000)

6

7

8

3

4

5

6x6 forceplate calibration matrix (Inverted Sensitivity) provided by the manufacturer

9 Xo Yo Zo True XYZ origin relative to the geometric center of the forceplate—in cm (provided by the manufacturer).

10 Xc Yc Zc XYZ location of the geometric center of the plate with respect to your video coordinate system.

(the video calibration system’s origin)—measured in cm

11

12

13

3x3 forceplate orientation matrix to make the forceplate coordinate system match the laboratory coordinate system

Figure E-2. Forcepla.cal File Structure For Kistler Forceplates.

Line# Description

1 Forceplate number followed by “K” to indicate a Kistler forceplate.

2 Forceplate scaling factor and [optional length and width of forceplate]

7

8

9

10

3

4

5

6

8x8 forceplate calibration matrix created by the user

11 Xo Yo Zo position of the forceplate transducers in cm

(provided by the manufacturer)

12 Xc Yc Zc XYZ location of the geometric center of the plate with respect to your video coordinate system.

(the video calibration system’s origin)—measured in cm

13

14

15

3x3 forceplate orientation matrix to make the forceplate coordinate system match the laboratory coordinate system

E-2

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

Forceplate File Data

Forceplate Number

A unique number is assigned to each forceplate in the system.

Forceplate Scaling

Factor and Optional

Length & Width

The scaling factor depends on the forceplate manufacturer and the gain setting. Length and width are optional and are measured in cm. Length and width orientation is also dependent on the forceplate manufacturer.

Refer to later sections specific to the manufacturer of your forceplate.

Forceplate

Calibration Matrix

True XYZ Origin

XYZ Location in

Video Coordinate

System

The calibration matrix transforms the output of the forceplate into forces and moments. Refer to the section later in this appendix specific to the manufacturer of your forceplate.

This is the offset of the origin of the forceplate XYZ coordinate system relative to the center of the top surface of the forceplate. Each manufacturer provides this offset value.

This tells the

EVaRT

system where the center of the top surface of the forceplate is located relative to the

EVaRT

video coordinate system. Once this is established, the video calibration frame must be placed in the same location each time you calibrate. The center of the top surface can be found by measurement or drawing diagonal lines from opposite corners.

The units of measurement are centimeters (cm).

3x3 Orientation

Matrix

This matrix describes the orientation of the forceplate relative to the laboratory or room coordinate system. It is a matrix of direction cosines of the angles between the forceplate coordinate system and the laboratory coordinate system. Using the terminology cos(X plate

, X lab

) to indicate the angle between the forceplate X axis and the laboratory X axis, the matrix takes the following form:

X

plate

Y

plate

Z

plate

X

lab cos

(

X

plate,

X

lab) cos

(

Y

plate,

X

lab)

Y

lab cos

(

X

plate,

Y

lab) cos

(

Y

plate,

Y

lab)

Z

lab cos

(

X

plate,

Z

lab) cos

(

Y

plate,

Z

lab) cos

(

Z

plate,

X

lab) cos

(

Z

plate,

Y

lab) cos

(

Z

plate,

Z

lab)

Since, in real situations, the forceplate should be aligned with the room coordinate system, the numbers in this matrix will always have one of three values:

Angle = 0

Angle = 90

Angle = 180 cos = 1 cos = 0 cos = –1

Example matrices are shown in the following figure:

E-3

Appendix E: Forcepla.cal File Format

Figure E-3. Forceplate Coordinates System

EVaRT 5.0 User’s Manual

E-4

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

Forceplate Scaling Factor, X-Width and Y-Length

The scaling factor depends on the forceplate manufacturer, and the forceplate amplifier gain setting and the voltage range.

Table E-1. Sample Forceplate Scaling Factors

Forceplate Manufacturer

AMTI

Bertec

Kistler

Scaling Factor

For Gain 4000—Use 25.0

Use 1.0

±10 Volt Amplifier—Use 1.0

±5 Volt Amplifier—Use 0.5

The x-width and y-length are the forceplate measurements in centimeters as measured in the forceplate coordinate system. Check the manufacturer’s specifications. If no x-width and y-length values are used, AMTI and Bertec forceplates default to 18-inches by 20-inches, and Kistler forceplates default to 50-centimeters by 50-centimeters.

E-5

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

Using AMTI and Bertec Forceplates

AMTI Gain Setting

For the AMTI forceplates, a gain of 4000 mV and a cutoff frequency of

1050 kHz is recommended. Using the method outlined in the AMTI literature, this gain yields a scaling factor for the

forcepla.cal

file of 25. The example in

Figure E-4 uses an AMTI forceplate.

Bertec Gain Setting

A gain setting of 10 for Bertec forceplates is recommended. In the

forcepla.cal

file you should set

scaling factor = 1 / gain

, yielding a

scaling factor

of 0.1.

The Calibration

Matrix

The 6x6 calibration matrix (Inverted Sensitivity) is provided by the manufacturer (AMTI or Bertec). It is used to transform the output of the forceplate into three force vectors and three moment vectors. The form of the

matrix is shown in Figure E-5 . The main diagonal of the matrix (upper

left to lower right) represents the basic channel sensitivities. The off diagonal terms represent the channel cross-talk.

Figure E-4. Example Forcepla.cal File for 2 AMTI Forceplates

1.

25 51 46.5

2.9350

0.0040

–0.0020

2.9930

–0.0270

0.0120

0.0130

0.0470

11.5420

–0.0240

0.0030

0.0000

–0.0070

0.0000

–0.0070

–0.0020

0.0000

–0.0480

–0.0400

0.0050

0.0080

1.5390

0.0000

–0.0250

0.0410

–0.0110

1.5350

0.0160

0.0010

–0.0040

–0.0040

0.7440

0.0020

0.0020

–0.0050

0.0020

–0.1000

–0.0260

–3.8000

5.6000

0.0000

–25.7000 –4.2000

1.0000

0.0000

1.0000

0.0000

2.

25 51 46.5

0.0000

0.0000

2.9340

0.0120

0.0050

–0.0040

0.0010

0.0020

0.0090

2.9750

0.0020

0.0010

–0.0020

0.0060

0.0000

–1.0000

0.0020

0.0450

–0.0130

0.0170

0.0310

–0.0200

0.0110

0.0340

11.5480

–0.0210

–0.0090

0.0070

0.0000

1.5470

–0.0020

–0.0030

0.0000

0.0000

–0.1000

0.0300

56.5

–4.2000

–25.7000 –4.2000

0.0020

0.0010

1.5450

–0.0080

–0.0060

0.7480

0.0000

–1.0000

0.0000

–1.0000

0.0000

0.0000

0.0000

0.0000

–1.0000

E-6

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

Figure E-5. The 6x6 Calibration Matrix forces moments a11 a12 a13 a14 a15 a16 a21 a22 a23 a24 a25 a26 a31 a32 a33 a34 a35 a36 a41 a42 a43 a44 a45 a46 a51 a52 a53 a54 a55 a56 a61 a62 a63 a64 a65 a66

The main diagonal represents the basic sensitivities for each channel:

Force units

The information provided by the manufacturer may include only the basic sensitivities for each channel with no values for cross-talk. In this case, the matrix should be filled with the basic sensitivities on the main diagonal and zeroes elsewhere.

Also notice that the upper right quadrant of the matrix contains the force sensitivities and the lower right contains the moment sensitivities. In every case, the force sensitivities are greater than the moment sensitivities.

EVaRT

uses this information to switch matrix quadrants (permute the matrix) if the manufacturer should supply the matrix with the moments on the left and force on the right.

Note:

The calibration matrix is intended to be used with your plate’s coordinate system, not the room’s. For this reason, if your plate is not aligned with the room, correct it with the 3x3 orientation matrix, not by switching wires or A/D signal names.

E-7

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

Using Kistler Forceplates

Signal Names

The Kistler forceplate has 8 outputs. Therefore, two forceplates will use

16 channels on the A/D card. The signal naming conventions are shown in

Figure E-6 . The names in the analog ANB (or ANA) file must appear

exactly as shown in the

EVaRT

ANB (or ANA) column.

Gain Setting

With the Kistler forceplate, the Charge Amplifier (model 9865) should be set on

range #3

for the X/Y and Z range settings (X and Y are set together). This is the 10,000 pC setting.

This setting can be changed if desired, but the

forcepla.cal

file will have to reflect the change. A

gain = 1

on the A/D board should be used since the Kistler outputs 10 V full scale.

Calibration Matrix

The Kistler forceplate requires an 8x8 calibration matrix. The matrix only contains non-zero data on the main diagonal (upper left to bottom right).

All non-diagonal cross-talk elements are zero.

To calculate the values to use on the main diagonal of the matrix (assuming nominal sensitivity values of 7.8 and 3.8 pC/N):

X and Y Scaling (10000 pC / 7.8 pC/N) / 10 V = 128.2 N/V

Z Scaling (10000 pC / 3.8 pC/N) / 10 V = 263.4 N/V

Figure E-6 shows an example 8x8 matrix in a

forcepla.cal

file.

True XYZ Origin

This is a measure of the X, Y, and Z distances to the piezoelectric transducers used to generate the signals in the Kistler forceplates. These numbers are supplied by the manufacturer.

Figure E-6. Example Forcepla.cal File For a Kistler Forceplate

0

0

0

0

1K

1.0

128.2

0

0

0

12.000

0

0

0.0

1

0

128.2

0

0

0

0

0

0

20.000

0.0

0

–1

0

0

0

128.2

0

0

0

0

0

–5.4000

0.0

0

0

–1

0

0

0

128.2

0

0

0

0

0

0

0

0

263.4

0

0

0

0

0

0

0

0

263.4

0

0

0

0

0

0

0

0

263.4

0

0

0

0

0

0

0

0

263.4

E-8

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

General Notes On

Kistler Forceplates

Note:

1.

2.

Since the Kistler forceplate format is flagged with a “K” after the forceplate number, Kistler and other forceplates may be included in a single system.

The proper way to orient the forceplate is the 3x3 orientation matrix, not the calibration matrix.

Do not switch the cables to the A/D board.

3.

4.

Keep the Long Term Constant turned off on the charge amplifier.

Reset the charge amplifier before each test, or at least every few tests.

This re-establishes the zero for the charge amplifier.

Using Kyowa Dengyo Forceplates

The Kyowa Dengyo force plate has now been incorporated into

EVaRT

.

The following is a description of the procedure for calibration using the

RealTime interface as well as a description of the

forcepla.cal

for Kyowa

Dengyo force plates.

The automatic calibration of the Kyowa Dengyo force plates is now implemented in

EVaRT

. At the end of the calibration procedure the Real

Time system creates a new

forcepla.cal

file containing the latest calibration values (zero, +cal and –cal) for computing the distortion conversion coefficient. The calibration procedure is as follows:

1.

Make sure that the Kyowa Dengyo force plates are connected to the

National Instruments A/D data acquisition hardware with the following channel assignments:

Table E-2. Force plate channels for Kyowa Force Plate number 1

Analog Channels

Channel 1

Channel 2

Channel 3

Channel 4

Channel 5

Channel 6

Channel 7

Channel 8

Kyowa Dengyo Force Plate

Channels

FZ11

FZ12

FZ13

FZ14

FX114

FX123

FY112

FY134

Follow a similar connection and naming (FZ21, FZ22 etc.) sequence for additional plates.

E-9

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

2.

Connect the ST-100 box terminal pins 25, 27, and 29 (pin 24 is ground) which correspond to bits 0, 1 and 2 of the digital I/0 of the

National Instrument A/D board to Cal (o), Cal (+) and Cal (-) of the

Kyowa Dengyo interface box.

Figure E-7. Kyowa Connection Block Diagram

Pin 24

Ground

Pin 25-bit 0

Pin 27 –bit 1

Cal 0

Cal +

Cal -

Pin 29 – bit 2

E-10

ST- 100 BOX

KYOWA

CONTROL BOX

3.

The Folder containing the

EVaRT

executables must also have a

forcepla.cal

file (previously created) according to the format for

forcepla.cal

file for Kyowa Force Plates.

Load a project file that has the Kyowa forceplates enabled in the Analog set up panel

There is a new button named

Calibrate Kyowa Force Plates

in the

Setup > Misc

sub-panel.

After making sure all the connections are properly made click on the

Calibrate Kyowa Force Plates

button.

The following sequence of events are initiated.

TTL pulses are sent, in sequence to CAL 0, CAL+, CAL- and CAL

ZERO terminals of the Kyowa control box as shown on the timing diagram.

100 samples of analog data are collected across all the Kyowa force plate channels, 5 seconds after the initiation of the TTL pulses.

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

Approximately 1 minute is required for the calibration.

The average of the 100 samples (in A/D units) are computed and data are written into a

forcepla.cal file

in the same folder as the loaded project file.

The timing sequence of the TTL pulses and the analog data acquisition is shown on the next page.

Figure E-8. Timing Diagram for Kyowa Force Plate Calibration

Cal 0

Cal +

Cal -

5 s

Settling time

5 s 5 s

Periods of Analog data Acquisition

(100 Analog Samples)

E-11

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

Description of

Forcepla.cal File for Kyowa

Dengyo

Forceplates

[Line 1] [Force plate number Kyowa] Example:1Kyowa, 2Kyowa etc,.

[Line 2] [Scale factor Width Length]

[Line 3] [Calibration Range Settings in the order FZ1 FZ2 FZ3 FZ4

FX14 FX23 FY12 FY34]

[Line 4] [Zero values in A/D units written by EVaRT Calibration step

[Line 5] [Cal + values in A/D units written by EVaRT Calibration step]

[Line 6] [Cal - values in A/D units written by EvaRT Calibration step]

[Line 7 ] [Load Conv.Coeff.*Voltage Conv. Coeff.*9.801] same order as above

[Line 8] [XY axis Conversion Coefficients(XY locations of Z-axis force

Transducers) in centimeters]

[Line 9] [X location of Fy and Y location of Fx transducers in centimeters]

[Line 10] [Location of Geometric Center of Force Plate with respect to video coordinate origin in video coordinates]X Y Z in Centimeters

[Line 11]

[Line 12] [Force Plate Orientation Matrix]

[Line 13]

E-12

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

Example Kyowa

Dengyo

Forcepla.cal File

Figure E-9. Example Forcepla.cal file for 4 Kyowa Dengyo Force Plates

Force Plate

Coordinate System

(+ Z up)

Y

X

4

1

X

Video

Coordinate System

(+ Z Up)

Y

3

2

E-13

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

1Kyowa

1.0

60 180

2000

500

2000

500

2000

-3 2 -4 3 6

1600 1601 1602 1599 803

2000

-6

802

500

5

800

-6

802

500

-1608 -1609 -1608 -1608 -805 -804 -804 -805

0.00103

0.00104

0.00103

0.00103

0.00073

0.00073 -

0.00071 -0.00071

-22.5017 –22.3307

22.8603

22.1836

55.7321 –55.2022

–54.9477

54.5368

23.35

51.00

90.0

30.0

0.00

0.0000

-1.0000

0.0000

2Kyowa

1.0000

0.0000

0.0000

0.0000

0.0000

1.0000

1.0

2000

500

-8

60 180

2000

5

500

-6

2000

3 -22

1609 1606 1606 1607 803

2000

26

803

500

-8

805

-11

803

500

-1608 -1607 -1605 -1607 -802 -804 -803 -802

0.00101

0.00102

0.00101

0.00101

0.00079

-22.5648

54.8645

23.35 51.00

0.00072

-22.3288

-55.3299

-0.00074

22.3755

-55.3140

-0.00075

22.7563

55.7631

-90.0

0.0000

-1.0000

0.0000

0.00072

-22.5481

55.2674

-55.6732

23.35 51.00

30.0

1.0000

0.0000

0.0000

0.00072

-22.3083

-54.6867

55.1962

0.0

0.0000

0.0000

1.0000

3Kyowa

1.0

60

2000

180

2000

500

-21 25

500

24

2000

22 -18

1613 1606 1605 1607 805

2000

21

806

500

-10

800

15

804

500

-1610 -1605 -1601 -1606 -803 -806 -808 -804

0.00102

0.00103

0.00101

0.00104

-0.00075

22.3729

-0.00075

22.7725

E-14

EVaRT 5.0 User’s Manual Appendix E: Forcepla.cal File Format

-90.0

0.0000

-1.0000

0.0000

4Kyowa

1.0

2000

-30.0

1.0000

0.0000

0.0000

60 180

2000 2000

0.00

0.0000

0.0000

1.0000

2000 500 500

500

-0 2

500

2 4 10

1602 1607 1604 1603 801

-22.3702

55.3667

23.35 51.00

90.0

-3

805

6

803

-5

803

-1608 -1614 -1612 -1610 -805 -808 -807 -806

0.00104

0.00072

0.00104

0.00073

0.00102

-0.00071

0.00103

-0.00072

-22.5673

-55.2103

22.6288

-54.9972

22.4559

55.2328

0.0000

-1.0000

0.0000

-30.0

1.0000

0.0000

0.0000

0.00

0.0000

0.0000

1.0000

E-15

Appendix E: Forcepla.cal File Format EVaRT 5.0 User’s Manual

E-16

Appendix F

SDK—Software Developers

Kit

Topic

SDK Overview

SDK Programming Example: Write your own Streaming

Plugin

Page

F-1

F-1

SDK Overview

The SDK is available for the advanced user who wishes to incorporate the output data stream from

EVaRT

into a software application.

The Software Developers Kit, which provides the tools for interfacing your program with

EVaRT

is available by special request from Motion

Analysis Corp. at [email protected]

SDK Programming Example: Write your own Streaming

Plugin

There is a Software Development Kit (SDK) which is written in the Visual C/Visual C++ language and an example C program that is available at no charge that demonstrates how to use the SDK. This allows our customers to use this as a starting point and creating their own program.

The sample program shows you how to connect to the

EVaRT

software and request that the kind of data that you want be transferred (Marker

XYZ data and/or Skeleton HTR data). The sample C program then writes out the data to a disk file. Please also note that the data can be streamed either from the live camera data (that is happening in real time), or from the Post Processing software when you press the Play button. The data is the same either way and the SDK program does not even need to know since it comes across the same way. So, the customer can write the program from previously edited XYZ data in the Post Processing part of

EVaRT

and get that to working. Press the

Play

button, you see the edited tracks and the data is streamed to the SDK. Then they can connect to the cameras and get the same XYZ or HTR type data from the live cameras.

F-1

Appendix F: SDK—Software Developers Kit EVaRT 5.0 User’s Manual

F-2

Appendix G

Import and Export File

Formats

Topic

Overview

mac_lic.dat

PRJ—EVaRT Project File

TRC—Track Row Column

HTR2—Hierarchical Translations and Rotations

HTR

ANC—Analog ASCII Row Column

TS—Time Series Files from the EVaRT Analysis Functions:

Velocity and Acceleration Calculations

Binary Files—ANB, TRB, and C3D

Page

G-1

G-2

G-3

G-3

G-5

G-9

G-15

G-16

G-18

Overview

The files generated by

EVaRT

fall into two main categories: ASCII and binary. ASCII files contain data in a text form that can be read by any text editor. They usually have descriptive headers that indicate the nature of the data that follows. Often these files are in a row and column format that allows data to be read and manipulated by spreadsheet programs such as

Excel

TM

. ASCII files are not compact and can be quite large.

Binary files contain raw binary data and are more compact than ASCII files. They cannot be read by a text editor. In general, binary files are not meant to be read by the end user.

G-1

Appendix G: Import and Export File Formats EVaRT 5.0 User’s Manual

mac_lic.dat

All Motion Analysis software requires a valid license to run. The license is keyed to a particular computer.

For a

Windows NT

computer, it is keyed to the number of the dongle supplied by Motion Analysis.

For

SGI

computers, it is keyed to the sysinfo-s number.

For

Sun/Sparc

computers, it is keyed to the host id number.

The license file is ASCII and has the unique name

mac_lic.dat

. It is located in the

Motion Analysis

directory. Only one license file is used for all Motion Analysis software. Each additional software application beyond

EVaRT

is given a separate line in the license file, with the license type enclosed in square brackets

[ ]

followed by two license codes.

If you acquire new Motion Analysis software, you must use a text editor to enter the new line in your license file to enable your new software. This line can be typed in or entered by cutting it from the file you receive and pasting it into the existing license file. The order of the licenses in the file is not important and only those lines that start with a

[

(left bracket) are read by the software.

Note:

If you are operating in Windows 2000, make sure that file extensions are not hidden, which is the default. This makes the

mac_lic.dat

look like

mac_lic,

which might be renamed to

mac_lic.dat.dat

. If this happens, the system will not recognize the license file.

Figure G-1. An Example of a Motion Analysis License File

Motion Analysis License File

Customer: MAC Customer

Platform: NT

SystemID: 19c

Created: 9/15/2005 1:42:26 PM

Sales Order#: 05-xxx

Entered By: Support

[EVa RealTime v4.4] aed50167

[Analog Input]

[OrthoTrak] b9806c31 b2df5e69

[Animation Plugins] b1a46160

[Director/Sequencer] e1a04e65

[RT2 Animation Plugins] e3f05340

[Analog Input] b9806c31

[Calcium 4] e7ed5923

[Skeleton Builder 4] a3f44279

[Reference Video 3.0] eb92592f

[Talon Streaming 4] ecb36136

[Talon Viewer 4]

[BioFeedTrak]

86fb0714 ac943872

[Motion Composer] c7f00e25

This license has no expiration.

873b2d56 d1567841

8964274a

805b5c49

85745819 a069081b d1567841 c363151f

99780c5b cf636a13 d65b4b14 f43d037e

92026c54 c534083f

G-2

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

PRJ—EVaRT Project File

Every motion capture session must have a project file containing all system settings, equipment parameters, and other information related to the project. This file contains both equipment parameters common to many different setups and calibration values unique to one particular session.

Among the data found in a project file are:

• the system setup

• the marker set

• calibration setup and results

• linkages between markers

SkB segment definitions, coordinate systems, and hierarchies

(optional) MoCap Solver segment definitions, joint types, and hierarchies

• camera type and parameters

In most cases, you will begin a session by loading an existing project file, editing it as necessary, and saving it in the directory where the motion data is to be saved. Any time you calibrate the system or edit project parameters, you should save the project file to disk to retain the new information.

Important

Project files contain ASCII data and it may be useful to view them using any text editor, however, you should never edit them in a text editor.

TRC—Track Row Column

The

.trc

file contains X-Y-Z position data for the reflective markers. This is an ASCII file in a Row/Column, horizontal tab delimited format that can be easily read into a spreadsheet program such as

Excel

TM

and

Lotus

TM

. The position data for each marker is organized into 3 columns per marker (X, Y and Z position) with each row being a new frame. The position data is relative to the global coordinate system of the capture volume and the position values are in the units used for calibration.

The file is made up of three parts:

• the file header,

• the position data header, and

• the position data.

All fields in this file type are separated by horizontal tabs.

File Header

The

.trc

file header occurs on the first three rows.

Row one contains the path file type label (string), path file type number (int), path file type descriptor (string) and original directory path and file name (string).

G-3

Appendix G: Import and Export File Formats

Data Header

Position Data

Empty Fields

Example

EVaRT 5.0 User’s Manual

Row two contains the data rate label (string), the camera rate label (string), the number of frames label (string) and the units label (string).

Row three contains the data rate value (real), the camera rate value (real), the number of frames (int) and data units (string).

The data header occupies rows four and five.

Row four contains the frame number label (string), the time label

(string) and followed by the marker name labels (string). There are three horizontal tab characters between each marker name label. These names usually correspond to the location where a reflective marker was placed on the subject.

Row five contains the column labels (string) for the position data starting on row six. For each marker name there is an X, Y and Z column. These axes labels have the trajectory numbers appended to them.

Position data begins at row six. Column one of the data fields contain the frame number (int). Column two contains the time (real) and columns from three on contain the X, Y and Z position data (real) for each trajectory. There are three columns for every trajectory.

An empty frame of position data (missing data) is represented as three consecutive horizontal tab characters.

Shown below is a portion of a file with the following attributes:

• captured rate = 60 frames per second

• total frames = 90

• total reflective markers = 33

• units of measure = mm

Data for only the first 3 markers is shown, the remaining markers would appear in columns to the right. Also, data for only the first 12 frames is shown, the remainder would appear in rows below frame 12.

G-4

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

Figure G-2. An Example of a TRC File

6

7

4

5

2

3

PathFileType4 (X/Y/Z)

DataRate CameraRate

/usr/people/evademo/Oct14/MichelleInit1.trc

NumFrames NumMarkers UnitsOrigDataRate mm 60.0

.60.0

60.0

Frame#Time Head_Top

1

90 33

LHead

X2

RHead

X3 X1 Y1 Z1 Y2 Z2 Y3 Z3

0.000

234.5437

1673.7619

232.2308

316.7533

1608.3785

218.5500

144.7963

1597.3691

274.4994

0.017

0.033

235.2399

235.2361

1673.4542

1673.4926

232.1284

232.1852

316.6074

316.1265

1608.2884

1608.1984

219.0597

219.7480

144.7684

144.7296

1597.8106

1597.8043

274.2137

274.3916

0.050

0.067

0.083

0.100

235.0781

235.0781

235.0781

235.3844

1673.4376

1673.4376

1673.4376

1673.4179

232.2152

232.2152

232.2152

232.3085

316.5659

316.7533

316.2800

316.4539

1607.9659

1608.3785

1608.3460

1608.1409

219.6344

218.5500

219.1676

219.6402

144.7652

144.7296

144.7963

144.7963

1597.5711

1597.8043

1597.3691

1597.3691

274.4420

274.3916

274.4994

274.4994

8

9

10

11

0.117

0.133

0.150

0.167

235.1416

235.4312

236.0334

235.7562

1673.4447

1673.4882

1673.4269

1673.7265

232.4143

232.7111

233.0335

233.4352

316.4539

316.1265

315.9777

316.4172

1608.1409

1608.1984

1608.3370

1608.1043

219.6402

219.7480

220.0773

219.9638

145.1222

144.7963

144.6867

144.7963

1597.3621

1597.3691

1597.5786

1597.3691

274.3569

274.4994

274.4743

274.4994

12 0.183

235.5336

1673.4414

233.5973

315.9770

1608.0798

220.2932

144.7963

1597.3691

274.4994

. . . . . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. . .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

. .

HTR2—Hierarchical Translations and Rotations

The HTR2 (

*.htr2

) file contains rotations (about X, Y and Z axis) for the body segments defined in the

EVaRT

Project and translations and rotations for the root segment. Rotations are calculated relative to a local coordinate system of each segment’s designated parent.

The HTR files translation are expressed in the units used for

EVaRT

system calibration and the rotations are calculated as Euler angles expressed in degrees. These Euler angles are either bounded or continuous.

Bounded

—indicates that when the angles are extracted they are bounded or constrained between

± 180 degrees for the X and Z directions, and

± 90 degrees for the Y direction.

Continuous

—means that the angles will be continuous, i.e. the angles are not bounded. With unbounded angles you can conceivably have an angle that goes from 0 to 1,000 degrees for each one of the X,Y and Z angles.

This HTR file has four main parts: the Header; the Segment Names and

Hierarchies, the Base Position, and the Data.

Typically, the Base Position frame is selected when the subject’s body is in a symmetrically oriented, neutral stance position. This Base Position frame is very important because this will be the position and orientation

G-5

Appendix G: Import and Export File Formats

Example

EVaRT 5.0 User’s Manual

where each segment’s translation and rotation is set to zero and the bone length scale factor is set to 1.0.

Since all segments are hierarchical, child segments have their translations and rotations relative to their parent. The origin of a child segment is found by applying the translations relative to the parent’s coordinate system. The orientation of the child segment can be established by rotating the child’s coordinate system relative to the parents coordinate system about each of the axes.

In the data section, the order of transformation is: translation followed by rotation. The segment names are keywords, for example

head

. Each segment’s data is contained in seven columns, translations in X, Y and Z, rotations about X, Y and Z and a scale factor for the Y axis or bone length.

Each frame of data is represented on one row.

The complete file contains data for each of the 20 segments. The position of each segment is recorded for 387 video frames. Such a file is quite large, so we have included an abbreviated version here.

The file begins with a

[Header]

section containing general information, such as the number of segments, the number of frames, the frame rate, and other parameters which apply to all data in the file.

This is followed by the

[SegmentNames&Hierarchy]

section which describes the child-parent relationships of the skeleton. Notice that only the

LowerTorso

segment relates to the

GLOBAL

coordinate system. All other segments motions are described in relation to a parent segment.

In the

[BasePosition]

section, the location of each segment’s origin and rotation are described in the skeleton’s base position, using the six available degrees of freedom:

Translation in X

Translation in Y

Translation in Z

Rotation about X axis

Rotation about Y axis

Rotation about Z axis

For the child segments, location and rotation are given in terms of the parent. In this skeleton, the origin of all the children lie near the bone (Y axis) of the parent and, therefore, have only Y values. The seventh column gives the length of the bone segment.

The remaining sections contain motion data for each segment, in each frame, relative to the base position. This is frame oriented, meaning each section holds all segments for that particular frame.

In this abbreviated example, only the first and last four frames are shown for the first three segments. In the actual file, all 387 frames for each of the 20 segments would appear. After all the segments, an

[EndOfFile]

section terminates the file.

G-6

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

Figure G-3. An Example of an HTR2 File

# Hierarchical Translation and Rotation (.htr) file

# Generated by EVaRT

[Header]

FileType htr

DataType HTRS

FileVersion 2

NumSegments 20

NumFrames 511

DataFrameRate 60

EulerRotationOrder ZYX

CalibrationUnits mm

RotationUnits Degrees

GlobalAxisofGravity Y

BoneLengthAxis Y

ScaleFactor 1

[SegmentNames&Hierarchy]

#CHILD

Head

PARENT

Neck

Neck

UpperTorso

LCollarBone

RCollarBone

LUpArm

UpperTorso

LowerTorso

UpperTorso

UpperTorso

LCollarBone

RUpArm

LLowArm

RLowArm

LHand

RHand

LowerTorso

LPelvis

RCollarBone

LUpArm

RUpArm

LLowArm

RLowArm

GLOBAL

LowerTorso

RPelvis

LThigh

RThigh

LLowLeg

RLowLeg

LFoot

RFoot

[BasePosition]

#SegmentName

Head

Neck

UpperTorso

LCollarBone

RCollarBone

LUpArm

RUpArm

LLowArm

RLowArm

LHand

RHand

LowerTorso

LPelvis

RPelvis

LThigh

LowerTorso

LPelvis

RPelvis

LThigh

RThigh

LLowLeg

RLowLeg

Tx Ty Tz Rx Ry Rz BoneLength

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

1.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

RThigh

LLowLeg

RLowLeg

LFoot

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

1.0

1.0

1.0

1.0

RFoot 0.0

0.0

0.0

0.0

0.0

0.0

1.0

#Beginning of Data.

G-7

Appendix G: Import and Export File Formats

12:

13:

14:

15:

16:

17:

18:

19:

20:

Frame 1:

0:

1:

2:

3:

4:

5:

6:

7:

8:

9:

10:

11:

265.65848

-5.56661

6.64226

-5.68993

0.21077

-2.91121

7.62039

4.45230

0.27224

10.70936

5.60677

-7.93797

7.92266

17.83143

5.29669

3.41580

0.59471

-6.65161

-8.69699

56.16376

69.34235

Frame 2:

0:

1:

2:

3:

4:

5:

6:

7:

8:

9:

10:

11:

12:

13:

14:

15:

16:

17:

18:

19:

20:

265.98917

-4.76026

6.24591

-5.45918

0.25053

-2.74117

7.56453

4.44328

0.13121

10.54723

6.09020

-7.83955

7.76307

17.85974

5.44536

3.05847

0.34028

-6.65867

-8.45317

56.25188

69.34779

.....

.....

[EndofFile]

958.52289

-1.03833

-10.44067

-4.37941

8.22242

-14.64063

4.41526

6.92158

1.42203

0.37209

0.99909

-1.63432

6.08590

2.88770

4.26448

-26.21330

18.20674

-0.51561

2.51242

-3.45440

4.62345

171.77657

-3.08909

-1.80951

2.25734

-101.23232

98.93079

6.57253

-1.13849

2.06549

-6.77907

-5.37607

4.66433

0.69348

-106.49151

108.76140

-60.12212

59.84934

-9.07148

6.54966

1.11591

-3.69895

222.84282

81.65315

279.37607

148.15902

122.55443

286.27760

273.89803

197.27960

227.77268

160.71960

138.17473

116.27502

122.39289

124.66989

448.80791

469.08647

349.00357

359.40872

184.70566

181.04145

958.55890

-0.93315

-10.44039

-4.48315

8.10135

-14.49931

4.56852

6.91024

1.30640

1.02512

1.13973

-2.10847

6.08555

3.07533

3.99245

-26.22045

16.47090

-0.44105

4.60744

-3.41367

4.49733

171.80534

-1.37413

-3.43725

2.16439

-101.08283

99.00355

6.41734

-1.35979

2.05696

-6.42520

-5.56063

4.52455

0.77588

-106.59285

108.62376

-60.08476

59.49600

-9.10057

7.28953

1.03058

-4.24123

222.40665

81.65838

279.15316

148.41689

122.50407

286.17036

273.86227

196.92114

227.48704

160.93207

138.30477

116.44541

122.55485

124.97174

448.35952

468.76412

349.54976

359.82323

184.85654

181.27314

EVaRT 5.0 User’s Manual

G-8

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

HTR

HTR files provide the same information as HTR2 files, but in a segment oriented method which is less suitable for information streaming. Refer to

“HTR2—Hierarchical Translations and Rotations” on page G-5 . The

HTR format cannot created in

EVaRT

. This feature will be added to future revisions.

Examples of

HTR Files

Example of an HTR

Version 1 File

Information about the structure and motion of hierarchical skeletons is stored in

.htr

files. There are two variations of .

htr

files: version 1 (HTR) and version 2 (HTR2). The skeleton data is identical in both file versions, however, the motion data is presented on a segment basis in version 1 files, while it is on a frame basis in version 2 files. Thus, version 1 files give the position data for all frames for the first segment followed by the position data for all frames for the second segment, etc. HTR Version 1 files are used:

1.

2.

3.

to save as HTR in

EVaRT

under the Post Skeleton function in the

Tools menu for input and output from

Si 2.0

for input and output for many of the animation package software

HTR2 files are output from steaming mode from

EVaRT 3.0

and

EVaRT

, but there is no software to import them. Version 2 file gives the position data for all segments for the first video frame followed by the position data for all segments for the second frame, etc.

The example of a version 1 file shown in Figure G-4 on page G-11

was generated by Motion Analysis

MoCap Solver

and contains data for the movement of a hierarchical skeleton with one root and 19 child segments.

Since the file was generated by

MoCap Solver

, the skeleton has fixed length bones.

The complete file contains data for each of the 20 segments. The position of each segment is recorded for 196 video frames. Such a file is quite large, so we have included an abbreviated version here.

The file begins with a

[Header]

section containing general information, such as the number of segments, the number of frames, the frame rate, and other parameters which apply to all data in the file.

This is followed by the

[SegmentNames&Hierarchy]

section which describes the child-parent relationships of the skeleton. Notice that only the

LowerTorso

segment relates to the GLOBAL coordinate system. All other segments motions are described in relation to a parent segment.

In the

[BasePosition]

section, the location of each segment’s origin and rotation are described in the skeleton’s base position, using the six available degrees of freedom:

Translation in X

Translation in Y

Translation in Z

Rotation about X axis

Rotation about Y axis

Rotation about Z axis

G-9

Appendix G: Import and Export File Formats EVaRT 5.0 User’s Manual

Example of a HTR2

File

For the child segments, location and rotation are given in terms of the parent. In this skeleton, the origin of all the children lie on the bone (Y axis) of the parent and, therefore, have only Y values. The seventh column gives the length of the bone segment.

The remaining sections contain motion data for each segment relative to the base position. Each section starts with the segment name followed by position data and the scale factor (SF) for each frame. The first segment,

[LowerTorso]

, is unique because it has six values describing its relation to the global coordinate system. All subsequent segments have only three rotational degrees of freedom. Therefore, if you wished to find the global coordinates of any segment in any frame you would follow these steps:

Calculate the three rotation values for each segment using the values from the desired frame and the base position for that segment. For the root, also calculate each of the three translation values. Be careful to use the correct rotation order as indicated in the header of the

.htr

file.

Using the positions and rotation of the root segment as a starting point, calculate the global positions of the origin of the first child’s coordinate system in the hierarchy.

Using this calculated global position, calculate the global position of the origin of the next child’s coordinate system.

Continue until you have reached the desired segment.

In the abbreviated example shown in

Figure G-4 , only the first and last

four frames are shown for the first three segments. In the actual file, all

196 frames for each of the 20 segments would appear. After all the segments, an

[EndOfFile]

section terminates the file.

The example of a HTR2 file shown in Figure G-5 on page G-14 was gen-

erated using Motion Analysis

SkB

and therefore does not have fixed bone lengths.

The information in the

[Header]

,

[SegmentName&Hierarchy]

, and

[BasePosition]

is very similar to the first example. However, note that the

FileVersion

is

2

instead of

1

. Also, notice that the use of scale factor and bone lengths are reversed from the usage in version 1 files. The bone lengths given in the base position are all 1.0. The actual bone length for each segment in each frame is given in the data section.

The data section starts at frame 1. The segment data is then given in the order defined in the

[SegmentNames&Hierarchy]

section. Segment 0 is the X, Y, and Z translation values for the root segment. The four values for each remaining segment are the X, Y, and Z rotations in degrees and the bone length in calibration units.

There is no end of file section, since the number of frames is already defined under

NumFrames

in the

[Header]

section.

G-10

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

Figure G-4. An Example of a HTR (Version 1) File

[Header]

FileType htr

DataType HTRS

FileVersion 1

# Header keywords are followed by a single value

# single word string

# Hierarchical translations followed by rotations and Scale

# integer

NumSegments

NumFrames

20

196

# integer

# integer

DataFrameRate 30 # integer

EulerRotationOrder XYZ # one word string

CalibrationUnits mm

RotationUnits Degrees

GlobalAxisofGravity Y

BoneLengthAxis Y

ScaleFactor 1.0000

# one word string

# one word string

# character, X or Y or Z

[SegmentNames&Hierarchy]

#CHILD PARENT

LowerTorso

UpperTorso

LCollarBone

RCollarBone

GLOBAL

LowerTorso

UpperTorso

UpperTorso

LUpArm

RUpArm

LLowArm

RLowArm

LHand

RHand

LPelvis

RPelvis

LCollarBone

RCollarBone

LUpArm

RUpArm

LLowArm

RLowArm

LowerTorso

LowerTorso

LThigh

RThigh

LLowLeg

RLowLeg

LFoot

RFoot

Neck

Head

LPelvis

RPelvis

LThigh

RThigh

LLowLeg

RLowLeg

UpperTorso

Neck

[BasePosition]

#SegmentName Tx, Ty, Tz, Rx, Ry, Rz, BoneLength

LowerTorso238.320832923.726971241.2948288.8069650.0000002.422863 141.720766

UpperTorso0.000000141.7207660.000000-9.396187-0.112582-0.4.226674324.970754

LCollarBone0.000000324.9707540.000000-8.9120410.520925-117.992062155.689602

RCollarBone0.000000324.9707540.000000-7.516124-0.516778118.838556127.553756

LUpArm 0.000000155.6896020.000000

11.284636

-3.556895 24.261557

273.483757

RUpArm 0.000000127.5537560.000000

12.990029

2.960321

-24.250841 285.322188

LLowArm0.000000273.4837570.000000

-15.450962 -0.327390 -5.501311

318.246332

RLowArm0.000000285.3221880.000000

-12.465152 -0.572074 -5.115813

305.910223

LHand 0.000000318.2463320.000000

10.682556

-5.620492 27.438027

85.440039

RHand 0.000000305.9102230.000000

5.730878

2.097665

-20.038204 98.351413

LPelvis0.0000000.000000 0.000000

RPelvis0.0000000.000000 0.000000

24.672467

39.808415

-6.972213 -127.877478122.309825

6.985094

127.739824 133.924520

LThigh 0.000000122.3098250.000000

-12.585726 -15.440180-52.153190388.012318

G-11

Appendix G: Import and Export File Formats

#Beginning of Data. Separated by tabs

[LowerTorso]

#Fr

1

Tx Ty Tz Rx Ry Rx SF

1262.497925-15.182068-2245.4418950.5336240.713565-1.0697651.000000

2

3

1262.534546-15.109411-2245.4282230.5902200.697084-1.1088881.000000

1262.367920-15.051557-2245.3151860.5882480.707301-1.1631961.000000

4 1261.811279-15.027791-2245.1335450.6069780.696131-1.3261221.000000

. . . . . . . .

. . . . . . . .

. . . . . . . .

193

194

-25.763012-11.505680-116.749229-85.78657582.456894-104.4607321.000000

-31.705627-12.149048-131.977005-81.80329980.946663-100.7663191.000000

195

196

-40.086079-12.445135-144.204391-77.85765879.605011-97.4399491.000000

-49.986629-12.544500-152.518951-74.27449278.596512-94.2411671.000000

[UpperTorso]

#Fr

1

Tx Ty Tz

0.0000000.000000 0.000000

2

3

0.0000000.000000 0.000000

0.0000000.000000 0.000000

4 0.0000000.000000 0.000000

. . . . . . . .

. . . . . . . .

. . . . . . . .

193

194

0.0000000.000000 0.000000

0.0000000.000000 0.000000

195

196

0.0000000.000000 0.000000

0.0000000.000000 0.000000

Rx Ry Rx SF

-2.259649

-2.946399 -0.475843

1.000000

-2.341502

-2.767608 -0.420984

1.000000

-2.331526

-2.552349 -0.414928

1.000000

-2.328715

-2.410058 -0.305621

1.000000

2.688713

-0.354111 4.239368

1.000000

0.312819

-2.503460 6.899297

1.000000

-1.617305

-3.427358 9.416446

1.000000

-2.357207

-3.057341 10.941742

1.000000

[LCollarBone]

#Fr

1

Tx Ty Tz

0.0000000.000000 0.000000

2

3

0.0000000.000000 0.000000

0.0000000.000000 0.000000

4 0.0000000.000000 0.000000

. . . . . . . .

. . . . . . . .

. . . . . . . . .

193 0.0000000.000000 0.000000

Rx Ry Rx SF

2.648904

1.239630

0.547259

1.000000

2.753487

1.291442

0.549240

1.000000

2.832925

1.339503

0.493109

1.000000

2.839463

1.330919

0.571445

1.000000

2.837519

0.820427

3.997483

1.000000

G-12

EVaRT 5.0 User’s Manual

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

194

195

196

0.0000000.000000 0.000000

0.0000000.000000 0.000000

0.0000000.000000 0.000000

1.948793

0.402037

3.829116

1.000000

1.872437

0.383394

3.696403

1.000000

2.747322

0.820163

3.692526

1.000000

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

[EndOfFile]

G-13

Appendix G: Import and Export File Formats EVaRT 5.0 User’s Manual

Figure G-5. An Example of a HTR2 File

Hierarchical Translation and Rotation (.htr) file

# Generated by EVaRT

[Header# Header keywords are followed by a single value

FileType htr# single word string

DataType HTRS# Hierarchical translations followed by rotations and Scale

FileVersion 2# integer

NumSegments 20# integer

NumFrames 511 # integer

DataFrameRate 60# integer

EulerRotationOrder ZYX# one word string

CalibrationUnits mm# one word string

RotationUnits Degrees# one word string

GlobalAxisofGravity Y# character, X or Y or Z

BoneLengthAxis Y

ScaleFactor 1

[SegmentNames&Hierarchy]

#CHILDPARENT

HeadNeck

NeckUpperTorso

UpperTorsoLowerTorso

LCollarBoneUpperTorso

RCollarBoneUpperTorso

LUpArmLCollarBone

RUpArmRCollarBone

LLowArmLUpArm

RLowArmRUpArm

LHandLLowArm

RHandRLowArm

LowerTorsoGLOBAL

LPelvisLowerTorso

RPelvisLowerTorso

LThighLPelvis

RThighRPelvis

LLowLegLThigh

RLowLegRThigh

LFootLLowLeg

RFootRLowLeg

[BasePosition]

#SegmentNameTxTyTz Rx Ry RzBoneLength

Head 0.0 1.0 0.0 0.0 0.0 0.0 1.0

Neck 0.0 1.0 0.0 0.0 0.0 0.0 1.0

UpperTorso 0.0 1.0 0.0 0.0 0.0 0.0 1.0

LCollarBone 0.0 1.0 0.0 0.0 0.0 0.0 1.0

RCollarBone 0.0 1.0 0.0 0.0 0.0 0.0 1.0

LUpArm 0.0 1.0 0.0 0.0 0.0 0.0 1.0

RUpArm 0.0 1.0 0.0 0.0 0.0 0.0 1.0

LLowArm 0.0 1.0 0.0 0.0 0.0 0.0 1.0

RLowArm 0.0 1.0 0.0 0.0 0.0 0.0 1.0

LHand 0.0 1.0 0.0 0.0 0.0 0.0 1.0

RHand 0.0 1.0 0.0 0.0 0.0 0.0 1.0

LowerTorso 0.0 0.0 0.0 0.0 0.0 0.0 1.0

LPelvis 0.0 0.0 0.0 0.0 0.0 0.0 1.0

G-14

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

ANC—Analog ASCII Row Column

ANC (

.anc

) files contain ASCII analog data in row-column format. The data is derived from

*.anb

analog binary files. These binary

*.anb

files are generated simultaneously with video

*.vc

files if an optional analog input board is used in conjunction with video data capture.

To create an

*.anc

file from an

*.anb

file, from the main menu select

File > Export ANC

. The data in ANC files is raw analog data in ASCII form and can be read and manipulated by a spreadsheet program.

Shown in

Figure G-6 is the beginning portion of an

*.anc

file.

Figure G-6. Example of an ANC File

File_Type: Analog R/C ASCII

Board_Type: National AT-MIO-64F-5

Trial_Name:1ndbfw Trial#: 8

Generation#:1

Polarity:Bipolar

Duration(Sec.): 6.000000#Channels:30

Name ant f1x f1y f1z m1x m1y m1z f2x f2y f2z m2x m2y m2zL tibialis

Rate 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000

Range 10000 10000 10000 10000 10000 10000 10000 10000 10000 10000 10000 10000 2500

0.0000

0.0010

0.0020

0.0030

0.0040

0

-1

-1

-1

0

1

0

0

0

0

-29

-29

-29

-29

-29

-1

-1

-1

-2

-1

-2

-2

-3

-3

-2

0

0

0

0

0

-1

-2

-2

-1

-2

0

-1

-1

-1

-1

-13

-13

-13

-13

-13

2

1

1

1

1

3

2

2

2

3

-11

-10

-10

-10

-10

-550

-392

-369

-440

-342

0.0050

0.0060

0.0070

0.0080

0.0090

0.0100

0.0110

1

0

0

0

0

0

0

1

0

0

0

1

0

0

-29

-29

-28

-28

-28

-28

-29

-1

-1

-1

-1

-1

-1

0

-2

-2

-2

-2

-2

-2

-2

-1

0

0

0

1

0

0

-2

-2

-2

-2

0

-1

-1

0

-1

-1

-1

0

0

-1

-13

-13

-12

-13

-12

-12

-13

1

1

1

1

2

2

1

3

3

3

2

3

4

3

-11

-11

-10

-10

-531

-803

-738

-10 -485

-10 453

876

-10 1401

0.0120

0.0130

0.0140

0.0150

0.0160

0.0170

0.0180

0.0190

0.0200

0.0210

0

-2

0

0

0

0

-1

0

-1

-1

1

-1

0

0

0

0

0

0

0

0

-29

-30

-29

-28

-29

-28

-29

-29

-29

-29

-1

-3

-1

-1

-1

-1

-1

-2

-1

0

-2

-3

-2

-2

-2

-2

-2

-2

-2

-2

0

0

0

0

0

0

0

0

0

0

-1

-3

-2

-2

-1

-1

-2

-2

-2

-1

-1

-3

-1

-1

-1

-1

-1

-2

-1

-1

-13

-13

-13

-13

-13

-13

-13

-12

-13

-13

1

2

1

1

0

2

2

1

1

1

3

3

2

2

2

2

3

2

2

3

-10 598

-10 -141

-11 -457

-10 446

-11 569

-10

-10

-10

-10

-11

507

689

501

348

209

G-15

Appendix G: Import and Export File Formats EVaRT 5.0 User’s Manual

TS—Time Series Files from the EVaRT Analysis Functions:

Velocity and Acceleration Calculations

EVaRT

Analysis Functions (F7) is a selectable view from the Data Views menu and allows the user to calculate velocity and accelerations of marker data, distances between markers and Included Angles. The Distances and Included angles are assumed to be self-documenting and a description of the Position, Velocity and Acceleration tabs is below. These data can be Exported to the .ts (Time Series) files.

Figure G-7. Analysis Functions

G-16

There is a Frames Factor, which can be set to 3, 5, 7, or 9 frames. This selects the number of frames to use for the velocity and acceleration calculations. If the Frames Factor is set to 3 frames, velocity data for the 5-th frame is calculated exclusively from frame 4 and frame 6. Velocity Data for frame 1 does not exist; velocity and frame data starts at frame 2. If the

Frame Factor is set to 5, the velocity data comes exclusively from –2 frames to +2 frames from the i-th frame. Larger Frame Factors have the effect of smoothing the data.

X1, Y1, Z1 Positional data is determined from the marker locations.

Velocity Calculation is done with a central difference. Let FR represent the Frame Rate of the camera. Time difference between frames = 1 / FR in the below calculations.

Velocity Calculation for frame i with a Frames Factor of 3: vX1 ( i ) = FR* ( X ( i+1) - X (i-1) ) / 2 vY1 ( i ) = FR* ( Y ( i+1) - Y (i-1) ) / 2 vZ1 ( i ) = FR* ( Z ( i+1) - Z (i-1) ) / 2

Velocity Calculation for frame i with a Frames Factor of 5: vX1 ( i ) = FR* ( X ( i+2) - X (i-2) ) / 4 vY1 ( i ) = FR* ( Y ( i+2) - Y (i-2) ) / 4 vZ1 ( i ) = FR* ( Z ( i+2) - Z (i-2) ) / 4

EVaRT 5.0 User’s Manual Appendix G: Import and Export File Formats

Velocity Calculation for frame i with a Frames Factor of 7: vX1 ( i ) = FR* ( X ( i+3) - X (i-3) ) / 6 vY1 ( i ) = FR* ( Y ( i+3) - Y (i-3) ) / 6 vZ1 ( i ) = FR* ( Z ( i+3) - Z (i-3) ) / 6

Velocity Calculation for frame i with a Frames Factor of 9: vX1 ( i ) = FR* ( X ( i+4) - X (i-4) ) / 8 vY1 ( i ) = FR* ( Y ( i+4) - Y (i-4) ) / 8 vZ1 ( i ) = FR* ( Z ( i+4) - Z (i-4) ) / 8

Resultant velocity scalar: vR1 (frame i) = SQRT( vX1**2 + vY1**2 + vZ1**2)

Accelerations for Frame i are calculated as the differences in velocity as:

A (frame i ) = Velocity (frame I >frame i+1) - Velocity (frame i-1

>frame i)

Time difference between frames = 1/ FR .

Acceleration Calculations sing the Frame Rate (FR) of the camera for a

Frames Factor of 3: aX1 ( i ) = FR*FR* ( X (i+1) - 2* X( i ) + X ( i-1 ) ) aY1 ( i ) = FR*FR* ( Y (i+1) - 2* Y( i ) + Y ( i-1 ) ) aZ1 ( i ) = FR*FR* ( Z (i+1) - 2* Z( i ) + Z ( i-1 ) )

For Frames Factor of 5: aX1 ( i ) = FR*FR* ( X (i+2) - 2* X( i ) + X ( i-2 ) ) / 4 aY1 ( i ) = FR*FR* ( Y (i+2) - 2* Y( i ) + Y ( i-2 ) ) / 4 aZ1 ( i ) = FR*FR* ( Z (i+2) - 2* Z( i ) + Z ( i-2 ) ) / 4

For Frames Factor of 7: aX1 ( i ) = FR*FR* ( X (i+3) - 2* X( i ) + X ( i-3 ) ) / 9 aY1 ( i ) = FR*FR* ( Y (i+3) - 2* Y( i ) + Y ( i-3 ) ) / 9 aZ1 ( i ) = FR*FR* ( Z (i+3) - 2* Z( i ) + Z ( i-3 ) ) / 9

For Frames Factor of 9: aX1 ( i ) = FR*FR* ( X (i+4) - 2* X( i ) + X ( i-4 ) ) / 16 aY1 ( i ) = FR*FR* ( Y (i+4) - 2* Y( i ) + Y ( i-4 ) ) / 16 aZ1 ( i ) = FR*FR* ( Z (i+4) - 2* Z( i ) + Z ( i-4 ) ) / 16

Resultant acceleration scalar: aR1 (frame i) = SQRT( aX1**2 + aY1**2 + aZ1**2)

G-17

Appendix G: Import and Export File Formats EVaRT 5.0 User’s Manual

Binary Files—ANB, TRB, and C3D

The following are binary files and cannot be directly read or manipulated by the end user. Their function and context are briefly described.

ANB

These files contain up to 64 channels of analog data collected simultaneously with video data by the optional analog board. The data in these files can be converted to readable ASCII form as either an ANA or ANC file. To do so, from the main menu select

File > Export ANC.

Note:

The data contained in the

*.anb

file has a dynamic range of –2048 to

+2047, which represents 12 bit signed numbers.

EVaRT

scales the specified input voltage range to this range of values.

TRB

These files contain the same 3D track data as ASCII

.trc

files, saved in a compact binary form. In addition to the data in

.trc

files,

.trb

files contain the following data for each frame:

• a list of the cameras used to calculate the 3D marker position

• the residual of the 3D position calculation

C3D

These are a special binary files that contains both scaled 3D track data and unscaled analog data. For more information, you can visit the C3D website at www.c3d.org.

G-18

Appendix H

SIMM Motion Module

Topic

Introduction

Opening Tracked Marker Files

Analog Data

Using the Mocap Model

Analog Configuration Files

SIMM Motion Module Guide to Mocap Model Markers

Page

H-1

H-2

H-6

H-9

H-16

H-17

Introduction

The Motion Module is an optional component to SIMM (Software for Interactive Musculoskeletal Modeling) that allows you to easily import data recorded by a motion capture system. It reads files containing tracked marker data (3D positions of markers in global space) using the TRC or

TRB file format developed by Motion Analysis Corporation. It can also read analog files in the ANB or ANC format with ground-reaction force and EMG data that was recorded in sync with the motion. The Motion

Module can also read C3D files, which contain both tracked marker and analog data in the same file. Additionally, the real-time version of the Motion Module can connect to a Motion Analysis system and receive and display motion and analog data in real-time, as it is being recorded.

Files of tracked marker data contain a sequence of frames, each representing a snapshot of the subject’s motion at a particular instant in time. Each frame contains the X, Y, and Z coordinates, expressed in a global coordinate system, of all the identified markers. A frame of marker data can thus be thought of as a “marker cloud” because the coordinates are not organized by body segment.

The Motion Module imports tracked marker data and fits a SIMM model within the marker cloud for each time frame. If the SIMM model contains markers whose names and positions match those of the markers placed on the subject, the Motion Module can adjust the model’s gencoord values to determine a “best fit” of the model to the marker cloud. The quality of a fit is determined by how closely each of the model’s markers is to its corresponding marker in the marker cloud. It then uses this best fit as the starting position for solving the next frame of data. The result is a SIMM motion that matches the tracked marker data. The model that is used to fit the data can either be one that you create or the pre-made model (the

mocap model

) that comes with the Motion Module.

H-1

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

The Motion Module has two primary components. The first component reads files containing tracked marker data (in the TRC, TRB, or C3D format) and creates SIMM motions from them, as described above. For more information on how this process works, and the various options for importing marker data, reference the SIMM User’s Guide.

The second component of the Motion Module creates a musculoskeletal model of a given individual by scaling a generic full-body model (the

mocap model

) based on tracked marker data from a static pose. The algorithms that are used to scale the model are the same as those used in

OrthoTrak, a full-body gait analysis package available from Motion Analysis. For more information on the mocap model and how it is created and used, see Section 5.3, Using the Mocap Model.

Opening Tracked Marker Files

SIMM can import tracked marker data that is stored in either a TRB or

TRC data file. These file formats, described in the

EVa

and

EVaRT

manuals, contain X, Y, and Z coordinates for each identified marker for each time frame. You can also import analog data files containing forceplate and EMG data recorded during the motion. These analog data files can be in either the ANB or ANC formats. The Motion Module can also read

XLS files containing other motion-related data that you may want to view in SIMM, such the kinetic data contained in an OrthoTrak single trial spreadsheet. For more information on importing analog and XLS files, see Section 5.2.3, Analog Data.

The Motion Module can also read C3D data files. These files contain tracked marker and analog data in the same file, so you only need to load one file to import all of your motion data from a trial.

When you open a tracked marker file (along with any associated analog files), SIMM attempts to map the data onto the current musculoskeletal model, thus creating a SIMM motion that is linked to the model. Therefore, to open a tracked marker file, you must already have loaded into

SIMM a model that contains the same marker set used in the marker file.

For best results, you should make sure that every marker in the tracked marker file is also in the SIMM model, and that their locations in the

SIMM model match where they were placed on the subject. The marker names should match exactly (except that they are case-insensitive). If the file contains markers that are not in the model, their data will be ignored by the Motion Module. Similarly, if the model contains markers that are not in the file, they will not be used to help fit the model to the motion data.

If you need to add, rename, or move markers in your SIMM model before loading a tracked marker file, you can use the Marker Editor to do so. See

Section 2.12, Marker Editor, for more details.

Selecting

Tracked Marker

Files

To import a tracked marker file into SIMM, first make sure that the model you want to apply it to is the current model (the topmost window in

SIMM). Then select

File > Open

from the menu bar. When the Windows file browser appears, change the Files of Type popup menu at the

H-2

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

bottom of the browser to

MAC Files

or to

C3D Files

then navigate to the folder containing the tracked marker file[s] you want to import.

Figure H-1. Windows File Browser

Next, select the appropriate marker file[s] in the file browser. Click the

Open

button to import the file[s]. At this point SIMM will display a dialog box allowing you to specify several options for importing each data file into SIMM.

Note: If your analog data files have the same base name as your TRB/

TRC file (

e.g.

,

subject14.trc, subject14.anb, and subject14.xls

), then it is not necessary to select analog files in the file browser. SIMM will automatically open any analog or XLS files with the same base name and in the same folder as the tracked marker file (there is an option in the dialog box to turn off this feature). If you are loading C3D files, this is not an issue since all of the data for the motion are stored in the C3D file.

Tracked Marker

Options Dialog

Import Frames, To,

Increment

Once you have selected one or more tracked marker files using the process described in the previous section, SIMM displays a dialog box for each one (in sequence), allowing you to set some options for importing the marker data. In many cases, you will want to use the default settings for these options, so you can simply click the

OK

button to import the motion. The following list describes each option in the dialog box.

These fields allow you to specify the range of frames to read from the marker file, as well as the increment. To use them, type into the first two fields the starting and ending frame numbers that you want to import. The third field specifies the increment to use when reading frames from the file. For example, to read every other frame from the file, enter an increment of 2. The starting frame number field and the increment field are initialized to 1. The ending frame number is initialized to the number of frames in the marker file.

H-3

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

Quick Solve

The Motion Module contains two optimization algorithms for fitting the musculoskeletal model to the marker data. The default method is fairly robust- it is designed to handle cases in which several markers are missing from a frame or in which the markers move large amounts between frames. The other method, called quick solve, is less robust but works up to twice as fast as the default method. If speed is an issue, and you know that your marker data is well-behaved, you may want to turn this option on to use the faster optimization algorithm.

Crop Ends

Tracked marker data files often have frames at the beginning and end of a motion that are missing some markers (because the subject is outside the camera volume). To automatically detect and ignore these frames as the file is read, turn on this option (it is on by default). When the option is on,

SIMM will start at the first frame and delete it if it is missing one or more markers. It will then continue to scan forward through the frames, deleting each one, until it encounters a frame containing all of the markers. It will then do the same procedure starting at the last frame and working backwards. SIMM will not remove frames with missing markers that are in between full frames, so there may still be frames in the motion that are missing markers.

Calculate

Derivatives

When loading a motion, SIMM has the capability of calculating derivatives of the motion variables. When this option is turned on, after SIMM has solved the marker data and created a SIMM motion, it will calculate first-order time derivatives of the generalized coordinate values (

i.e

., joint velocities) during the motion. It will also calculate derivatives of any force or EMG data in the analog file (if present). These derivatives can then be plotted using the Motion Curves command in the Plot Maker (see

Section 2.5.2 for more details).

Figure H-2. Tracked Marker Import Dialog Box

H-4

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

Show Markers

X Axis Units, Start at Zero

Auto-Load Analog

Data

Calibrate Forces

Remove Old

Forceplates

Read Marker

Names From

This checkbox turns on the display of the global marker positions in each frame when playing back a motion. When it is on, SIMM will add spherical motion objects to the motion, representing the location of each marker, as recorded in the marker file. When you animate the model according to the motion, the blue spheres represent these actual, recorded marker locations. These are the marker locations that the Motion Module is trying to fit the model to, for each frame. It can be helpful to display them in the model window in order to visualize how good the fit is, and to help debug problems with the data.

These options give you control over the specification of the X axis of the motion that is created from the marker data. The units along the X axis can be either time (in seconds), or frame number. The starting X value of the motion will be 0.0 if the units are time, and 1 if the units are frame number, unless frames of data are cropped because of missing markers

(see

“Crop Ends” on page H-4 ). For example, if 12 frames of data are

cropped from the beginning of the motion (and the data frequency is 60

Hz), the starting X value will be 0.2 seconds for units of time, and 13 for units of frame number. If you want the X values to start at 0.0 (or 1 for frame number) even if frames are cropped, turn on the start at zero option.

When this box is checked, SIMM will look for and automatically load any analog (ANB, ANC) or XLS data files with the same base name as the

TRB/TRC file. If SIMM did not detect the presence of any analog files when the TRB/TRC file was selected, this option is grayed out.

If you selected a C3D file with the file browser, then this box controls whether or not the analog data will be read from the C3D file.

If an analog file is present, and the auto-load analog data box is checked

(see above), then this box is active and gives you control over the calibration of the forceplate data. When this box is checked, SIMM determines the baseline of each forceplate channel and automatically subtracts these baseline values from the data, thus “zero-ing out” the force data.

In order to display forceplate data that is in the analog file, SIMM creates graphical objects in the model window representing the forceplates. Each time you load a tracked marker file with corresponding analog data,

SIMM creates a new graphical object for each forceplate in the file. In most cases you will want to remove the existing forceplate objects from the model when loading a new file, so that the display is not cluttered with multiple (or redundant) sets of objects. Thus this option is turned on by default. If you load a series of marker and analog files that all have the same forceplate definitions, then you should leave this option turned on.

For C3D import only:

This option allows you to choose from which parameter field in the C3D file to read the names of the tracked markers. Because the

POINT:LABELS

field in a C3D file is limited to four characters, some software packages (

e.g.,

EVaRT) store the full marker name in the

POINT:DESCRIPTIONS

field. Since the marker names in the tracked file must exactly match the names used in the mocap model, if your C3D file does not contain full marker names in the

POINT:DESCRIPTIONS

field, you may have to edit the mocap model so that the marker names match the four-character names stored in the

POINT:LABELS

field.

H-5

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

Save HTR File

Save Motion File

This option allows you to save an HTR file containing the motion that

SIMM calculates from the marker data. This HTR file cannot be read back into SIMM, but is useful if you want to import the motion into another software package. If this box is checked, a browse button is enabled that allows you to specify the name and location of the HTR file.

This option allows you to save a SIMM motion file containing the motion that SIMM calculates from the marker data. This file contains exactly the same data that is in the motion that SIMM loads onto the model. You can load this motion file into SIMM at a later time, rather than re-importing the marker file. If this box is checked, a browse button is enabled that allows you to specify the name and location of the motion file.

Analog Data

Analog data files contain forceplate and EMG data that was collected in sync with motion data. When loading C3D files, there are no separate analog files; all of the analog data is contained in the C3D file. When loading TRB/TRC files, you can load analog files only if they correspond to the chosen TRB/TRC files. If the analog file has the same base name as the TRB/TRC file, then the Motion Module will load it automatically when you select the marker file. Otherwise, you should select the analog file as well in the file browser. The same holds for XLS files, which are not actually analog files, but are treated similarly. XLS files can contain other data corresponding to the recorded motion, such as kinetic data calculated by OrthoTrak and stored in a “single trial spreadsheet.”

SIMM can recognize three types of analog data: ground reaction forces,

EMG activation levels, and “other” data (usually kinetic data from an

XLS file). These data types, and how they are interpreted by SIMM, are described below:

Forceplate Data

SIMM displays forceplate data by drawing a vector in the model window at the appropriate point of application and with a size corresponding to the magnitude of the force. Forceplate data in an analog file are voltages measured by forceplate transducers. These voltages are converted into forces using a calibration file,

forcepla.cal

. This file is the same one used by

EVa, EVaRT, and OrthoTrak. To use it with SIMM, you should put a copy of it in the same folder as your motion data, or in the folder

SIMM\Resources\ mocap\misc

. If you have only one forceplate configuration for your motion capture system, it is preferable to put

forcepla.cal

in

SIMM\Resources\mocap\misc,

rather than copying it into every folder of motion data.

Note:

C3D files that contain force plate data also contain the calibration information for the plates. Thus there is no separate calibration file that SIMM reads when importing C3D files.

SIMM also uses another configuration file,

importVariables.txt

, to map forceplate channels to SIMM variables. This file is located in

SIMM\Resources\mocap\misc

, and contains mappings for typical channel names for up to six forceplates. You will only need to change this file if you use more than six forceplates, or use forceplates that have exotic channel con-

H-6

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

EMG Data

Other Data

figurations. This file is used when loading ANB/ANC files and when loading C3D files.

SIMM displays EMG data by varying the sizes and colors of the corresponding muscles in the SIMM model. EMG data in an analog file are voltages measured by the EMG system. SIMM rectifies and smooths these data, and then scales them based on an MVC value (maximum voluntary contraction), resulting in a smooth muscle excitation level that varies between 0.0 and 1.0. If MVC values are located in the configuration file

importVariables.txt

, SIMM will use them to scale the EMG data. If

MVC values are not present, SIMM will use each muscle’s maximum voltage in the analog file to scale that muscle’s EMG data (thus each muscle’s excitation will peak at 1.0 sometime during the motion). The file

importVariables.txt,

located in

SIMM\Resources\mocap\misc

, contains mappings between typical EMG channel names and the muscle names in the mocap model. It does not contain any MVC values. In most cases, however, it is sufficient to not specify them and use SIMM’s default scaling method.

“Other” data is contained in XLS files, and can represent any motion variable that you choose to calculate and store in the file. It is usually reserved for kinetic data (

e.g

., joint moments and powers) that OrthoTrak calculates and stores in its spreadsheet (XLS) format. It may also include motion events, such as toe-off and heelstrike, that are stored at the top of the

XLS file. SIMM does not perform any calculations on these data, but does import them so that you can create plots of them in the Plot Maker. SIMM will only import “other” data that are identified as such in

importVaria b l e s . t x t

. T h i s c o n f i g u r a t i o n f i l e , l o c a t e d i n

S I M M \ R e sources\mocap\misc

, contains mappings between OrthoTrak and the mocap model of all forces, moments, and powers for the hip, knee, and ankle joints. You will only need to edit this file if you want to import data other than these.

Real-time Import

In addition to importing tracked marker files, SIMM can import motion data that is sent over the network in real-time from EVaRT. SIMM is thus able to animate a musculoskeletal model and plots of joint angles and muscle lengths while the subject’s motion is being recorded. For this realtime connection,

EVaRT

solves tracked marker data using the mocap model. It then sends generalized coordinate values (as well as analog data) over the network to the SIMM computer. If the same mocap model is loaded into SIMM, these generalized coordinates will drive the animation of the model in real-time, with a small delay (whose length depends on the network speed and the graphics speed of the SIMM computer).

Follow these steps to use the real-time connection between EVaRT and

SIMM:

First-time setup only:

1.

Find the folder

SIMM\EVaRT

on your SIMM computer and look for

mocap.jnt

and

solver.dll

(

solver.dll

may be hidden in the folder view because it is a system file). Copy both files to the folder on the

EVaRT

machine (you’ll need to exit

EVaRT

first, if it is running).

This will guarantee that

EVaRT

is using the same mocap model and the same scaling algorithms as SIMM uses.

H-7

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

2.

3.

Open the text file

SIMM\Resources\preferences

in a text editor such as Notepad or Wordpad. Locate the line that reads:

EVART_MACHINE

<hostname>

, and change the hostname to the name of your

EVaRT

machine. Save and close the preferences file (make sure that the file is saved as ASCII text with the name preferences—Wordpad likes to surreptitiously append a

.txt

extension when it saves files that don't already have a filename extension).

If your motion capture system includes forceplates, copy the file

forcepla.cal

from your

EVaRT

computer onto your SIMM computer and put it in the folder

SIMM\Resources\mocap\misc

.

Each motion capture session:

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Copy the folder containing the motion data from the EVaRT computer to the SIMM computer (or make it shared). If there is a

personal.dat

file for this data, make sure it is in the folder too.

Launch

EVaRT

. Load the appropriate project.

Select

File > Load Tracks File

and select the tracked marker file corresponding to the static trial for the subject.

Under

Setup > Misc

, click on the radio button for

SIMM OrthoTrak

Solver

, located in the Skeleton Options area.

Launch SIMM.

Select

File > Open Mocap Model

and navigate to the motion folder.

Choose the tracked marker file containing the static pose.

Set the options as desired in the dialog box and click

OK

.

Open the Model Viewer window.

In the Model Viewer window, choose

Start > Realtime Connection to <hostname>

. SIMM will display a dialog box allowing you to set some options for the connection. The motion buffer size options control how many seconds of motion data are saved in SIMM’s buffer.

The

time scale

options let you specify the minimum and maximum values, in seconds, for the time scale of the motion. If you want the scale to remain fixed between minimum and maximum, check the sliding checkbox, otherwise the scale will continue to increase as new data is received.

SIMM will now wait to receive data from the

EVaRT

computer. Once the connection is established, SIMM will display “connected” in its message window, and the SIMM model will begin tracking the motion of the subject in real-time. You can pan, zoom, rotate, and change the draw mode of the SIMM model as it is tracking the motion. You can also create plots of kinematic variables and muscle properties and see the plots change in real-time.

To disconnect SIMM from the real-time stream, click the

Stop

button in the Model Viewer. You can play back the last N seconds of the motion.

Note:

When analog data is imported into SIMM in real-time, it is processed slightly differently than when the data is post-processed in SIMM. This is because the real-time analog data is processed frame-by-frame, without the benefit of the full data set. This has the following implications:

To set the baseline for the forceplates, the first frame of force data is used as the “zero” level. Thus when you first connect SIMM to

EVaRT

, you should make sure that nothing is on the forceplates.

H-8

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

If no MVC levels have been specified for some of the muscles, a running tally of each muscle’s maximum level will be kept, and used to scale the EMG signals into the range 0.0 to 1.0. Thus if you want to accurately scale EMG levels throughout a real-time SIMM motion import, you should either specify MVC levels, or have the subject perform MVCs just after connecting SIMM to EVaRT.

Using the Mocap Model

SIMM has the ability to read tracked marker data and convert it into a motion by fitting a musculoskeletal model to it. For this to work well, the body segment lengths, marker names, and marker locations in the model must exactly match those for the subject whose motion is being recorded.

Because it is time consuming to measure and scale the body segments, and measure and record the offsets of all of the markers, the Motion Module has the ability to automatically scale a pre-made model (the mocap model) to fit the subject.

To use the mocap model, select

Open Mocap Model

from the

File

menu.

SIMM will display a Windows file browser and ask you to select the name of a static pose file. This static pose is used to calculate joint center locations and segment lengths for the subject, using the same algorithms implemented in OrthoTrak. In other words,

the Motion Module recreates the OrthoTrak skeletal model from the static pose, and then maps this skeletal model onto the mocap model

. Thus to use the mocap model, you need to use the same motion capture protocol as you would for OrthoTrak. You can use either the Helen Hayes or Cleveland

Clinic marker sets (plus your own additional markers, if desired), as long as the marker names and locations match the protocol defined in the

OrthoTrak manual. The Motion Module uses the tracked marker data from the OrthoTrak static pose, and also segment information from

personal.dat,

to scale the mocap model to the subject. The algorithms for calculating joint center locations and segment lengths have been designed to be as similar as possible to the OrthoTrak algorithms. This was done so that motion information in SIMM (

e.g

., joint angles, EMG levels) would match the corresponding information in OrthoTrak, and also so that you would not have to change your OrthoTrak protocol in order to use SIMM.

The mocap model and the algorithms used to scale it are described in the following sections.

The Mocap

Model

The mocap model is a full-body SIMM model that has been customized for gait analysis, but can be used to import and display any type of fullbody motion. The model has 41 body segments, 41 joints, 40 degrees of freedom, and 88 lower-extremity muscles. It represents an adult male, approximately 175 cm. tall, with a mass of 78 kg. The model is scaled to match the size of the motion capture subject using algorithms described in

Section 5.3.4. The model’s joints have been carefully constructed to represent normal joint motion as closely as possible.

To load the mocap model into SIMM, the software looks for the

MOCAP_MODEL

variable in the preferences file (

SIMM\Resources\preferences

) to get the name of the joint file that comprises the mocap model.

T h e d e f a u l t s e t t i n g f o r t h i s v a r i a b l e i s

S I M M \ R e s o u r c e s \

H-9

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual mocap\mocap.jnt.

This joint file includes the file

mocap.msl

to get the definitions of the muscles.

You may change the mocap model however you wish. For example, you can add or remove muscles from the model, or change the tendon and fiber parameters of existing muscles. You can also add degrees of freedom to the model, in order to more accurately represent a particular motion (

e.g.

, adding toe joints and gencoords to examine toe motion in greater detail). If you modify the mocap model, however, you should keep in mind two things.

First, the model has been set up to correspond to the skeletal model that

OrthoTrak uses when processing gait data. The lower-extremity body segments and orientations of the reference frames closely match those in the

OrthoTrak model. Also, each body segment in the mocap model is scaled to fit the subject by relating its length to the length of an OrthoTrak segment. These relations are specified in the mocap model by defining scale segments and scale factors for each body segment. If you add, delete, or modify joints or body segments in the mocap model, you should make sure that each segment still properly relates to an OrthoTrak segment.

Second,

mocap.jnt

contains several macros that are used to properly define the orientation of the floor, and to automatically remove the upper body segments if there are no upper body markers. When SIMM reads a joint file, it performs these macros but does not save them internally. Thus when it writes out a joint file, all of the macros have been removed. If you make changes to the mocap model in SIMM and then save the new model to a file, do not replace

mocap.jnt

with the new file. Instead, copy the relevant portions of the new file into

mocap.jnt

using a text editor, thus preserving the macros and comments.

The Static Pose

When you open the mocap model, SIMM prompts you for the name of a tracked marker file containing a static pose of the subject. This static pose is the same one used by OrthoTrak, and for it you can use any of the six marker sets identified by that software package:

Cleveland Clinic

Lower Body

,

Cleveland Clinic Full Body

,

Cleveland Clinic Full

Body with Head

,

Helen Hayes Lower Body

,

Helen Hayes Full Body

, and

Helen Hayes Full Body with Head

. It is also strongly recommended that you include the medial knee and ankle markers in the static pose, for more accurate calculation of knee and ankle joint centers. You can also supplement the OrthoTrak marker set with your own custom markers, as long as you do not move or remove any markers from the identified set. Lastly, the marker set used in the static trial must include all of the markers you plan to use for capturing motion. This is because the

Motion Module calculates the locations of all markers in the mocap model based on their locations in the static trial. These are the steps you should follow when collecting the static trial:

1.

2.

3.

Choose which of the six OrthoTrak marker sets you would like to use for capturing motion.

Add the medial knee and ankle markers, for better calculation of knee and ankle centers (not required, but highly recommended).

Add any additional markers that you would like to use (

e.g.

, extra markers on the feet, more markers on the arms). These markers must also be added to the mocap model, The Marker Set.

H-10

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

average from frame

These fields allow you to specify the starting and ending numbers for the sequence of frames that are averaged together to determine the static pose. These fields are initialized to 1 and the number of frames in the file, meaning that all frames will be averaged. If frames in the chosen sequence are missing some markers, locations for markers that are present will still be used in the average.

load personal.dat

This option gives you control over the automatic loading of

personal.dat.

When SIMM loads the static marker file, it looks for a file called

personal.dat

in the same folder. This file is identical to the one created and used by OrthoTrak. If the file is present, SIMM will automatically load it and read model parameters from it, such as foot length and hip origin offsets. It will use these parameters to determine joint center locations and segment lengths, using the same algorithms that OrthoTrak does. If there is no

personal.dat

file present in the folder, this option will be grayed out. If it is checked and you do not want to load

personal.dat

, click the box to turn it off.

read marker names from

For C3D import only:

This option allows you to choose from which parameter field in the C3D file to read the names of the tracked markers. Because the

POINT:LABELS

field in a C3D file is limited to four characters, some software packages (

e.g.,

EVaRT

) store the full marker name in the

POINT:DESCRIPTIONS

field. Since the marker names in the tracked file must exactly match the names used in the mocap model, if your C3D file does not contain full marker names in the

POINT:DESCRIPTIONS

field, you may have to edit the mocap model so that the marker names match the four-character names stored in the

POINT:LABELS

field.

subject mass

4.

5.

Capture the static trial using the protocol outlined in the OrthoTrak manual. The subject should have their arms either down by their sides, or straight out from their body with their thumbs facing forward.

Remove the medial knee and ankle markers, and any others that you do not want to use for capturing motion.

Note:

If you use a marker set with no upper extremity markers, the Motion

Module will remove the upper extremity from the mocap model and display only the pelvis and legs.

Once you have selected the static pose file to be used for opening the mocap model, SIMM displays a dialog box, allowing you to set some options for importing the static pose. In many cases, you will want to use the default settings for these options, so you can simply click the

OK

button to import the motion. The following list describes each option in the dialog box:

This field allows you to specify the total mass of the model after it has been scaled to fit the size of the subject. After the scaling is done, all the body segments' mass parameters are scaled up or down by a single percentage so that the total mass of the model equals the number entered into this field. This field has no effect if mass properties are not specified in the model file.

H-11

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

preserve mass distribution

This option gives you control over how the mass properties of the individual body segments are scaled. If this option is off, then each body segment's mass is scaled proportionally with its size. If this option is on, then each segment's mass parameters are not scaled with their change in size

(i.e., the distribution of body mass specified in the model file is preserved). In either case, after the model has been scaled, all the body segments' mass parameters are scaled up or down by a single percentage so that the total mass of the model equals the number entered into the "subject mass" field.

Figure H-3. Static Trial Import Dialog Box

Save JNT File, Save

MSL File

These options allows you to specify if SIMM will write out joint and muscle files containing the musculoskeletal model that is scaled to fit the subject. After SIMM has loaded the mocap model and scaled it based on the data in the static marker file and

personal.dat

, it will write out corresponding joint and muscle files, depending on the states of these check boxes. You may want to create these files so that you can make changes to them or to be able to re-load the model without going through the scaling process again.

H-12

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

Calculation of

Joint Centers

Scaling the

Mocap Model

Once the static pose has been loaded, the Motion Module recreates the

OrthoTrak skeletal model from the marker cloud. The first step is determining the locations of the joint centers for all of the joints in the

OrthoTrak model. The pelvis, hip, knee, and ankle centers are all found using the same procedure used by OrthoTrak.

The hip center is determined using percentage offsets from the pelvis markers. The Motion Module reads these offsets from

personal.dat

, as written by OrthoTrak. The default values for these offsets are taken from

Bell et al. Journal of Biomechanics, 23(6), 1990, pp. 617-21

: posterior displacement: 22% lateral displacement: 32% inferior displacement: 34%

To change these values, edit the file

personal.dat,

as described in Appendix D of the OrthoTrak manual.

The knee and ankle centers are found using the medial and lateral markers. It is strongly recommended that you use medial markers for a more accurate calculation of joint centers. If you choose not to use them, you should enter knee and ankle diameter measurements into

personal.dat

.

The Motion Module will use them to locate the knee and ankle centers if no medial markers are used.

The default method for determining shoulder, elbow, and wrist joint centers uses percentage offsets from the appropriate marker locations. If medial elbow and wrist markers are used in the static trial, their locations are averaged to get the joint centers, as is done with the knee and ankle. It is recommended that you use medial elbow and wrist markers in the static trial if you want an accurate representation of arm motion.

Once the locations of the OrthoTrak joint centers have been calculated from the static pose, the Motion Module determines the orientations of the OrthoTrak segment reference frames. It then can measure the lengths of the OrthoTrak segments and use them to scale the mocap model to match the size of the subject.

The reference frames for the foot, shank, thigh, pelvis, and torso are all determined using the procedure described in Appendix H of the

OrthoTrak manual.

OrthoTrak does not create reference frames for the upper and lower arms, but the Motion Module does this using one of several methods. If medial elbow and wrist markers are used in the static trial, then the arm reference frames are found in the same way in which the thigh frames are found. If no medial markers are present, then the upper and lower arm reference frames are found using the line between the joint centers as the X axis, and using the same Y axis as the torso. The Z axis is then determined by crossing X and Y.

Once all of the segment reference frames have been determined, the length of each segment is calculated. For most segments, the length is simply the distance from one joint center to the next. For the foot, the Motion Module reads the length from

personal.dat.

If there are no foot

H-13

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

The Marker Set

length measurements in

personal.dat

, then the foot length is assumed to be 1.4 times the distance from the heel marker to the toe marker (the toe marker is actually placed on the top of the foot just posterior to the toes).

Each body segment in the mocap model contains scaling information that tells the Motion Module how to scale it based on an OrthoTrak skeletal segment. The scaling information consists of the name of the OrthoTrak segment and X, Y, and Z reference numbers that correspond to the unscaled length of the SIMM segment. For example, the right femur in the mocap model contains the line: gait_scale R_THIGH 0.3960 0.3960 0.3960

This tells the Motion Module that the unscaled femur is 0.3960 meters long. Once the length of the corresponding OrthoTrak segment is known

(R_THIGH), the femur can be scaled accordingly. If the R_THIGH segment were 0.35 meters long, then the femur would be scaled by a factor of

0.35/0.396. In most cases the three reference values are the same number, indicating that the segment should be scaled uniformly in X, Y, and Z.

The two exceptions are the TORSO and PELVIS, which are scaled differently in two dimensions. For SIMM segments that do not map directly to an OrthoTrak segment, their scaling information is copied from the most relevant segment. For example, the right hand in the mocap model copies the scaling information from the right lower arm, so that the hand is scaled the same amount as the lower arm.

The marker set in the mocap model that comes with SIMM includes every marker used in all six marker sets that OrthoTrak recognizes, plus the medial knee and ankle markers. In addition, many other markers have been added, such as medial elbow and wrist markers. For a complete list of the markers in the model, as well as information on when they should be used and where they should be placed on the subject, read the

Guide to

Mocap Model Markers

document. The mocap model contains over 80 markers, which is more than the number used in most applications. When the static trial is loaded, any marker in the mocap model which is not in the static trial is removed from the model. Thus it is not a problem to have extra markers in the mocap model. In fact, you should add to the model whatever extra markers you may need for any of your motion capture applications. Then for a particular application the mocap model will have all the necessary markers, and the unused ones will automatically be removed when the model is loaded into SIMM.

To add or change markers in the mocap model, use the Marker Editor.

You should be careful not to overwrite the original

mocap.jnt

file. Instead, after editing the marker set, save the model to a new file name, and copy the altered markers into

mocap.jnt

.

All of the markers in the mocap model have X, Y, and Z offsets that put them in realistic locations given the dimensions of the generic model.

Thus if you load the unscaled mocap model into SIMM (using the

File >

Open

command, not the

File > Open Mocap Model

command, which will scale it), the markers will appear in positions corresponding to where they are placed on the subject.

These offsets are purely decorative, to help you view the marker set. They are not used by the Motion Mod-

H-14

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

ule to process any marker data

. To explain why this is so, we must first introduce the concept of

critical

markers and

non-critical

markers.

Critical markers are ones that must be present in the static trial in order for the Motion Module to load and scale the mocap model. For the lower body, these markers are: V.Sacral, R.ASIS, L.ASIS, R.Knee (or

R.Knee.Lateral), R.Ankle (or R.Ankle.Lateral), R.Heel, R.Toe, L.Knee

(or L.Knee.Lateral), L.Ankle (or L.Ankle.Lateral), L.Heel, and L.Toe. If any of these markers is missing from the static trial, the SIMM model of the lower body will not be loaded. For the upper body, the critical markers are: V.Sacral, R.ASIS, L.ASIS, R.Shoulder, R.Elbow, R.Wrist, L.Shoulder, L.Elbow, L.Wrist. Note that the ASIS and sacral markers are critical for both portions of the body. If one of these markers is missing from the static trial, you will get an error when trying to load the mocap model.

Non-critical markers are all other markers in the set.

Once the Motion Module has determined the locations of the joint centers and the orientations of the segment reference frames from the static pose, it calculates the proper offsets for all of the critical markers (plus the static-only medial markers). For example, once the right thigh reference frame has been oriented within the static pose marker cloud, the exact positions of the critical markers attached to the right thigh can be measured directly from the static pose and entered into the mocap model, thus overwriting whatever offsets were in the model input file.

After the offsets of all the critical markers have been determined in this fashion, the mocap model is “fit” to the static pose marker cloud using only the critical markers to find the best fit. This process orients the mocap model within the marker cloud, so that the offsets of the non-critical markers can be measured directly from the static pose. These offsets are then entered into the model, overwriting whatever values were in the model input file.

To summarize, the Motion Module uses a two-step process to calculate proper offsets for all of the markers in the mocap model. The first step determines the offsets of the critical markers, which the OrthoTrak algorithms can definitively locate without knowing anything about the mocap model. Then these critical markers are placed on the mocap model, and the model is fit to the static pose marker cloud. Now the offsets of the other markers can be measured, because every body segment in the mocap model is now correctly placed in the static pose.

H-15

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

Analog Configuration Files

SIMM can include analog data such as ground reaction forces, EMG activation levels, and kinetic data when importing a motion. SIMM uses a configuration file named

importVariables.txt

to determine which analog variables to import from an analog file, and how the data for each variable should be interpreted. This configuration file is used for both TRB/TRC import (with corresponding ANB/ANC analog files) and for C3D import

(where the analog data is contained in the C3D file itself). SIMM can interpret analog data as one of three types:

Forceplate Data

These variables specify voltages representing force or moment components as measured by a forceplate transducer. Given the voltages generated by a forceplate (6 channels for an AMTI or Bertec forceplate, 8 channels for a Kistler forceplate) SIMM can calculate and display a force location and vector for the forceplate.

EMG Data

These variables define activation levels for one or more muscles in the

SIMM full-body model. SIMM rectifies, smooths, and scales EMG data so that it can be plotted, and used to control the width and color of muscles during an animation.

Other Data

Any variables that are not forceplate or EMG data are classified as other data. SIMM does not perform any calculations on these data variables, but they may be included in SIMM plots.

importVariables.

txt

The

importVariables.txt

file, located in

SIMM\Resources\ mocap\misc

, contains a list of variable names and attributes. When SIMM processes an analog data file or an OrthoTrak XLS file, it consults the

importVariables.txt

file to decide which variables to import and how to interpret them.

Each row in

importVariables.txt

defines a variable to be imported. The first column in a row specifies the name of the variable as it appears in the analog or XLS data file. Since certain analog files support variable names with spaces in them, the first column of the

importVariables.txt

file

must

be terminated by a tab character. SIMM considers all characters from the beginning of a row until the first tab character to be the name of the import variable. SIMM does a case-insensitive comparison when matching variable names defined in

importVariables.txt

with variable names in an analog data file. Therefore the name “Rt Tibialis” would be considered the same as “rt tibialis”.

The second column in a variable definition specifies the type of the variable. It must be one of the following keywords: force_plate

, muscles

, or other_data

. These keywords must be lowercase. Following each keyword is information describing the variable:

This keyword specifies a ground reaction force variable. It must be followed by three values:

1.

2.

The forceplate number (

1

,

2

,

3

, etc.), then

The keywords force

or moment

, then

H-16

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

3.

The channel component ( x

, y

, or z

for AMTI or Bertec forceplates, or x12

, x34

, y14

, y23

, z1

, z2

, z3

, z4

for Kistler forceplates).

This keyword specifies an EMG variable. It must be followed by one or more SIMM muscle names. The keyword mvc

may optionally appear after the last muscle name. If mvc

appears, then it must be followed by an integer number that SIMM uses as the voltage for the maximum voluntary contraction when scaling that EMG channel. If no MVC value is specified, then the channel is scaled such that its maximum value is 1.0. EMG scaling is performed after the EMG channel's data has been smoothed and resampled to the motion's frequency.

This keyword specifies a data channel that exists simply to be included in

SIMM plots. This keyword may be optionally followed by a single word that will be used to label this channel in SIMM plots. If no name follows the other_data

keyword, then the name of the imported variable will be used.

forcepla.cal

When importing analog data from ANB/ANC files, SIMM uses the same calibration file as EVaRT and OrthoTrak for processing forceplate data.

Therefore, you can simply copy the

forcepla.cal

file from your EVaRT folder into the

Resources\mocap\misc

folder. For users who need to create a

forcepla.cal

file to describe their forceplate(s), refer to Appendix

C of the OrthoTrak manual.

Note: forcepla.cal

is not used for C3D import since C3D files contain the necessary calibration information for the force plates.

SIMM Motion Module

Guide to Mocap Model Markers

This guide describes the markers used by the Motion Module in SIMM to load each Mocap Model, scale it to fit the subject, and import recorded motions. For details on how the Motion Module processes the marker data and the model, see Chapter 5 of the SIMM User Guide. For a tutorial of the Motion Module, click on Help-> SIMM Tutorials -> Motion Module Demo in the SIMM menu bar. This document focuses on the names and locations of the markers, and when they are needed by the Motion

Module.

Definitions

Static Trial

A TRC, TRB, or C3D file of a motion capture subject in a static pose, usually the “T” or “scarecrow” pose

Motion Trial

A TRC, TRB, or C3D file of a subject performing an activity, such as walking or throwing

H-17

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

Mocap Model

A SIMM musculoskeletal model that can be loaded into SIMM, scaled to fit a subject using a static trial, and used to animate motion trials of that subject. The primary model is a full-body model with lower-extremity muscles, but others are available as well.

Critical Marker

A marker that is required in the static trial, and which must be placed in a specific location on the subject, according to instructions in the OrthoTrak manual. The coordinates of the marker in the static trial are used to determine joint centers and body segment lengths.

Semi-critical Marker

A marker that is optional in the static trial, but if used, must be placed in a specific location on the subject, according to instructions in the OrthoTrak manual. The coordinates of the marker in the static trial are used to improve the joint center calculations.

Optional Marker

A marker that is optional in the static trial, and whose placement on the subject does not need to be in a specific location

Fixed Marker

An optional marker whose X, Y, Z offsets are not automatically calculated when the static trial is processed. Rather, the offsets in the marker definition in the Mocap Model file are used to position the marker on the model

(these offsets are scaled with the body segment, however).

The Motion Module comes with four different Mocap Models for you to choose from. Each of them contains parameters that turn on and off different portions of the model, depending on which of the critical markers are present in the static trial. When you load a Model Model with a static trial, the Motion Module reads the list of markers from the trial and sets the values of the model parameters so that the appropriate portions are included. For example, if the critical markers on the right hand are present, then the degrees of freedom in the fingers are activated. If they are not present, the hand is modeled as one rigid body segment, with movement only at the wrist.

The Mocap Model that you will most likely want to use is

mocap.jnt

.

This is a model of a full body, with lower extremity muscles and [optionally] movable fingers in each hand. There is also a right arm model and a left arm model (

rightArm.jnt

and

leftArm.jnt

). These should be used if you want to capture motion of one arm without any torso or pelvis markers. Lastly,

mocap3D.jnt

is similar to

mocap.jnt

, but it includes 3D muscle surfaces for 18 key lower extremity muscles, rather than the lines of action for all 86 muscles. These muscle shapes look more realistic, but they do not have force-generating parameters, so you cannot calculate the lengths or forces in these muscles during the recorded motion.

H-18

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

Table H-1 on page H-19 shows the available combinations of model com-

ponents. To determine which Mocap Model you should use, find the row that best describes the model you want, then locate the filename in the last column. All of these files are located in

SIMM\Resources\mocap

. Once you h ave determ ined w hich o ne to use, y ou can either set the

MOCAP_MODEL variable in

SIMM\Resources\preferences

to that file, or choose that file using the

Options…Choose Model Model

command in the SIMM menu bar.

Table H-1. Combinations for Model Components lower extremity

yes yes no no no yes yes yes yes no no no

upper extremity

yes yes no yes yes right arm only right arm only left arm only left arm only yes yes no

movable fingers

no yes no yes yes no no yes no yes no no

muscles

legs only legs only legs only none none none none none none legs only, 3D legs only, 3D legs only, 3D

file name

mocap.jnt

mocap.jnt

mocap.jnt

mocap.jnt

mocap.jnt

rightArm.jnt

rightArm.jnt

leftArm.jnt

leftArm.jnt

mocap3D.jnt

mocap3D.jnt

mocap3D.jnt

It is important to note that the critical and semi-critical labels for markers

are relevant only for the static trial. For motion trials, all markers are op-

tional. That is, after recording the static trial, you can remove any of the markers from the subject before recording motion trials. Generally, however, you will want to keep all of the markers on the subject for the motion trials, with the possible exception of the medial joint markers. Also, once the static trial has been recorded, you must be careful not to move any of the markers on the subject (except for removing them completely).

SIMM uses the static trial to calculate the coordinates of each marker relative to its body segment, so if you move a marker or add additional markers, you must re-record the static trial and re-load the Mocap Model.

All of the markers described in this document are already part of the primary Mocap Model, located in SIMM\Resources\mocap\mocap.jnt. To use any of them, you do not need to make any changes to the file; just place the markers on the appropriate locations on the subject, and make sure the marker names in the static trial match the names shown in the figures below. Many of the markers can have one of several names, as listed

H-19

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

in the box pointing to each marker in the figures. These names are caseinsensitive, and may contain spaces.

If you want to add markers to the Mocap Model, you can do so with the

Marker Editor in SIMM. This tool allows you to create new markers, attach them to the appropriate body segments, and specify their X,Y, Z offsets. The exact values of the offsets are not important; they are used only for display of the marker while creating it. The offsets will be overwritten with values calculated by the Motion Module when the static trial is processed and the model is scaled to fit the subject. This process is described in more detail in Chapter 5 of the SIMM User Guide, but here is a brief summary. After loading the static trial, the Motion Module places all of the critical markers that are in the trial on the Mocap Model in their corresponding locations. The Mocap Model is then scaled to match the subject, and then a least-squares optimization fits the model within the cloud of static trial markers, considering only the critical markers. This positions the model within the marker cloud so that the Motion Module can then directly calculate the offsets from the optional markers to the model segments to which they are attached. If you do not want the offsets for a marker to be calculated in this manner, then you must turn on the “fixed” button for that marker in the Marker Editor, and enter accurate X, Y, Z offsets into the number fields. This tells the Motion Module to scale the marker’s offsets when the model is scaled, but not to recalculate their values as it does for the other optional markers.

Note on adding markers: You can create new markers using the Marker

Editor, and then save the model by writing out a joint file, but you should n o t r e p l a c e t h e o r i g i n a l m o d e l f i l e ( e . g . ,

S I M M \ R e sources\mocap\mocap.jnt

) with this new file. This is because the model file contains many comments and special parameters that enable SIMM to automatically modify it for a particular static trial, as described above.

However, when this file is loaded into SIMM and then written back out, these comments and parameters are lost. Thus after saving your new joint file, you should use a text editor to copy the new marker definitions from the file and paste them into the existing model file.

Shown in Figure H-4 on page H-21 are the critical and semi-critical mark-

ers for upper body and lower body motion recording. If any of the lower body critical markers is missing from the static trial, the legs will not be loaded with the Mocap Model. Similarly, if any of the upper body critical markers is missing from the static trial, the torso, head, and arms will not be loaded. Note that the sacral, left ASIS, and right ASIS markers are critical for both upper and lower body motion recording. If any of these markers is missing, the Motion Module will print an error and not load the

Mocap Model. The head and hand markers are semi-critical. If used, they allow the Motion Module to track motion at the neck and wrist. If not used, these joints will remain fixed during animation of motion trials in

SIMM.

H-20

EVaRT 5.0 User’s Manual

Figure H-4. Critical and Semi-Critical Markers

Appendix H: SIMM Motion Module

H-21

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

The markers shown in

Figure H-5 are optional. If any of these markers is

in the static trial, its location on the corresponding body segment in the

Mocap Model will automatically be determined after the model has been scaled using the critical markers (i.e., these optional markers are not

“fixed,” so their X, Y, Z offsets in the model file will be overwritten when the model is loaded). These markers will then be used to help solve the frames of data in a motion trial.

Figure H-5. Optional Markers

H-22

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

The markers shown in

Figure H-6 on page H-24

are used by the Motion

Module to control the degrees of freedom in the hand. If the three critical markers are present in the static trial, the Motion Module will load a detailed model of the hand with three joints in each finger. By default, all of the finger joints are fixed. SIMM converts them into hinge joints as it detects the presence of markers to control the joints. For example, if

R.Finger2.M1, R.Finger2.M2, and R.Finger2.M3 are all present, SIMM will create three hinge joints in the index finger, each with its own degree of freedom. If only R.Finger2.M1 is present, SIMM will create the proximal finger joint with a degree of freedom, and make the two distal joints dependent on the proximal one (so that all three joints will flex when the proximal one does). Any combination of the optional markers can be used to create a hand model with the desired degrees of freedom. All of the optional hand markers are defined as “fixed” in the model file. This means that the offsets specified in the file are used for solving motions (the Motion Module does not overwrite them), and thus you should place the markers on the subject according to how they are shown in the figure below.

H-23

Appendix H: SIMM Motion Module

Figure H-6. Critical and Optional Markers for Hands

EVaRT 5.0 User’s Manual

H-24

Starting with SIMM 4.0, support has been added for alternative critical marker sets for use with the Mocap Model. For example, the sacral marker can be replaced with two PSIS markers, and the lateral wrist marker can be replaced with the radius marker. It is thus difficult to display in a single picture of the body the complete set of markers that are required. On the following pages are descriptions of the critical and semicritical marker sets for each portion of the body. Also, for each marker, the complete list of acceptable names is shown. Any one of these case-in-

sensitive names in the list can be used to identify the marker in the

EVaRT

project.

EVaRT 5.0 User’s Manual Appendix H: SIMM Motion Module

Lower Body

Critical Markers

Semi-critical

Markers

The lower body portion of the Mocap Model will be loaded if the critical markers listed below are present in the static trial. The thigh, shank, and feet segments will each be scaled separately, based on measurements made from the static trial. Each of these segments will be scaled uniformly in the X, Y, and Z dimensions. The pelvis segment will be scaled independently in the X, Y, and Z dimensions. It is not possible to load only one leg of the Mocap Model.

1.

Right ASIS.

acceptable names:

R.ASIS RASIS RASI

2.

Left ASIS.

acceptable names:

L.ASIS LASIS LASI

3.

Posterior pelvis: a. Sacrum.

acceptable names

: V.SACRAL V.SACRUM SACRAL

SACRUM SACR VSAC or

b. Right PSIS.

acceptable names:

R.PSIS RPSIS RPSI and

Left PSIS.

acceptable names:

L.PSIS LPSIS LPSI

4.

Right lateral knee

. acceptable names

: R.KNEE R.KNEE.LAT-

ERAL R.KNEE.LAT RKNE

5.

Left lateral knee

. acceptable names:

L.KNEE L.KNEE.LATERAL

L.KNEE.LAT LKNE

6.

Right lateral ankle.

acceptable names:

R.ANKLE

R.ANKLE.LATERAL R.ANKLE.LAT RANK

7.

Left lateral ankle

. acceptable names

: L.ANKLE L.ANKLE.LAT-

ERAL L.ANKLE.LAT LANK

8.

Right heel.

acceptable names:

R.HEEL RHEE

9.

Left heel.

acceptable names:

L.HEEL LHEE

10. Right toe.

acceptable names:

R.TOE RTOE

11. Left toe

. acceptable names:

L.TOE LTOE

1.

Right medial knee.

acceptable names:

R.KNEE.MEDIAL

R.KNEE.MED

2.

Left medial knee.

acceptable names:

L.KNEE.MEDIAL

L.KNEE.MED

3.

Right medial ankle.

acceptable names:

R.ANKLE.MEDIAL

R.ANKLE.MED

4.

Left medial ankle.

acceptable names:

L.ANKLE.MEDIAL

L.ANKLE.MED

H-25

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

Upper Body

Critical Markers

Semi-critical

Markers

The upper body portion of the Mocap Model will be loaded if the critical markers listed below are present in the static trial. The upper arm and lower arm segments will each be scaled separately, based on measurements made from the static trial. Each of these segments will be scaled uniformly in the X, Y, and Z dimensions. The torso segment will be scaled independently in two dimensions (the X is scaled the same as the Z). It is not possible to load the upper body with only one arm. To load only one arm (without the rest of the upper body), use the SIMM file rightArm.jnt

or leftArm.jnt as the Mocap Model.

1.

Right ASIS.

acceptable names:

R.ASIS RASIS RASI

2.

Left ASIS.

acceptable names:

L.ASIS LASIS LASI

3.

Posterior pelvis: a. Sacrum

. acceptable names:

V.SACRAL V.SACRUM SACRAL

SACRUM SACR VSAC or

Right PSIS

. acceptable names:

R.PSIS RPSIS RPSI and

Left PSIS.

acceptable names:

L.PSIS LPSIS LPSI

4.

Right shoulder

. acceptable names:

R.SHOULDER RSHO

5.

Left shoulder

. acceptable names:

L.SHOULDER LSHO

6.

Right lateral elbow

. acceptable names

: R.ELBOW

R.ELBOW.LATERAL R.ELBOW.LAT RELB

7.

Left lateral elbow.

acceptable names

: L.ELBOW L.ELBOW.LAT-

ERAL L.ELBOW.LAT LELB

8.

Right wrist

:

a. Lateral

. acceptable names:

R.WRIST R.WRIST.LATERAL

R.WRIST.LAT RWRI or

b. Radius

. acceptable names:

R.RADIUS RWRA

9.

Left wrist

:

a. Lateral

. acceptable names:

L.WRIST L.WRIST.LATERAL

L.WRIST.LAT LWRI or

b. radius

. acceptable names:

L.RADIUS LWRA

1.

Right medial elbow.

acceptable names:

R.ELBOW.MEDIAL

R.ELBOW.MED

2.

Left medial elbow

. acceptable names:

L.ELBOW.MEDIAL

L.ELBOW.MED

3.

Right wrist

:

a. Medial

. acceptable names:

R.WRIST.MEDIAL R.WRIST.MED

H-26

EVaRT 5.0 User’s Manual

Right Arm

Critical Markers

Semi-critical

Markers

Left Arm

Critical Markers

Appendix H: SIMM Motion Module

or

b. ulna

. acceptable names:

R.ULNA RWRB

4.

Left wrist

:

a. medial.

acceptable names:

L.WRIST.MEDIAL L.WRIST.MED

or

b. ulna

. acceptable names:

L.ULNA LWRB

To load only the right arm, set the MOCAP_MODEL parameter in your

SIMM preferences file to rightArm.jnt, or choose that file using the O

ptions…Choose Model Model

command in the SIMM menu bar. Then use the markers listed below.

1.

Right shoulder

. acceptable names:

R.SHOULDER RSHO

2.

Right lateral elbow.

acceptable names

: R.ELBOW

R.ELBOW.LATERAL R.ELBOW.LAT RELB

3.

Right wrist: a. Lateral.

acceptable names:

R.WRIST R.WRIST.LATERAL

R.WRIST.LAT RWRI or

b. Radius

. acceptable names:

R.RADIUS RWRA

1.

Right medial elbow

. acceptable names:

R.ELBOW.MEDIAL

R.ELBOW.MED

2.

Right wrist

:

a. Medial.

acceptable names:

R.WRIST.MEDIAL R.WRIST.MED

or

b. Ulna

. acceptable names:

R.ULNA RWRB

To load only the left arm, set the MOCAP_MODEL parameter in your

SIMM preferences file to leftArm.jnt, or choose that file using the Op-

tions…Choose Model Model command in the SIMM menu bar. Then use the markers listed below.

1.

Left shoulder

. acceptable names:

L.SHOULDER LSHO

2.

Left lateral elbow

. acceptable names

: L.ELBOW L.ELBOW.LAT-

ERAL L.ELBOW.LAT LELB

3.

Left wrist

:

a. Lateral.

acceptable names:

L.WRIST L.WRIST.LATERAL

L.WRIST.LAT LWRI or

a. Radius

. acceptable names:

L.RADIUS LWRA

H-27

Appendix H: SIMM Motion Module

Semi-critical

Markers

Right Hand

Critical Markers

Semi-critical

Markers

Left Hand

Critical Markers

1.

Left medial elbow.

L.ELBOW.MED

EVaRT 5.0 User’s Manual

acceptable names:

L.ELBOW.MEDIAL

2.

Left wrist

:

a. Medial

. acceptable names:

L.WRIST.MEDIAL L.WRIST.MED

or

b. Ulna

. acceptable names:

L.ULNA LWRB

The right hand will always be included when the right arm is loaded, even if there are no markers on the hand. The presence of critical markers controls how the hand is scaled and what degrees of freedom it has. The right hand will be scaled separately from the right lower arm if the three critical markers listed below are present in the static trial. The individual finger gencoords will be added to the model if the three critical hand markers and the appropriate finger markers are present in the static trial.

1.

Right thumb.

acceptable names:

R.THUMB R.THUMB.M3

2.

Right middle finger.

acceptable names:

R.MIDDLE.FINGER

R.FINGER R.FINGER3.M3

3.

Right wrist: a. Lateral.

acceptable names:

R.WRIST R.WRIST.LATERAL

R.WRIST.LAT RWRI or

b. Radius.

acceptable names:

R.RADIUS RWRA

1.

Right wrist

:

a. medial.

acceptable names:

R.WRIST.MEDIAL R.WRIST.MED

or

b. Ulna.

acceptable names:

R.ULNA RWRB

The left hand will always be included when the left arm is loaded, even if there are no markers on the hand. The presence of critical markers controls how the hand is scaled and what degrees of freedom it has. The left hand will be scaled separately from the left lower arm if the three critical markers listed below are present in the static trial. The individual finger gencoords will be added to the model if the three critical hand markers and the appropriate finger markers are present in the static trial.

1.

Left thumb

. acceptable names:

L.THUMB L.THUMB.M3

2.

Left middle finger

. acceptable names:

L.MIDDLE.FINGER

L.FINGER L.FINGER3.M3

3.

Left wrist

:

a. Lateral.

acceptable names:

L.WRIST L.WRIST.LATERAL

L.WRIST.LAT LWRI or

H-28

EVaRT 5.0 User’s Manual

Semi-critical

Markers

Head

Critical Markers

Appendix H: SIMM Motion Module b. Radius

. acceptable names:

L.RADIUS LWRA

1.

Left wrist

:

a. Medial

. acceptable names:

L.WRIST.MEDIAL L.WRIST.MED

or

b. Ulna

. acceptable names:

L.ULNA LWRBb

The head will always be included when the upper body is loaded, and the neck will contain three degrees of freedom. If the critical markers listed below are present in the static trial, the head will be scaled separately from the torso. Otherwise, the head will be scaled uniformly by the scale factor used for the Y (height) of the torso. If no markers (critical or optional) are included on the head in the static trial, then the degrees of freedom in the neck will remain fixed during imported motions.

1.

Rear of head.

acceptable names:

HEAD.REAR REAR.HEAD

HEADREAR REARHEAD

2.

Top of head.

acceptable names:

HEAD.TOP TOP.HEAD HEAD-

TOP TOPHEAD

3.

Front of head

. acceptable names

: HEAD.FRONT FRONT.HEAD

HEADFRONT FRONTHEAD

H-29

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

The following optional markers are already defined in

mocap.jnt

. To use them, just put their exact names in your

EVaRT

project:

pelvis: R.Trochanter

L.Trochanter

thorax: Offset

Sternum

T10

CLAV

STRN

RBAK

cerv7: C7

C7 Spinous Process

head: R.Ear

L.Ear

RBHD

RFHD

LBHD

LFHD

HEDO

HEDP

HEDA

HEDL

H-30

EVaRT 5.0 User’s Manual

clavicle_l: L.Clavicle

scapula_l: L.Scapula

L.Scapula.Top

L.Scapula.Bottom

L.Angulus Acromialis

L.Trigonum Spinae

L.Angulus Inferior

humerus_l: L.Bicep

L.Biceps.Lateral

ulna_l: L.Forearm

femur_l: L.Thigh

L.Thigh.Upper

L.Thigh.Front

L.Thigh.Rear

LTHI

tibia_l: L.Shank

L.Shank.Upper

L.Shank.Front

L.Shank.Rear

LTIB

foot_l: L.MedFoot

L.LatFoot

Appendix H: SIMM Motion Module

clavicle_r: R.Clavicle

scapula_r: R.Scapula

R.Scapula.Top

R.Scapula.Bottom

R.Angulus Acromialis

R.Trigonum Spinae

R.Angulus Inferior

humerus_r: R.Bicep

R.Biceps.Lateral

ulna_r: L.Forearm

femur_r: R.Thigh

R.Thigh.Upper

R.Thigh.Front

R.Thigh.Rear

RTHI

tibia_r: R.Shank

R.Shank.Upper

R.Shank.Front

R.Shank.Rear

RTIB

foot_r: R.MedFoot

R.LatFoot

H-31

Appendix H: SIMM Motion Module EVaRT 5.0 User’s Manual

H-32

Appendix I

Synchronizing Digital Video with EVaRT

Topic

EVaDV Overview

System Requirements

Installation

Using EVaDV

Currently Open Known Issues

Page

I-1

I-1

I-2

I-2

I-3

EVaDV Overview

EVaDV

is a Digital Video (DV) capture application for use with Motion

Analysis Corporation's

EVaRT

software for the synchronized capture of color video data on a separate Windows PC as

*.avi

files. You can directly transfer digital information back and forth between a DV camcorder and your computer with the use of the IEEE 1394 standard, also known as a

Firewire or i.Link connection. If your computer does not come with this interface built into it, you will need to purchase an inexpensive card that provides the correct port.

You can run

EVaDV

on your local machine that has

EVaRT

running on it

(not recommended) as well as with remote machines that are connected to digital video cameras.

Note:

Any standard DV camcorder should be sufficient for use with the

EVaDV

software. All

EVaDV

testing and product development was done with a

Sony TM DCR-TRV520 NTSC model DV camcorder.

System Requirements

Recommended

Minimum

Specifications

Microsoft Windows 2000

or

XP

256 MB RAM

80GB (or more) free hard drive space (for captured files)

CD-RW drive

Ethernet card

I-1

Appendix I: Synchronizing Digital Video with EVaRT EVaRT 5.0 User’s Manual

Installation

We recommend you install the software into the

C:\Program Files\Motion Analysis\EVaDV

directory, but

EVaDV.exe

will run from any folder.

Note:

No dongle or license file is required to run this application. But to collect synchronized color video (AVI files) in

EVaRT

, you need the [Reference

Video 3.0] line in your mac_lic.dat license file.

Using EVaDV

Capturing Digital

Video Using

EVaDV

1.

2.

3.

4.

5.

6.

Launch

EVaDV.

Select the desired camera/capture device from Camera dropdown. If there is a single video camera connected to the host system,

EVaDV

connects to this camera automatically.

Select a Capture Folder to indicate where captured files should be stored.

Select a Capture File to indicate the name of the file that will be created.

To begin recording, press the

Record

button (red circle). A red

RECORD will display next to the player control panel indicating the system is recording, and a message will be displayed in the Message

Bar indicating record start time.

To stop recording, press the

Stop

button. The red RECORD will disappear and a message will be displayed in the Message Bar indicating recorder filename, stop time, and any available stats.

Capturing Digital

Video in EVaDV from EVaRT

Note:

[EVaXX]

indicates the host machine on which to perform the action.

1.

Ensure the

EVaDV

host and the

EVaRT

host are connected to each other via a TCP/IP network connection.

2.

[EVaDV]

Launch

EVaDV

on the video capture host.

3.

[EVaRT]

Launch

EVaRT

on the

EVaRT

host.

Note—

launch order is unimportant, however if

EVaRT

grabs the camera/capture device first (i.e. displays color video window)

EVaDV

will not have access to the camera.

EVaRT

will grab the camera when you press the

F1

key in a capture window and release it when you select another key (

F2, F3

, etc.).

4.

[EVaDV]

Select the desired camera/capture device from Camera dropdown. If there is a single video camera connected to the host system,

EVaDV

connects to this camera automatically.

5.

[EVaRT]

Activate the Color Video (*.avi) checkbox in the

Motion

Capture > Output

sub-panel. Press the

F1

key to show video and grab the DV camera.

6.

[EVaRT]

Press the

Record

button.

I-2

EVaRT 5.0 User’s Manual Appendix I: Synchronizing Digital Video with EVaRT

7.

[EVaDV]

On Record,

EVaDV

will record the file specified by

EVaRT

to the directory specified by

EVaRT

or, if not present, to the local

Capture Folder. The Message Bar indicates the directory requested by

EVaRT

.

8.

[EVaDV]

On Record, A red "RECORD" will display next to the player control panel indicating system is recording, and a message will be displayed in the Message Bar indicating record start time.

9.

[EVaRT]

Stop recording by pressing either the

Stop

button or after reaching the duration specified.

Note:

If you have more than one computer running

EVaRT

on the same network, the

EVaDV

software does not consume the plugin port (as does the streaming Alias/Motionbuilder online plugin, for example). The message about starting and ending the data capture is broadcast to the x.x.x.255 address which means that all computers connected can hear the message and start and stop the recording. So you should be able to connect as many

EVaDV

recorder-computers up as you need.

Note:

You probably will not want the AVI files streaming across the network.

Note:

When VC files are collected, AVI files are collected. If VC files are not collected, AVI files are not collected.

Note:

You must have the color video window (F1 function) open in the

EVaRT

interface to record AVI files.

Suggestion for

DV Camera

Setup

Sony

TM

DV cameras have a setup mode called

FRAME/FIELD

. The

FRAME

setup mode works best as it eliminates the “motion blur” which results from the

FIELD

(also known as Interleaved) mode. On Canon

TM

DV cameras, you will want to select

MOVIE

mode.

Currently Open Known Issues

Interface is evolving.

Resizing video window by dragging corner can cause a hang-up.

Stop, resize, play seems to work around this.

File recompression functionality unstable with certain compressors.

Multiple camera support is working, however certain cameras (Sony) seem to dislike it, and will give a "fail on run" message.

Snapshot .

tga

image is incorrect.

Only Type-1 (interleaved) AVI files are supported (import/export).

Known Graphic

Card Issues

There is a known problem displaying the AVI files when a 3D window is displayed on a Windows XP system. The AVI file plays in a jerky motion and the screen is sometimes sliced into horizontal blocks.

This is a problem with Open GL and Direct X displaying at the same time. Right-click on your desktop, then select

Properties > Settings >

Advanced > Troubleshoot

. Slide the Hardware Acceleration down a few points. Quit

EVaRT

, then re-launch. See if that fixes the problem. The settings depend on what your graphics card does and has for features that

I-3

Appendix I: Synchronizing Digital Video with EVaRT EVaRT 5.0 User’s Manual

changes from one computer to another. If that does not fix the problem, try to slide the Hardware Acceleration to

None

. Quit

EVaRT

and then relaunch. If that still does not fix the problem, try to change some of the

Open GL settings found under the name of your graphics card (found under the

Properties > Setting > Advanced

tab). Also, change the Vertical Sync setting to

On by Default

and restart your computer.

I-4

Appendix J

Using EVaRT with Jack

Software

Topic

Introduction

Installation Instructions

Jack – Using the Motion Analysis Calcium Tracker Module

Jack5.0a Required & Recommended Marker Sets in EVaRT for Calcium Skeleton Generation

Page

J-1

J-2

J-2

J-5

Introduction

Jack is an ergonomics and human factors product used in various industries to improve the ergonomics of product designs and workplace tasks.

This software enables users to position biomechanically-accurate digital humans of various sizes in virtual environments, assign them tasks and analyze their performance. Jack (and Jill) digital humans can tell engineers what they can see and reach, how comfortable they are, when and why they're getting hurt, when they're getting tired and other important ergonomics information. This information helps organizations design safer and more effective products faster and for less cost. Ultimately, Jack helps companies bring factories on-line faster and optimize productivity while improving worker safety. For more information, please visit the UGS website at www.ugs.com.

Before You Start

Jack 5.0a does not need a new license file. It uses the same license as Jack

4.2. The license is based on your Ethernet card's Physical Address. To find it:

1.

2.

3.

4.

Select

Start > Run

. This command. brings up a DOS command window.

Type ipconfig /all

Find the line that looks like:

Physical Address . . . . .

000-50-DA-8E-BC-BD

<<--

Note yours and then contact UGS.

Installation of the Jack5.0a software requires very specific installation as per the instructions in the next section. If you deviate from the settings specified it will not launch.

J-1

Appendix J: Using EVaRT with Jack Software EVaRT 5.0 User’s Manual

The .jk.tcl file will not easily copy itself to a native windows environment. This will be fixed for the release version of Jack 5.0. The .jk.tcl acts as a launching tool for the Calcium plugin window. Currently, copying this file from a web-based browser (like Hotmail or Gmail) will allow you to save it as a .jk.tcl file (MS Windows thinks it doesn’t have a filename).

If you are using an MS product or Novell Webmail, it will try to rename it to .jk.tcl[1] and you cannot rename it. If you open the .jk.zip folder and extract the file, it will copy itself correctly, without appending the [1] to the end of the filename.

Installation Instructions

Note:

1.

Download the latest Jack-5.0alpha from ftp://specialdevftp.eai.com/private/jack_dist/FordDW/Jim/Jack50a

2.

3.

4.

Unzip each zip file to the

C:\Jack50a

folder.

Copy your Jack license file to the

C:\Jack50a\license

folder.

Copy the attached jk.tcl file to the

C:\Documents and Settings\<your_login_id>

folder.

If you have a HOME environment variable set on your machine, copy the .jk.tcl file to your HOME folder.

5.

To start Jack, double click on the

jack50.bat

file in the

C:\Jack50a

folder.

For debugging, you can use the

Start > Run…

command in a

DOS window. Then cd \Jack50a and run the batch file:

jack50.bat.

This gives more information.

Jack – Using the Motion Analysis Calcium Tracker Module

Loading

CalciumTracker module in Jack

-OR-

1.

2.

3.

4.

5.

Start Jack.

Click

Modules | Plug-ins..

.

In the Add-On Modules Dialog, select the

CalciumTracker

module.

Click

Load

and

OK.

Click

Modules | CalciumTracker | MotionAnalysis.

1.

2.

Start Jack (auto loading of Calcium Tracker occurs via the .jk.tcl

script).

[Optional]

If you are using .jk.tcl file, click

Modules | CalciumTracker Dialog.

Connecting to

MotionAnalysis

1.

2.

3.

In the

MotionAnalysis-Calcium

dialog, click on the

Devices

tab.

Enter the

EVaRT

or localhost computer name as the Host name, and then click

Connect

.

Check/Uncheck

Display Bodies

to turn on/off the visible bodies in

Jack.

Set the Origin Rotation Check box to

ON

, select

X Rotation –90

,

Z

Rotation –90

.

J-2

EVaRT 5.0 User’s Manual Appendix J: Using EVaRT with Jack Software

Auto-Scaling,

Constraining and Positioning the Subject

1.

Create a male or female Jack figure.

This may be done by selecting either of the following buttons.

Figure J-1. Male or Female Selection Button

2.

Click the

pick

button and select the subject from the scene.

The subject turns yellow when it is picked. Left-click on the subject.

The name “Human” should fill the blank box beside the pick icon.

3.

4.

5.

6.

Click on the

Subject

tab.

Have the actual subject in standing straight posture and click

Auto

Scale.

Click

Constrain.

Click

Move

. Using the mouse in the Jack window, move the

MA_ORIGIN figure to position the subject.

Creating Two

Channel Eye

View Windows

(First Person)

[OPTIONAL]

1.

2.

3.

In the MotionAnalysis-Calcium dialog, click the

pick

button and select the subject from the scene.

Click on the

Eye View

tab. Click

Apply

(minimize all other windows). Note that the main Jack window is disabled. To enable, right middle button on the TJ_Window and uncheck

Disabled

.

[Optional]

You can change eye view window parameters.

Setting Up

Collision

Detection

[OPTIONAL]

1.

2.

In the MotionAnalysis-Calcium dialog, click the

pick

button and select the subject from the scene.

Click on the

Collision

tab. Check the required segments and select an algorithm for the collision check. Then click

Apply.

Note:

You should have a scene loaded before you setup the collision detection.

J-3

Appendix J: Using EVaRT with Jack Software

List of Segments

Required by

Jack

Figure J-2. Jack Interface

8.

9.

10.

11.

12.

13.

4.

5.

6.

7.

1.

2.

3.

Root

Head

Neck

Spine1/Spine2/Spine3/Spine4

LClavicle/RClavicle

LUpperArm/RUpperArm

LLowerArm/RLowerArm

LHand/RHand

LHip/RHip

LUpperLeg/RUpperLeg

LLowerLeg/RLowerLeg

LFoot/RFoot

LToes/RToes

EVaRT 5.0 User’s Manual

J-4

EVaRT 5.0 User’s Manual Appendix J: Using EVaRT with Jack Software

Jack5.0a Required & Recommended Marker Sets in EVaRT for Calcium Skeleton Generation

Table J-1. Required markers for Jack 5 SIMM-OrthoTrak Model: 22 (absolute minimum)

Markers

R.Shoulder, L.Shoulder

R.Elbow, L.Elbow

Location

markers on top of shoulder above shoulder joint lateral side of elbow, on elbow hinge axis. On top of elbow with arms in T-pose position

Wrist has 2 options: R.Radius, L.Radius REC-

OMMENDED (on distal radius, thumb side of your hand), and R.Ulna, L.Ulna (on distal ulna, pinky side of your hand)

-and/or-

R.Wrist, L.Wrist: top of wrist

R.ASIS and L.ASIS

L.BackOffset

Back of pelvis has 2 options:

R.PSIS and L.PSIS: RECOMMENDED Posterior

Superior Iliac Spine

-and/or-

V.Sacral (All 5 is RECOMMENDED:

ASIS, PSIS, V.Sacral)

Anterior Superior Iliac Spine offset marker for asymmetry

R.Knee and L.Knee

R.Ankle and L.Ankle

R.Heel and L.Heel

R.Toe and L.Toe

lateral knee, close to knee axis lateral ankle, on fibular malleolus heel, at same height as toe markers toe markers, center of foot at proximal base of toe joint

J-5

Appendix J: Using EVaRT with Jack Software EVaRT 5.0 User’s Manual

The following markers are a recommended marker set only. These incorporate the above markers, include some additional markers, and match the data set Jack5_41Markers_AutoScale.prj in the Sample Data files.

29.

30.

31.

32.

25.

26.

27.

28.

21.

22.

23.

24.

17.

18.

19.

20.

37.

38.

39.

40.

41.

33.

34.

35.

36.

13.

14.

15.

16.

9.

10.

11.

12.

5.

6.

7.

8.

1.

2.

3.

4.

R.Heel

R.Toe

R.Foot

L.Thigh

L.Knee

L.Shank

L.Ankle

L.Toe

L.Heel

L.Foot

L.Bicep

L.Elbow

L.Forearm

L.Radius

L.Ulna

L.Thumb

L.Pinky

R.ASIS

L.ASIS

R.PSIS

L.PSIS

V.Sacral

R.Thigh

R.Knee

R.Shank

R.Ankle

Top.head

Back.Head

Front.Head

L.Head_Offset

R.Shoulder

L.Shoulder

Neck

L.BackOffset

R.Bicep

R.Elbow

R.ForeArm

R.Radius

R.Ulna

R.Thumb

R.Pinky

J-6

EVaRT 5.0 User’s Manual

Jack 5 Marker Set

Figure J-3. Front View

1

3

4

Appendix J: Using EVaRT with Jack Software

5

6

9

23

12

13

28

16

24

35

20

19

29

36

30

37

31

34

33 40

38

41

1 - Top.Head

3 - Front.Head

4 - L.Head_Offset

5 - R.Shoulder

6 - L.Shoulder

9 - R.Bicep

12 - R.Radius

13 - R.Ulna

16 - L.Bicep

19 - L.Radius

20 - L.Ulna

23 - R.Asis

24 - L.Asis

28 - R.Thigh

29 - R.Knee

30 - R.Shank

31 - R.Ankle

33 - R.Toe

34 - R.Foot

35 - L.Thigh

36 - L.Knee

37 - L.Shank

38 - L.Ankle

40 - L.Toe

41 - L.Foot

J-7

Appendix J: Using EVaRT with Jack Software

Figure J-4. Right Side View

1

2 3

7

6

9

11

10

12

14

13

15

EVaRT 5.0 User’s Manual

1 - Top.Head

2 - Back.Head

3 - Front.Head

6 - L.Shoulder

7 - Neck

9 - R.Bicep

10 - R.Elbow

11 - R.Forearm

12 - R.Radius

13 - R.Ulna

14 - R.Thumb

15 - R.Pinky

25 23

28

29

30

32

31

34

33

23 - R.Asis

25 - R.Psis

28 - R.Thigh

29 - R.Knee

30 - R.Shank

31 - R.Ankle

32 - R.Heel

33 - R.Toe

34 - R.Foot

J-8

EVaRT 5.0 User’s Manual

Figure J-5. Left Side View

Appendix J: Using EVaRT with Jack Software

21

19

22

20

18

17

3

16

1

2

4

5

7

8

1 - Top.Head

2 - Back.Head

3 - Front.Head

4 - L.Head_Offset

5 - R.Shoulder

7 - Neck

8 - L.Back_Offset

24

26

16 - L.Bicep

17 - L.Elbow

18 - L.Forearm

19 - L.Radius

20 - L.Ulna

21 - L.Thumb

22 - L.Pinky

24 - L.Asis

26 - L.Psis

35

36

37

41

40

38

39

35 - L.Thigh

36 - L.Knee

37 - L.Shank

38 - L.Ankle

39 - L.Heel

40 - L.Toe

41 - L.Foot

J-9

Appendix J: Using EVaRT with Jack Software

Figure 16. Rear View

4

1

2

EVaRT 5.0 User’s Manual

6 7

5

8

16

9

18

21

19

22

20

17

26

27

25

10

11

13

15

12

14

36

29

1 - Top.Head

2 - Back.Head

4 - L.Head_Offset

5 - R.Shoulder

6 - L.Shoulder

7 - Neck

8 - L.Back_Offset

9 - R.Bicep

10 - R.Elbow

11 - R.Forearm

12 - R.Radius

13 - R.Ulna

14 - R.Thumb

15 - R.Pinky

16 - L.Bicep

17 - L.Elbow

18 - L.Forearm

19 - L.Radius

20 - L.Ulna

21 - L.Thumb

22 - L.Pinky

25 - R.Psis

26 - L.Psis

27 - V.Sacral

29 - R.Knee

31 - R.Ankle

32 - R.Heel

33 - R.Toe

34 - R.Foot

36 - L.Knee

38 - L.Ankle

39 - L.Heel

40 - L.Toe

41 - L.Foot

41

40

38

39

31

32

33

34

J-10

Appendix K

Questions and Answers for

Specific Applications

Question

Answer

Question

Answer

Question

Answer

Does

EVaRT

require a specific order for markers and linkages?

For the identification of markers to work swiftly in real-time the order of marker definitions is important. You should follow these rules.

1.

2.

Markers should be ordered such that each successive marker builds the character top to bottom through linkages (i.e. Head to Neck, down one arm, then down the other, down the torso to the hips, down one leg, then the other). Do not backtrack.

If the first markers are linked into a stiff triangle, marker identification will be swift. For this reason, the head markers should always be first. Linkage order may affect the rectify process. For the head, a linkage order of 1-2, 1-3, and 2-3 works well.

How can I prevent ghost markers from appearing?

1.

2.

3.

Ghost markers may appear if the Max. Residual value is set too low.

This parameter is set in the tracking function in the Motion Capture mode (except for centroid function instead of tracking).

Set the minimum number of cameras to 3.

Increase the minimum number of lines per marker

How do I control the length of a recorded file in

EVaRT

?

All the recording options are set in the Output function in the

Motion Capture mode.

When the record button is clicked using the trigger or the mouse, a new file is recorded and saved with the name, directory, and output type(s) you have specified. The recording will stop when either:

• the duration in seconds is reached

• the

Stop

button is clicked using the trigger or mouse

The default duration is 60 seconds. If you always want to control the end of the take with the trigger or mouse, we recommend setting the duration to a number that is higher than the trials you usually capture such as 10 minutes (600 seconds).

K-1

Appendix K: Questions and Answers for Specific Applications EVaRT 5.0 User’s Manual

Question

Answer

Can

EVaRT

be used in a large capture volume (for example 50’x50’, 50 cameras)?

If care is taken during setup, motion capture will work well in a large volume. The four areas requiring attention are:

1.

Camera Setup

—Are your cameras covering your volume efficiently? Are you using an overlapping volume setup? If not, you may have too many cameras seeing the same area. Over coverage can result in an over abundance of data, slowing the system down.

Note:

Try to have no more than three cameras see any given area of the capture volume from one direction and 10 to 12 cameras total.

2.

Calibration

—Tracking residuals should be below 2.0 mm. If not, try raising the Max. Residual value. Too low a value may cause ghost markers to appear.

3.

Template

—Verify that marker identity is being performed quickly.

Click the

Reset IDs

button several times while the actor is in the capture space and see if there is a lag in acquiring marker identification.

If there is a lag, you may need to create a better template. Verify that the first three markers are the head markers and that the first three links form a rigid triangle. Finally, verify that remaining link definitions flow down the body following the marker definitions.

4.

Frame to Frame Rectifying

—This is mainly influenced by your 3D data quality and tracking parameters. Too many extra, stretchy linkages can cause problems here.

Question

Answer

Can I use MoCap Solver, Si 2.0, or Calcium in

EVaRT

?

Yes. Export a MOD file from Si/Calcium and name it the same as the project file. Select

Model Edit > Tree View

and then in the

Setup > Misc

sub-panel, select

Calcium Solver

in the Skeleton Engine field.

Question

Answer

What is the order of the data in the TRC or TRB files when you use MTOs for tracking?

The resulting

.trc

file from MTO tracking should match the marker order of the resulting "Merge MarkerSets" operation.

Question

Answer

If I have a Solver skeleton setup from

EVa 6.x

, can I use it in

EVaRT

?

Yes, if you go to the

Setup > Misc

sub-panel and in the Skeleton Engine field, you check

Calcium Solver.

in

EVaRT.

You must also copy the

MOD file into your current project folder with the same name as your current PRJ file.

K-2

EVaRT 5.0 User’s Manual Appendix K: Questions and Answers for Specific Applications

Question

Answer

Question

Answer

Question

Answer

Question

Answer

Question

Answer

If I have the

EVa 6.X

Solver setup, can I get an HTR file from

EVaRT

?

Yes. Select the Calcium Solver Skeleton Engine button (Bone button, in lower left of the Post Process dashboard), then select

File > Export HTR

File....

What are the tradeoffs in capturing at 120 fps or faster or slower with

Eagle cameras?

With the older analog cameras, there were tradeoffs in image quality as the frame rate went up. With the Eagle cameras, we see no degradation

AT ALL with the higher frame rates, which is great for high speed captures. The images are all taken off of the Eagle sensors in 2 msec which corresponds to the 500 fps. Waiting longer between frames does not degrade or enhance the image quality. The only considerations are Ethernet bandwidth (not generally a problem), disk space used, and time to process

(or post-process) the raw VC files.

Are there any rules that should be followed when deciding which camera should be set as "Master"?

Any camera can be master. If you have an A-D system, the master camera must be connected to it using the A-D sync cable.

When you see a marker in the 2D display, are you simply seeing a digital representation of what the camera sees at the CCD, or are any of the tracking parameters incorporated into determining whether the system

"sees" a marker (i.e. marker size, centroid parameters, etc.)?

The black data is the raw edge data, affected only by the lighting and the

Threshold. The red dots (lens corrected and/or not) are the calculated centroids. To calculate a centroid, there are two main things: 1) Min Lines per

Marker (usually set to 2 or 3 lines), 2) Max. lines per Marker (usually set to a BIG number like 100), and Shape Analysis (None, Normal or Weak), normally to Normal. But sometimes it is set it to None if it is tossing out centroids, like during a L-Frame seed calibration.

Can you connect 7 or 8 cameras to an EagleHub? We tried connecting 8 cameras to an EagleHub2 and the data transfer to the gigaswitch became quite unstable. Note that this pertains to the older 8-port EagleHubs only

(not the 12-port EagleHubs).

There are only 8 useful ports on an older model EagleHub, which means 7 cameras can be connected to the EagleHub, and one more is used for the uplink to the Network Interface Card (NIC). An eighth camera can be connected directly to one of the remaining open port son the NIC using a patch cable.

K-3

Appendix K: Questions and Answers for Specific Applications EVaRT 5.0 User’s Manual

Question

Answer

Question

Answer

Question

Answer

Is there a way to have two templates for two people in the project that identify them both in real time at once, rather than using one huge template that includes everything?

This is the MTO (Multiple Tracking Object) item. Refer to “Multiple

Tracking Objects” on page 9-7 .

What does the extend template option do?

Extend Template adds new linkage stretch to the existing template if you need it. For example, you can make a one frame template, ID some motion, then you can extend the template so that it knows about the new motion linkage stretches as well as the old ones.

What are the latest specifications for the Eagle system performance?

From the 5 person Eagle camera data included in the

EVaRT

release package under the Samples folder:

Trial: FivestarsAgainandAgain.VCX: 1800 Frames, 120 Frames/sec.

per the EVaRT software.

Biggest VC file: 3900 KB (highest data rate), Smallest VC File: 1100

KB, Avg. VC. size about 2500 KBytes.

Data rate per Eagle camera: Max size file: 3900 KBytes X 120

Frames/sec X 1/1800 Frames = 260 KBytes/sec or about 2600 Kbps

(kilobits/sec) or 2.6 Mbps or about 2.6% of the 100 Mbps Ethernet or about 0.26% of the 1000 Mbps Ethernet

Avg. size file: 2500 KBytes X 120 Frames/sec X 1/1800 Frames =

166 KBytes/sec or about 1660 Kbps or about 1.6% of the 100 Mbps

Ethernet

So, what does this mean for your 12 Camera Setup?

For a 5 person Eagle camera capture, with an average data rate for 120 Hz capture per camera, it works fine (about 1660 Kbps X 12 cameras =

19,920 Kbps or about 20% of the available Ethernet bandwidth). We used a Gigabit Ethernet NIC and Switch for our 24 camera setup. That used about 4% of the 1000 Mbps Ethernet, but would have been 40% of the

100 Mbps Ethernet, which could result in lost packets. The

EVaRT

software is robust enough to deal gracefully with lost packets by ignoring the empty frames and continuing with the capture.

Another dataset in the Samples folder,

Eagles Face and Body

: Eagle one person, 60 Frames/sec, 400 Frames, 300-600 KBytes per camera, average maybe 450 KBytes (450 KBytes X 60 Frames/sec X 1/400 Frames = 68

KBytes /sec or about 680 Kbits/sec or about 0.7% of the 100 Mbps Ethernet).

For your 12 camera setup, this would be: 680 Kbps X 12 cameras = 8160

Kbps or about 8% of the 100 Mbps Ethernet.

K-4

EVaRT 5.0 User’s Manual Appendix K: Questions and Answers for Specific Applications

Question

Answer

Question

Answer

Question

Answer

Question

Answer

Question

Answer

Why is there is a problem displaying the AVI files when a 3D window is displayed on a Windows XP system? The AVI file plays in a jerky motion and the screen is sometimes sliced into horizontal blocks.

This is a problem with Open GL and Direct X displaying at the same time. Right-click on your desktop, then select

Properties > Settings >

Advanced > Troubleshoot

. Slide the Hardware Acceleration down a few points. Quit

EVaRT

, then re-launch. See if that fixes the problem. The settings depend on what your graphics card does and has for features that changes from one computer to another. If that does not fix the problem, try to slide the Hardware Acceleration to

None

. Quit

EVaRT

and then relaunch. If that still does not fix the problem, try to change some of the

Open GL settings found under the name of your graphics card (found under the

Properties > Setting > Advanced

tab). Also, change the Vertical Sync setting to

On by Default

and restart your computer.

We came across a problem while in a two person w/ prop motion capture session. Whenever we recorded a motion, the recorded trc file would be missing a good portion of the marker data. Any idea why the data would just disappear?

This is most likely cause by insufficient marker slots. The default marker slot setting is 192. In your case, you should increase the value. The marker slot setting maybe adjusted under

Setup > Misc

.

We are only using 93 markers (40 per actor, 5 per prop, and 3 for the ball). Shouldn’t 192 marker slots be enough?

Not necessarily. You should always have at least twice as many marker slots then actual markers. The number of slot is dependent on the tracking parameters settings. Each snippet of trajectory requires its own slot and different parameter values will create different sets of trajectories.

How will I know if I need more marker slots?

After loading the file (trb/trc). In Post Processing Mode, scroll down the unnamed marker list (u_marker). If the all the slots are filled then you should increase the number of marker slots.

I just made some changes to my project file, but I do not want to recapture the entire motion list over again. Is there any way to rebuild the trb/trc data from the VC files?

Yes. You can re-record the tracks files using the updated set of parameters in

EVaRT

. First load the VC files. Then go to the

Motion Capture > Output

panel. Select to export a trb or trc file, you have to select the option

OK to Overwrite

if the tracks file already exists. Click on

Record

. The recording will automatically stop at the end of the VC file as long as the option “Loop Raw Files” is not checked (in the

Setup > Misc

panel). If it is checked then you will have to stop it by pressing the

Stop

button (same as the Record button, it changes name during the recording).

K-5

Appendix K: Questions and Answers for Specific Applications EVaRT 5.0 User’s Manual

Question

Answer

Our templates in EVaRT are always either extremely good or extremely poor. Can you give us some time so we can get a consistently good template each time?

The first step of the process is to collect both an init pose and a

ROM(range of motion) for the actor. Do a manual identification of the makers in the init pose (T-pose or A-pose). Use this to create your template.

You should next be able to ID the first frame of the ROM (which should be a T-Pose) and do a Rectify through the whole ROM. Once the ROM has been completely identified, save the changes and then use Extend

Template from the Create Template Dialog box.

The next step is to process all the easier motions. This way you can use these motions to extend your template further which will allow you to have a more complete template by the time you need to track the harder motion files.

K-6

Appendix L

Useful Blank Forms

Motion Capture Log

Human Body Outline—Front

Topic

Human Body Outline—Side

Human Body Outline—Back

Page

L-2

L-3

L-4

L-5

The following blank forms may be useful to prepare for and document a motion capture session. Feel free to make copies as needed.

L-1

Appendix L: Useful Blank Forms

Motion Capture Log

EVaRT 5.0 User’s Manual

Date ______/_____/________

Take

#

EVaRT

Filename # Seconds

Client ____________________________________________

Ref.

Video

Counter Comments Results

L-2

EVaRT 5.0 User’s Manual Appendix L: Useful Blank Forms

Human Body Outline—Front

Project_____________________________________________Date______/_____/____

L-3

Appendix L: Useful Blank Forms EVaRT 5.0 User’s Manual

Human Body Outline—Side

Project_______________________________________________Date______/_____/____

L-4

EVaRT 5.0 User’s Manual Appendix L: Useful Blank Forms

Human Body Outline—Back

Project_______________________________________________Date______/_____/____

L-5

Appendix L: Useful Blank Forms EVaRT 5.0 User’s Manual

L-6

Index

Numerics

10 Camera Setup

Typical

, 5-7

10 EMG Channels Connections

, B-17, B-19

12 Camera Setup

Typical

, 5-8

14 Camera Setup

Typical

, 5-8

16 Camera Setup

Two-Tier

, 5-4

Typical

, 5-9

16 Camera, 2 EagleHub Configuration

, A-9

16 Camera, 3 EagleHub Configuration

, A-10

2 Markers

, 11-16

2 Panes Top/Bottom

, 6-15

24 Camera, 4 EagleHub Configuration

, A-11

28 Camera, 3-Tier Setup

, 5-5

2D Display

, 6-7, 6-14, 6-19 options

, 6-19 pop-up menu

, 6-19

3 Markers

, 11-16

3 Point Average Filter

, 10-9

32 Camera Setup

Typical

, 5-10

3D Display

options

, 6-18

pop-up menu

, 6-17

3D Studio Max

, D-1

4 Panes Layout

, 6-14

5 Point Average Filter

, 10-9

6 Camera Setup

Typical

, 5-6

8 Camera Setup

Typical

, 5-7

8 Camera, 1 EagleHub Configuration

, A-7

8 Camera, 2 EagleHub Configuration

, A-8

8 Eagle Camera Setup

Typical

, 5-11

A

Accuracy

, 11-16

Additional Tracking Objects

, 9-9

Adjusting Thresholds

, 7-12

All Markers Radial Button

, 6-35, 10-11

All On Button

, 8-11

Amplitude Zoom

, 6-40

EVaRT 5.0 User’s Manual

AMTI Forceplates

Calibration Matrix

, E-6

forcepla.cal

, E-2

Using

, E-6

AMTI Forceplates Connections

, B-17, B-19

AMTI Gain Setting

, E-6

AMTI or Bertec Forceplates

, B-4

Analog ASCII Row Column

, G-15

Analog Camera

System Configuration

, A-14

Analog Channel Names

Replace

, 6-10

analog data files

opening

, H-6

Analog data graphs

, 6-7

Analog Display

, 6-14, 6-20

pop-up menu

, 6-20

analog forceplate data

, 6-14

Analog Input Channel Connections

, B-13, B-16, B-

18

Analog Input Hardware Connections

, B-1

Overview

, B-1

Analog Setup

, 7-13

Grid

, 7-13

Analysis

, J-1

Exporting Information

, 10-30

graphs

, 10-28

Analysis Graphs F7

, 6-7

ANB

, G-18

ANB file

, 6-14

ANC

, G-15

Example

, G-15

Animation

, C-1

Animation Plugins

, 1-15

Auto Scale

, 10-3

Autozero Forces

, 7-20

B

Backup Media

, 4-4

Batch Processing

, 9-13

Batch Processing Options

, 6-8

Bertec Forceplates

Calibration Matrix

, E-6

forcepla.cal

, E-2

Using

, E-6

Bertec Gain Setting

, E-6

Binary Files

, G-18

ANB

, G-18

C3D

, G-18

Index-1

EVaRT 5.0 User’s Manual

TRB

, G-18

BioFeedTrak

, 1-11, 12-7

BioFeedTrak Sub-Panel

, 12-7

Biomechanics

, C-3

Blank Forms

, L-1

Human Body Outline—Back

, L-5

Human Body Outline—Front

, L-3

Human Body Outline—Side

, L-4

Motion Capture Log

, L-2

Building a Template

, 9-5

Face

, D-5

Butterworth Filter

, 10-6

C

C3D

, G-18

C3D files

see tracked marker files

Calcium

, 1-7

Calcium Segments

, 11-15

Calculate Solver Skeleton

, 6-8

CalFloor.vc1

, 8-13

calibrating system overview

, 5-20

Calibrating Your System

, 8-1

Calibration

, 6-12

Coordinate System

, 5-20

Face

, 8-13

Floor

, 8-13

from Previously Collected Files

, 8-18

Refining

, 8-15

Simulated

, 8-18

Calibration Frame Tab

, 8-5

Calibration Settings Window

, 8-5

Calibration Settings Window Tabs

, 8-4

Calibration Frame

, 8-4

Capture Volume

, 8-8

Origin Offsets

, 8-5

Calibration Setup

, 6-7

Calibration Square

, 5-20, 5-21, 5-22, 8-1

Placing

, 5-22

Worksheet

, 5-21

Calibration Sub-Panel

, 8-1, 8-3

Calibration Wand

, 8-1

CalSeed.vcX

, 8-10

CalWand.vcX

, 8-12

Camcorder

, 4-4

Camera and Strobe Settings

Summary

, A-21

Camera Buttons

, 6-32

Index-2

Right-Click

, 6-32

camera buttons

Real Time Dashboard

, 6-24

Camera Connections

, A-17

Camera Connector Assembly

, A-17

Camera Positioning

, 8-8

Camera Problems

, 5-24

Camera Software

, 7-5

Camera Type drop down list

, 7-1

Cameras

Optimum Number

, 5-2

Properly Seeded

, 8-10

Setting Up

, 5-11

Square (Seed) Calibration

, 8-8

Wand Calibration

, 8-11

Cameras Sub-Panel

, 7-1

Capture Volume

Sizes

, 5-5

Capture Volume and Marker Size

Relationship

, 5-29

Capturing Facial Motion

, D-1

Accessory Kit

, D-1

Cheeks Marker Placement

, D-5

Chin Marker Placement

, D-5

Examples

, D-7

Eyebrows Marker Placement

, D-4

Eyelids Marker Placement

, D-5

Four Camera

, D-3

Head Marker Placement

, D-4

Jaw Marker Placement

, D-5

Lips Marker Placement

, D-5

Marker Placement

, D-4

Nose Bridge Marker Placement

, D-5

Nose Marker Placement

, D-5

Overview

, D-1

Setup

, D-1

System Configuration

, D-1

Three Camera

, D-3

Centroid Parameters

, 9-2

Channels Table

, 6-20

Clear Marker Set Button

, 11-2

Clearing Masks

, 7-12

Cohu 4915 60 Hz

Camera Connections

, A-20

Collect and Calibrate Button

, 8-10

Collect Calibration Square Button

, 8-7

Color Video

, 6-6

Color Video option

, 6-14

Colors

, 6-9

Colors Form

, 6-9

Connect Cameras Button

, 6-31

Connect to Cameras

, 7-1

Connections

10 EMG Channels

, B-17, B-19

AMTI Forceplates

, B-17, B-19

Analog Input Channel

, B-13, B-16, B-18

Analog Input Hardware

, B-1

EagleHub1

, A-7

EagleHub2

, A-7

EagleHub3

, A-5

Specific Cameras

, A-18

Control Points

, 5-21

Convert .anc File...

, 6-5

Coordinate System

Calibration

, 5-20

Create Linkages Button

, 11-3

Create Orthotrak Model

, 7-17

Create Template

, 6-35, 9-6, 10-11

Creating Masks

, 7-12

Creation Date

, 15-7

Current Camera Information

, 7-4

Current Frame

, 10-2

Cut

, 6-37, 10-13

D

Data

Viewing

, 10-2

Data Capture

MA Quickstart

, 2-11

NM Quickstart

, 3-9

Data Painting

, 10-26

Data Views

, 6-3, 6-6

Dedicated Interface

, 7-7

Degrees of Freedom

, 11-17

Delete All Linkages Button

, 11-6

Delete Outside Volume

, 12-2

Delete Short Snippets

, 12-2

Digital Video Option

, 6-23

Director/Sequencer

, 1-18

Directory List

, 6-11

Display Codes

, 7-9

Distance Between Two Markers Tab

, 10-29

Draw 3D Points Button

, 11-3

Dynamic Template Stretch Limits

, 9-4

E

Eagle Camera

Connections

, A-12

EVaRT 5.0 User’s Manual

Dimensions

, 5-17

Display

, 7-9

Physical Dimensions

, 5-17

Power Consumption

, A-3

Settings

, 7-2

System Configuration

, A-2

EagleHub

, A-2–A-12

EagleHub1 Connections

, A-7

EagleHub2 Connections

, A-7

EagleHub3 Connections

, A-5

Edit Thresholds

, 6-8, 7-12

Editing Tracked Data

, 10-1

EMG

Muscle Name Selection

, 7-14

Signal Name Conventions

, B-4

Enable External Trigger

, 9-12

Euler Angle Order

, 13-7

EVaDV Software

, 1-13, 6-23

EVaRT Project File

, G-3

EVaRT.ini

, 6-18

Examples

Marker Sets

, C-4

Exchange

, 6-36, 10-12

Export File Formats

, G-1

Export Forces File...

, 6-5

Export.ts (Time Series) File... button

, 10-30

Extend Template

, 2-14

Extending the Seed Calibration

, 8-19

External Trigger Mechanism

, 9-14

Extra Stretch

, 11-12

Eye Movement

, D-14

F

Face

Building a Template

, D-5

Face Calibration

, 8-13

Facial Animation Techniques

, D-9

Falcon Camera

Connections

, A-18

Settings

, 7-12

System Configuration

, A-14

Threshold Monitor

, 7-8

F-F Ethernet adapter

, A-7

FIFO slider

, 6-14

file management

, 6-3

Filters

, 10-4

Flash Drive

Software Installation

, 1-4

Floor Calibration

, 8-13

Index-3

EVaRT 5.0 User’s Manual

Force Plate Channel Order

, G-15

Force Vector Scale

, 7-20

forcepla.cal

, E-1

General Information

, E-1

Forcepla.cal File Format

, E-1

Forceplate

3x3 Orientation Matrix

, E-3

Calibration Matrix

, E-3

File Data

, E-3

Number

, E-3

Optional Length & Width

, E-3

Scaling Factor

, E-3

True XYZ Origin

, E-3

XYZ Location in Video Coordinate System

, E-3

Forceplate Forces

, 6-9

Forceplate Scaling Factor

, E-5

Forceplates

6x6 Calibration Matrix

, E-7

Frame Counter

, 6-32

Frame Offset

, 7-19

Frame Rate

, 7-3

Frames, Selecting

, 10-4

G

gait analysis

, 5-2

gait.lib

, E-1

Gauss Newton

, 11-15

ghost markers

, K-1

Global Marker Data Adjustments

, 12-3

Global Scale

, 11-15

Going Live

, 7-12

H

Hardware required

, 1-3

Hawk Camera

Connections

, A-12

Dimensions

, 5-18

Display Codes

, 7-9

Physical Dimensions

, 5-18

Power Consumption

, A-3

Settings

, 7-2

System Configuration

, A-2

Helen Hayes Marker Set

, 2-10, C-3, C-4

Help Menu

, 6-11

Hide Markers

, 6-36, 10-12

Hierarchical Translation & Rotation Data

, 11-17

Hierarchical Translations and Rotations

, G-5

High Frame Button

, 10-2

Hinge Joints

Example

, 11-19

Hot Keys and Tips

, 6-39

HTR

, G-9

HTR Graphs pop-up menu

, 6-22

HTR Version 1

Example

, G-9

HTR Version 2

Example

, G-10

HTR2

, G-5

Example

, G-6

I

Identifying

, 6-24

Identifying Markers

, 10-11

Import File Formats

, G-1, H-1, I-1, J-1

Included Angles Tab

, 10-30

Information Center

, 6-13

Installing Software

, 1-4

International Society of Biomechanics

, 5-21

IP Addresses

, A-13

ISB

, 5-21

J

Join

, 10-22

Join Virtual

, 6-31, 6-33, 10-22

Guidelines

, 6-39, 10-25

Joining Gaps in Data

, 10-4

K

Keyframe Animation

, D-13

KinTrak

, 1-18

Kistler Forceplates

, B-4

Calibration Matrix

, E-8

Example

, E-8

forcepla.cal

, E-2

Gain Setting

, E-8

General Notes

, E-9

Signal Names

, E-8

True XYZ Origin

, E-8

Using

, E-8

Kyowa Dengyo Forceplates

, E-9

Laboratory

Conditions

, 5-1

Motion Capture

, 5-1

L

Index-4

Supplies

, 5-1

layout control

, 6-3

Lenses/Orientation

, 8-7

Levenberg-Marquart

, 11-16

Library Error

, 5-29

License File

Sample

, 1-5

Linkage Stretch Parameters

, 9-3, 10-16

Linkages

Selecting

, 6-41

Links

, 11-12

Load .ini Preferences...

, 6-6

Load Analog Setup...

, 6-6

Load Calibration...

, 6-6

Load Into EVaRT

, 15-8

Load Last Capture

, 9-14

Load Marker Set...

, 6-6

Load Project

, 6-14

Load Tracks File

, 6-15

Loading New Camera Software

, 7-5, 7-7

Low and High Selected Frames

, 10-3

Low and High Visible Frames

, 10-3

Low Frame Button

, 10-2

M

mac_lic.dat

, 1-5, G-2

Main Marker Set

, 9-9

Make Unnamed

, 6-35, 10-11

marker cloud

defined

, H-1

Marker ID

, 6-36, 10-12

Marker Placement

Capturing Facial Motion

, D-4

Cheeks

, D-5

Chin

, D-5

Eyebrows

, D-4

Eyelids

, D-5

Head

, D-4

Jaw

, D-5

Lips

, D-5

MA Quickstart

, 2-9

NM Quickstart

, 3-8

Nose

, D-5

Nose Bridge

, D-5

Marker Sets

, C-1

Animation

, C-1

Biomechanics

, C-3

Developing

, C-5

Examples

, C-4

EVaRT 5.0 User’s Manual

Overview

, C-1

Marker Size

, 5-30, 9-3

Marker Slots

Number of

, 7-19

Markers

Identifying

, 10-11

Selecting

, 6-41

Unnamed

, 10-10

Markers Panel

Create Linkages Button

, 11-3

Markers Sub-Panel

, 11-2

Clear Marker Set Button

, 11-2

Draw 3D Points Button

, 11-3

Select and Edit Button

, 11-3

Masks

Clearing

, 7-12

Creating

, 7-12

MasterProjectsList.mdb

, 15-2

Matrix Method

, 11-15

Max Acceptable

, 9-4, 9-5, 10-16

Max Horizontal Lines per Marker

, 9-3

Max Prediction Error

, 9-3

Max Residual

, 9-3

Max Speed (mm/frame)

, 10-20

Max Target Speed

, 9-3

Max. Prediction Error (mm)

, 10-21

Maya

, D-1

Menu Bar

, 6-3

Merge Marker Sets

, 9-9

Mesh Deformation

, D-9

MIDAS Connections

, A-16

Min. Cameras To Use

, 9-3

Min. Horizontal Lines per Marker

, 9-2

Misc Sub-Panel

, 7-16

mocap model

, H-15

joint center calculations

, H-13

marker set

, H-14

scaling

, H-13

static pose

, H-10

Mode Panel Buttons

, 6-12

Model Adjustments

, 12-3

Model Edit

, 6-12, 11-1

Overview

, 11-1

Morphing

, D-10

Motion Analysis Corp.

Contact Information

, 1-20

Motion Analysis License File

Example

, G-2

Motion Capture

, 6-12

Index-5

EVaRT 5.0 User’s Manual

Overview

, 9-1

Motion Capture Laboratory

Setting Up

, 5-1

Motion Capture of Hands

, D-15

Motion Capture Panel

, 9-1

Motion Capture Terminology

, 4-5

Motion Composer

, 1-9, 12-6

Motion Composer Sub-Panel

, 12-6

Motion Module

, H-1–H-17

opening analog data files

, H-6

opening C3D files

, H-2

opening TRB/C files

, H-2

opening XLS files

, H-6

real-time import

, H-7

Motionbuilder

, D-1

Move

, 4-5

msscript.ocx

Installing

, 14-7

MTO

, 9-7

Multiple Tracking Objects

, 9-7

Muscle Name Selection

EMG

, 7-14

N

National Instruments

, B-14

Network Configuration

, A-13

Network Interface Card (NIC)

, A-7, A-13

New Subject Button

, 6-24

New Subject button

, 6-25

Next Frame Button

, 10-2

NI PCI-6071E

, B-1, B-14

NI USB-6218

, B-1, B-5

NIDAQ Software

, B-3

O

OBJ

, 11-7

Objects Sub-Panel

, 9-8

Options

Post Process Panel

, 6-36, 10-12

Orient Body

, 11-16

OrthoTrak

, 1-17

Intended Use

, 1-17

Output Files

, 9-12

Output Sub-Panel

, 9-11

P

P3D

, G-15

PCS_16Camera_2Tier.prj

, 5-4

Index-6

Phoneme Recognition

, D-13

Pig-Tail Cable

, A-17

Pinging a camera

, 5-27

Play Backward Button

, 10-2

Play Forward Button

, 10-2

pop

, 6-22

pop-up menu

, 6-17

2D Display

, 6-19

3D Display

, 6-17

Analog Display

, 6-20

HTR Graphs

, 6-22

XYZ Graphs

, 6-21

Pose ID Options

, 6-8

Position, Velocity, and Acceleration Tab

, 10-28

Positioning Cameras

, 8-8

Post Process

, 6-12

Post Process Dashboard

, 6-34

Post Process mode

, 6-15, 6-34

Post Process Toolbar

, 6-35, 10-11

Post Processing

, 10-1

Square Data

, 8-22

Strategies and Tips

, 10-31

Wand Data

, 8-22

Post Skeleton Options

, 6-8

Post Trigger Mode

, 9-12, 9-14

Power Consumption

, A-3, A-15

Eagle

, A-3

Hawk

, A-3

Preview Calibration check box

, 8-7

Previous Frame Button

, 10-2

PRJ Files

, 6-13, G-3

Project Databases

, 15-4

Project Initialization

MA Quickstart

, 2-2

NM Quickstart

, 3-2

Prop Definition

, 9-6

Props

, 4-4

Protect Lens Correction

, 7-19

Pulnix Camera

Switches and Connections

, A-19

Push-Button Switch

, A-22

Q

Questions and Answers

, K-1

Quick ID

, 6-36, 10-12

quick solve defined

, H-4

QuickDB

, 15-1

Quick-Start

, 15-2

Terminology

, 15-3

User Interface

, 15-3

QuickDB Sub-Panel

, 12-9

Quick-Start Tutorial

Movement Analysis

, 2-1

R

Range of Motion

, 3-9, 9-5

Raw Files

, 6-14

Raw Video Button

, 6-31

RB Join

, 6-38, 10-14

Real Time Dashboard

, 6-12, 6-23

camera buttons

, 6-24

real-time import

see Motion Module

Record

, 6-11

Recording Data

, 9-14

Rectify

, 6-36, 6-37, 10-12, 10-13, 10-15

Rectify Functions

, 6-37, 10-13, 10-15

Rectify Unnamed

, 6-35, 10-11, 10-17

Refine Tracks

, 12-3

Refining a Calibration

, 8-15

Requirements

System

, 1-3

Reset IDs Button

, 6-24

Rigid Body Rectify

, 6-36, 10-12, 10-17

rigid objects

, 9-9

ROM

, 3-9, 9-5

rom.bin

, 7-5

Rotating

, 6-39

Rotation Offset

, 11-14

Run Button

, 6-31

functions

, 6-31

Run mode

, 6-23

S

Sample Data

Viewing

, 6-14

Sample Form

, 4-5

Samples directory

, 6-14

Save .ini Preferences...

, 6-6

Scaling

, 10-3

ScriptPlugin.ocx

Installing

, 14-7

SDK

, F-1

Seed Calibration

Extending

, 8-19

Segment

Selecting

, 6-41

EVaRT 5.0 User’s Manual

Select All Frames

, 10-4

Select and Edit Button

, 11-3

Select Marker Set

, 9-9

Selected Markers Radial Button

, 6-35, 10-11

Selecting Frames

, 10-4

Buttons

, 10-4

Selecting Linkages

, 6-41

Selecting Markers

, 6-41

Selecting Segments

, 6-41

Selecting Virtual Markers

, 6-41

Set as Master

, 7-4

Set Master Camera button

, 7-4

Setting Up

Cameras

, 5-11

Settings

Output Sub-Panel

, 9-12

Setup

, 6-12

Setup Analog

MA Quickstart

, 2-8

Setup Panel

, 7-1

Shape Analysis

, 9-3

Show

, 6-18

Show and Use Masks

, 7-8

Show Camera Field of View

, 5-12

Show Residuals and Cameras

, 10-3

Show Skin

, 11-7

Show Threshold

, 7-8

Show Video

, 7-8

Show Volume

, 5-13

Shutter Speed

, 7-3

SIMM

, 1-18

SIMM OrthoTrak Model

, 7-17

SIMM Solver

Options

, 7-19

Simulated Calibration

, 8-18

Simulation Mode Speed

, 7-21

SkB

, 1-8

SkB Segments

, 11-14

Skeleton Builder

, 1-8

Skeleton Definitions

, 6-5

Skeleton Graphs

, 6-7

Skeleton Options

, 7-16

Skeleton Types

, 13-1

Skin File

, 11-6

Skin Tranparency

, 11-6

Sky

FAQ

, 14-7

Functions

, 14-3

Graphical User Interface

, 14-2

Index-7

EVaRT 5.0 User’s Manual

Script

, 14-4

Script Examples

, 14-5

Sky Files

Global

, 14-3

Local

, 14-3

Sky Interface

Toolbar

, 14-2

Sky Scripting Interface

, 14-1

Sky Sub-Panel

, 12-5

Sky Writer

, 14-1

Slate Board

, 4-4

Smear Display

, 8-11

Smoothing Options Tab

, 10-5

Snippets

, 12-2

Socket Error

, 5-29

SoftImage

, D-1

Software

Installing

, 1-4

required

, 1-4

Software Developers Kit

, F-1

Sound Effects

, 7-19

Square (Seed) Calibration

Cameras

, 8-8

Square Calibration

, 8-2

Standard System Configuration

, A-2

Starting EVaRT

Quickstart

, 2-1

static pose

see Motion Module

Status Bar Messages

, 6-12

Still Camera

, 4-4

Streaming Options

, 7-19, 7-20

Studio Preparation

, 4-1

Subject Preparation

, 4-6

Sub-Panel Buttons

, 6-12

Sub-sampling Rate

, 7-8

Sync with EVaRT

, 15-8

Synchronizing Video

, I-1, J-1

System Calibrating Process

, 5-20

System Calibration

MA Quickstart

, 2-4

NM Quickstart

, 3-3

System Configuration

Analog Camera

, A-14

System Hardware Interconnection

, A-1

Overview

, A-1

Standard System Configuration

, A-2

System Objects

, 9-9

System Requirements

, 1-3

Index-8

T

Take

, 4-5

Talon Plugins

, 1-16

Talon Viewer

, 1-17

Target Marker

, 10-3

Template

Building

, 9-5

Template ID

, 6-35, 10-11

Template ID Details

, 10-18

Template Rectify

, 6-35, 10-11, 10-16

The

, 6-38, 10-14

Threshold Monitor

, 7-8

Thresholds

Adjusting

, 7-12

Edit

, 7-12

Time Code

, 6-41

Time Code Counter

, 6-32

Time Lines

, 6-8, 10-27

Time Series Files

, G-16

Time Zoom

Method 1

, 6-39

Method 2

, 6-40

Time Zoom Slider

, 10-2

To Reconsider

, 9-4, 10-16

Tool Menu

, 6-13

tools

, 6-3

Tools Menu

, 6-7

Track Row Column

, G-3

Tracked Data

Editing

, 10-1

Viewing

, 6-16

tracked marker files

cropping ends

, H-4

opening

, H-2

Tracking

, 6-24

With More Than 8 Cameras

, 5-15

Tracking mode

, 6-34

Tracking Parameters

, 9-2

Tracking Sub-Panel

, 9-2

Translating

, 6-39

TRB

, G-18

TRB file

, 6-14

TRB/TRC files

see tracked marker files

TRC

, G-3

Data Header

, G-4

Empty Fields

, G-4

Example

, G-4

File Header

, G-3

EVaRT 5.0 User’s Manual

Position Data

, G-4

Tree View Sub-Panel

, 11-5

Delete All Linkages Button

, 11-6

V-Marker Definition Button

, 11-5

Trial

, 4-5

Trial Fields

, 15-9

Trial List

, 15-7

tripod mounting points

, 5-17, 5-18, 5-19

Troubleshooting

Diagnosing Camera Problems

, 5-24

Eagles and Hawks

, 5-24

Ethernet

, A-4

TS

, G-16

Typical Camera Setup

, 5-11

U

Undo

, 6-38, 10-14

Unhide Markers

, 6-36, 10-12

Uniform Scale

, 10-3

Unload Tracks Button

, 6-43

Unnamed Markers

, 10-10

Use Joint Limits

, 11-16

User Apps

, 6-12, 12-1

Overview

, 12-1, 13-1, 14-1, 15-1

User Interface

, 6-1, 6-2

V

VC file

, 6-14

Video Display Options

, 7-8

Video Processor (MIDAS) Connections

, A-16

Viewing Your Data

, 10-2

Virtual Marker

Quickstart Example

, 11-20

Types

, 11-17

Virtual Marker Definitions

, 6-8, 11-16

Virtual Marker Tracks

Calculating

, 11-20

Virtual Markers

, 11-16

Selecting

, 6-41

Visible Channels

, 6-20

Visible header

, 6-20

V-Marker Definition Button

, 11-5

V-Marker Definitions Button

, 11-20

VMarkers

, 11-11

W

Waldos

, D-14

Wand

, 8-1

Wand Calibration

, 8-2, 8-11

Wand Calibration Coverage

, 2-6

Wand Processing Status

, 8-12

Wand Processing Status Window

Accept Button

, 8-12

Extend Seed Button

, 8-12

Reject Button

, 8-12

Run Again Button

, 8-12

Stop Button

, 8-12

X

X Offset

, 11-11

X Sub-Panel

, 12-2

XYZ Graphs

, 6-7, 6-21

pop-up menu

, 6-21

Y

Y Offset

, 11-11

Z Offset

, 11-11

Zoom

amplitude

, 6-40

Zoom In-Zoom Out

, 10-3

Zooming

, 6-39

Z

Index-9

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement