Nebulosity 3 Manual

Nebulosity 3 Manual
Version 3.0
February, 2012
Craig E. L. Stark, Ph.D.
Table of Contents
ACKNOWLEDGEMENTS ......................................................................................................................................5
LICENSE ...................................................................................................................................................................6
INTRODUCTION ....................................................................................................................................................7
WHAT’S NEW IN 3.0? ...........................................................................................................................................8
FEATURES ............................................................................................................................................................10
MAIN SCREEN ......................................................................................................................................................15
IMAGE WINDOW ......................................................................................................................................................................15
DISPLAY PANEL .......................................................................................................................................................................15
CAPTURE PANEL .....................................................................................................................................................................17
Camera Section ...................................................................................................................................................................17
Exposure Section .................................................................................................................................................................17
Capture Section ...................................................................................................................................................................18
STATUS BAR ............................................................................................................................................................................20
CUSTOMIZATION OF THE INTERFACE ...................................................................................................................................20
Notes and History ...............................................................................................................................................................21
Camera-­‐speci<ic Dialogs ..................................................................................................................................................21
External Filter Wheel .......................................................................................................................................................22
Pixel Stats ...............................................................................................................................................................................22
Link to PHD Guiding ..........................................................................................................................................................22
Macro Tool ............................................................................................................................................................................23
CAPTURING IMAGES .........................................................................................................................................24
MONOCHROME VS. COLOR CAMERAS ...................................................................................................................................24
ONE-­‐SHOT COLOR CAMERAS: SHOULD I CAPTURE RAW OR RGB? ................................................................................25
FILE FORMATS ..........................................................................................................................................................................26
CAMERA GAIN AND OFFSET ...................................................................................................................................................27
How should I set my gain and offset to set it and forget it? .............................................................................27
OVERVIEW OF IMAGE PROCESSING .............................................................................................................29
Deciding on Bad Pixel Mapping vs. Dark Subtraction ........................................................................................29
Applying the darks, <lats, and biases ..........................................................................................................................30
Using Bad Pixel Mapping ................................................................................................................................................31
Converting RAW images to Color and/or Pixel Squaring (aka Reconstruction) ....................................32
Normalize Images (optional) ........................................................................................................................................32
Grading and Removing Frames (optional) .............................................................................................................33
1
Stacking: Align and Combine ........................................................................................................................................33
Crop off the edges ...............................................................................................................................................................34
Remove the Skyglow Color .............................................................................................................................................35
Stretching ..............................................................................................................................................................................35
IMAGE PRE-­‐PROCESSING: THE DETAILS ....................................................................................................37
PRE-­‐PROCESSING: THEORY ....................................................................................................................................................37
IMAGE SET PRE-­‐PROCESSING TOOL .....................................................................................................................................38
BAD PIXEL MAPPING ..............................................................................................................................................................40
RECONSTRUCTION: DEMOSAIC-­‐ING AND PIXEL SQUARING ................................................................42
PREVIEWING AND GRADING IMAGES .........................................................................................................44
LRGB COLOR SYNTHESIS .................................................................................................................................45
RGB MODE ..............................................................................................................................................................................45
LRGB: TRADITIONAL HSI MODE ........................................................................................................................................45
LRGB: COLOR RATIO MODE .................................................................................................................................................45
IMAGE NORMALIZATION AND HISTOGRAM MATCHING ......................................................................46
STACKING IMAGES .............................................................................................................................................47
OVERVIEW OF TRADITIONAL STAR-­‐BASED ALIGNMENT .................................................................................................47
COMBINATION METHODS .......................................................................................................................................................48
Averaging vs. Adding vs. Adaptive Stacking ...........................................................................................................48
Standard Deviation Based Stacking (Sigma-­‐clip) ................................................................................................49
ALIGN AND COMBINE OPTIONS .............................................................................................................................................50
Align and Combine: Fixed ...............................................................................................................................................50
Align and Combine: Translation ..................................................................................................................................50
Align and Combine: Translation + Rotation (+ Scaling) ...................................................................................52
Drizzle alignment (Translation & Rotation + Resolution enhancement) ..................................................52
Colors in Motion (Simultaneous translation align + Color reconstruction) .............................................54
AUTOMATIC ALIGNMENT (NON-­‐STELLAR) .........................................................................................................................55
IMAGE ADJUSTMENT ........................................................................................................................................57
DEMOSAIC’ING AND PIXEL SQUARING ..................................................................................................................................57
Demosaic Image ..................................................................................................................................................................57
Square B&W ..........................................................................................................................................................................57
Reconstructing Images from One-­‐shot Color Cameras and Line Filters ....................................................57
GEOMETRIC MANIPULATIONS ................................................................................................................................................59
Crop ..........................................................................................................................................................................................59
Mirror / Rotate ....................................................................................................................................................................59
Binning ....................................................................................................................................................................................59
Resize .......................................................................................................................................................................................60
STRETCHING AND INTENSITY SCALING ................................................................................................................................60
Levels / Power Stretch .....................................................................................................................................................60
CURVES .....................................................................................................................................................................................62
Digital Development Processing (DDP) ....................................................................................................................63
Zero Min .................................................................................................................................................................................64
Scale Intensity / Pixel Math ...........................................................................................................................................64
COLOR ADJUSTMENT ...............................................................................................................................................................64
Adjust Color Offset .............................................................................................................................................................64
Adjust Color Scaling ..........................................................................................................................................................65
Auto Color Balance ............................................................................................................................................................65
Hue / Saturation .................................................................................................................................................................65
Discard Color and Convert to Color ............................................................................................................................65
LRGB COLOR SYNTHESIS TOOL ...........................................................................................................................................65
2
SHARPENING, BLURRING, AND NOISE REDUCTION ............................................................................................................65
Sharpening and Tighten Star Edges ..........................................................................................................................66
Gaussian Blur .......................................................................................................................................................................66
3x3 Pixel Median .................................................................................................................................................................67
Vertical Smoothing (deinterlace) ................................................................................................................................67
NOISE REDUCTION ..................................................................................................................................................................67
OTHER TOOLS .....................................................................................................................................................70
LAUNCH NEW INSTANCE ........................................................................................................................................................70
CHANGE LANGUAGE ................................................................................................................................................................70
IMAGE INFO AND FITS HEADER ...........................................................................................................................................70
DSS LOADER ...........................................................................................................................................................................70
MEASURE DISTANCE ...............................................................................................................................................................70
CHECK / UPDATE LICENSE ....................................................................................................................................................71
SUPPORTED CAMERAS ....................................................................................................................................72
SBIG .........................................................................................................................................................................................72
STARLIGHT XPRESS .................................................................................................................................................................72
CANON DIGIC II, III, & 4 DSLRS ........................................................................................................................................73
Color images and FITS/CR2 ...........................................................................................................................................73
White balance ......................................................................................................................................................................74
Long exposures / bulb triggers .....................................................................................................................................74
Mirror lockup .......................................................................................................................................................................75
Mode dials and lens settings ..........................................................................................................................................75
Troubleshooting Connections .......................................................................................................................................76
ASCOM CAMERAS ..................................................................................................................................................................76
SAC7 AND LONG EXPOSURE WEBCAMS .............................................................................................................................77
CAMERA ADVANCED PANEL ..................................................................................................................................................77
TAKING GOOD IMAGES ....................................................................................................................................79
YOUR TELESCOPE ....................................................................................................................................................................79
YOUR MOUNT ..........................................................................................................................................................................80
Polar Alignment ..................................................................................................................................................................80
Periodic Error and Guiding ............................................................................................................................................80
FOCUS ........................................................................................................................................................................................81
EXPOSURE SETTINGS ...............................................................................................................................................................82
Rule #1: Use the Histogram to keep your background above the <loor and bright bits below the ceiling. ......................................................................................................................................................................................82
Rule #2: Take lots of images ..........................................................................................................................................83
Rule #3: Don't over-­‐tax your mount ..........................................................................................................................83
What do gain and offset do? ..........................................................................................................................................83
Gain's downside: Bit depth and dynamic range ....................................................................................................84
How do manufacturers determine gain and offset for cameras that don't allow the user to adjust them? ........................................................................................................................................................................................84
MENU QUICK REFERENCE ...............................................................................................................................86
FILE MENU ...............................................................................................................................................................................86
EDIT MENU ..............................................................................................................................................................................86
BATCH MENU (SEE ALSO THE PRE-­‐PROCESSING SECTION) .............................................................................................87
IMAGE MENU (SEE ALSO THE IMAGE ADJUSTMENT SECTION) .........................................................................................87
VIEW MENU .............................................................................................................................................................................89
HELP MENU .............................................................................................................................................................................89
PREFERENCES .....................................................................................................................................................90
CAPTURE ...................................................................................................................................................................................90
OUTPUT ....................................................................................................................................................................................91
3
PROCESSING .............................................................................................................................................................................91
COLORS .....................................................................................................................................................................................92
MISC ..........................................................................................................................................................................................92
SCRIPTS .................................................................................................................................................................93
SCRIPT EDITOR ........................................................................................................................................................................93
FULL COMMAND LIST .............................................................................................................................................................94
Capture Setup Commands ..............................................................................................................................................94
Control Commands ............................................................................................................................................................95
Advanced Camera Control Commands ......................................................................................................................96
All text and images Copyright Craig Stark, Stark Labs 2005-2012
Last updated February, 2012
4
1 Acknowledgements
The author would like to extend his heartfelt thanks to several individuals who have
helped in the creation of Nebulosity. In particular, I would like to thank Michael Garvin,
Chuck Kimball, Sean Prange, and Dave Schmenck for all their perennial help. Several
individuals worked very hard in translating Nebulosity into other languages and I am
very grateful for the help of Christoph Bosshard (German), Denis Bram (French),
Michele Palama (Italian), Rodolphe Pineau (French), and Ferry Zijp-Herzberg (Dutch).
I would also like to acknowledge the fine wxWidgets cross-platform GUI library used
extensively here. Without it, I would not have written Nebulosity. I would like to
acknowledge use of the FreeImage, LibRAW, and CFITSIO libraries for image input and
output. In addition, note that the noise reduction software GREYCstoration and the
automatic alignment software ANTS, are included as binary applications bundled with
and called from Nebulosity.
5
2 License
Nebulosity ("Software") and all of its documentation are Copyright 2005-2012 Craig
Stark (Stark Labs). This is commercial software and is not freeware. Please read the
following "Fair" End User License Agreement carefully.
1. We grant you one license to install and use this software on one computer at a time. If you do not
agree to the following terms of this license, please uninstall and remove all copies. If you have
purchased this software and it was not included with a camera purchase, return the product within 10
days of your purchase for a full refund.
2. You may install and use the software on another computer, but the software should not be in use on
more than one computer at a time unless you purchase additional licenses (e.g., a laptop used for image
collection and a desktop used for subsequent image processing). You may make back-up copies of the
software for archival purposes. You may permanently transfer your license to use the software to another
party who will be bound by this agreement, provided you do not retain any copies of the software.
3. The software is protected by the copyright laws of the U.S. and other countries, and we retain all
intellectual property rights in the software. You may not separately publish, sell, market, distribute, lend,
lease, rent, or sublicense the software. However, this license is not to be construed as prohibiting or
limiting any fair use sanctioned by copyright law.
LIMITED WARRANTY
4. We warrant that the software will provide the features and functions generally described in the product
specification on our website when you purchased it and in the product documentation. Media on which
the Software is furnished, if any, will be free from defects in materials and workmanship.
5. We have taken all reasonable steps to keep the software free of viruses, spyware, "back door"
entrances, or any other harmful code. We will not track or collect any information about you, your data,
or your use of the software except as you specifically authorize. We will not intentionally deprive you of
your ability to use any features of the software or access to your data.
6. We do not warrant that the software or your ability to use it will be uninterrupted or error-free. To the
extent permitted by applicable law, we disclaim any implied warranty of merchantability or fitness for a
particular purpose.
LIMITATIONS ON LIABILITY
7. Your exclusive remedy under the above limited warranty shall be, at our option, either a full refund of
the purchase price (if purchased separately from a camera) or correction of the defective software or
media. To the fullest extent permitted by applicable law, we disclaim all liability for indirect or
consequential damages that arise under this license agreement. Nothing in this agreement limits our
liability to you in the event of death or personal injury resulting from gross negligence, fraud, or knowing
misrepresentation on our part.
GENERAL PROVISIONS
8. If any part of this agreement is found to be invalid or unenforceable, the remaining terms will stay in
effect. This agreement does not prejudice the statutory rights of any party dealing as a consumer.
9. This agreement will be governed by the laws, including Article 2 of the Uniform Commercial Code, of
the state in which you reside. If your state has enacted the Uniform Computer Information Transactions
Act (UCITA) or substantially similar law, said statute shall not govern any aspect of this agreement and
instead the law as it existed prior to such enactment shall govern.
10. This agreement does not supersede any express warranties we made to you. Any modification to
this agreement must be agreed to in writing by both parties.
Additional license information for included executables is provided in the installer.
6
3 Introduc6on
Welcome to Nebulosity 3! Nebulosity began in 2005 with the goal of providing a means
of easily and efficiently capturing images of DSO objects using your CCD camera. It
has grown considerably since then, but has remained true to those goals. It is
designed to be a powerful, but simple to use capture and processing application for
your CCD or DSLR camera. Its goal is to suit people ranging from the novice imager
who wants to create his or her first images and the advanced imager who wants a
convenient, flexible capture application for use in the field. As such, an emphasis has
been placed on easy access to commonly-used camera controls, as nobody wants to
navigate through many menus in order to simply capture a series of images. There is a
lot of power under the hood, but what’s presented on the surface should be easy to
use.
An emphasis has also been placed on compatibility with other applications. For many
imagers, the tools provided here will be well-suited to produce images that are ready to
be touched up in a graphics editing package (e.g., Adobe Photoshop). The tools
provided are the tools most of us want and need to make great images. For more
advanced imagers who already use more sophisticated astronomical image
manipulation software, Nebulosity might serve as a suitable capture application and
provide a set of key processing tools. Nebulosity supports a wide range of output
formats, including various FITS formats and other 16-bit per color formats, so that your
images can be easily imported into whatever software you use.
What Nebulosity is not designed to do is to be an all-inclusive, general-purpose,
observatory control and image capture / analysis package. There are several of these
on the market already and all are fine packages. All are very large, place more
substantial demands on your computer, and, by virtue of being large and all-inclusive,
do not typically present a simple, clean, interface for basic image capture control. The
author of Nebulosity routinely stands in cold, dark fields with a laptop and a camera
taking pictures. Under these situations, when gloves must be removed to operate the
computer, simple, dedicated user interfaces are exceptionally welcome.
That said, the author is also a stickler for power and accuracy. You get quite a few
"serious" tools in Nebulosity. The ones you get are purpose-built - tools that you will
want for processing raw DSO images into beautiful pictures.
One more thing - I’m also a stickler for “fair use” in licensing. Your Nebulosity 3 license
is not tied to one machine or even one operating system. Feel free to use it on a laptop
for image capture and a desktop for processing. Feel free also to upgrade to any new
release in the v3 series.
7
4 What’s New in 3.0?
Version 3 of Nebulosity has brought a number of new features and a lot of work under
the hood to speed up processing, lighten the load on your computer, and be more
stable than ever before. An emphasis has been placed on enhancing usability and on
improving existing features to make them both more powerful and more approachable
in addition to adding new features. It’s easy for software to become bloated and
unwieldy. A goal of Version 3 was to prevent this and to lay a foundation for more
growth to come. Since Nebulosity 2.5 we have:
• The ability to run more than one instance of Nebulosity via a new Launch
another instance entry in the File menu. This will launch another instance of
Nebulosity to let you run more than one camera or to let you work on several
images or projects at once.
•
Creating scripts was not very user-friendly in previous versions. These are very
powerful ways to automate image capture and now there is a new script
creation tool that lets you write scripts far more easily. Just select commands
from the list and double-click to add them to the script or choose from a set of
predefined scripts. If you change the parameter values before double-clicking,
the command and parameter will go in there. If not, just edit the script being
formed to fill in the desired values. No more having to remember all the
commands! Also, if you leave this window with a script in place that you're
working on, it will remember it for next time (so long as you don't quit Neb).
•
In addition to the new tool, we have some new script commands and
revisions to existing commands to make them easier to use.
o Times (like exposure durations) are now in seconds with fractions
allowed. So, if you want 10.5 seconds you can say "SetDuration 10.5".
You should not say the old "SetDuration 10500" (the old msec format).
This parallels the main UI. You can also use an "m" in here to denote
minutes. For example 'SetDuration 2.5m' would be 2 minutes, 30
seconds. '2m30' would also be this.
o SetTEC command added to set the TEC regulation temperature (e.g.,
"SetTEC -10.0")
o SetPHDDither command added to control the dither level on the PHD
link. 0=none, …, 5=extreme
•
Ever want to apply the same processing to another image or set of images?
There is now a new Macro tool in place to do just this. Copy/paste from your
history (or triple-click, Append to Macro in History or just write it manually) and
build up a processing sequence to run on other images.
•
Numerous image adjustment tools and internal bits significantly sped up
through use of all available processors on your machine and more efficient code.
You should notice that pre-processing and image adjustment is a more
responsive, especially on big images and if your computer has several
processing cores. In addition, the memory load for color images has been
8
reduced considerably.
•
Extensive support for internationalization and foreign language support.
Virtually the entire user interface is now setup for other languages. Currently, we
have Dutch, French, German, and Italian!
•
New debayer / demosaic algorithms have been added (a setting in
Preferences controls which one is used). Two simple ones (color binning and
bilinear), PPG (patterned pixel grouping by Alain Desboilles), and AHD (Adaptive
Homogeneity-Directed by Keigo Hirakawa, Thomas Parks, and Paul Lee) are
now added to the VNG (variable number of gradients by Ting Chen) previously
available.
•
Sliders work most of the time for setting values while manipulating images, but
sometimes you want to specify the value exactly. Now, simply double-click on
the value and you can specify the image manipulation parameters directly.
•
Ever want to force an image to be “color” even if it’s mono (e.g., to use in preprocessing color images)? We now have a convert to color tool to do just this.
•
What’s an update without updates to the supported camera list?
o "Atik 3xx,4000,10000" entry changed to "Atik Universal" (Windows).
o The original "ASCOM Camera" has been removed (early-bound interface
that is no longer used in ASCOM as of v5 and v6) and the previous
"ASCOMLate Camera" renamed "ASCOM Camera" (Windows).
o Canon 600D support and updates for T3 and T3i
o The QSI 500 cameras on the Mac shifted to open-source based libftdi for
better stability with the new 600 series cameras. Both have been
renamed “QSI 500/600”
o SBIG has been updated to 4r73B7-based driver for STF-8300 support
•
The main window’s user-interface has gotten a facelift and a number of the
colors are now customizable (see Preferences). In addition, a 10% zoom option
is in place and zooming in and out is faster and more predictable.
•
For Mac users, one change is that the PPC processors are no longer supported
and neither is OS X 10.4 (Tiger). You must be on an Intel, 64-bit capable
processor running 10.5 (Leopard) or later. 32-bit Intel processors (e.g., the
original Core Duos) will work for most, but not all features (e.g., auto-align nonstellar).
9
5 Features
•
•
•
Cross-platform support
o Mac OS X 10.5 - 10.7 (Intel)
o Windows XP, Vista, and 7
o Same license code works on both platforms and all functionality is
available across platforms (Note, not all cameras are supported on all
platforms for image capture.)
Simple, but powerful interface
o All basic controls are present on the main screen. No need to navigate
through lots of menus during an imaging session. Nebulosity was
designed to be easily operated in the field by someone who actually
operates it in an actual field.
o Interface can be customized to show, arrange, or hide tools to suit your
needs.
o Ability to have multiple instances of Nebulosity running for processing
different images or controlling different cameras.
o Support for multiple languages (currently, English, Dutch, French,
German, and Italian)
o By default, all displays are auto-scaled. Any scaling (including inverted) of
the data onto the display possible using easy sliders.
o Histogram gives a quick view of how much of the valid data range is
being used during each capture.
o Pixel statistics / area statistics pop-up window for real-time readout
o Zoom button lets you rescale the displayed image quickly.
o Image pan mode (shift key)
o Unlimited undo/redo (0, 3, or unlimited levels of undo).
o Small clock to show local time, UTC, GMST, local sidereal, Polaris RA, or
current CCD temperature
Capture control
o All basic capture parameters present on main screen. Duration of
exposure, number of exposures per captured series, delay between
captures, name of series, camera gain and camera offset all in one simple
panel.
o 1x1 - 4x4 binning (depending on camera) and fast-readout modes
supported
o Quick Preview button captures one frame with current settings and
displays it on the screen without saving. Helps in focus, composition, and
tuning of capture parameters.
o Frame and Focus mode: Loops a quick, binned image to assist in rapid
initial focus and framing (crosshairs can be enabled or disabled via
Preferences).
o Fine Focus mode with HDR focus metric: Loops a very quick image
around a selected star in full resolution and provides running statistics
(and linegraphs of the history of the statistics) to assist in fine-tuning
focus.
10
•
•
o Support for using LiveView during Frame and Focus and/or Fine Focus on
Canon DSLRs supporting LiveView.
o Capture one-shot color in RAW CCD format or reconstruct color on the fly
– your choice.
o Capture status able to be shown in large red display for easy viewing
when away from computer.
o Link to PHD Guiding to enable pausing guiding during main image
download and to dither location of images.
o Focus information available to other programs (see Preferences, Save
Fine Focus info and CaptureSingle script command).
o Control of both on-board and external filter wheels both in the main userinterface (Tools) and automated via scripts
Multiple file formats supported
o Read virtually any FITS file to process images from virtually any camera
(RGB color, black and white, compressed or uncompressed, any bit
depth)
o Load and process data from FITS, PNG, PPM/PNM/PGM, TIFF, BMP,
JPEG and just about any DSLR “RAW” format.
o Captured data saved in FITS as 16-bits (0-65,535) per color channel, 32bit floating point, or in 15-bits (0-32767) per color channel.
o One-shot color data captures may be saved in RAW CCD format or as
reconstructed full-color images in an RGB FITS format (Maxim / AstroArt
style or ImagesPlus style) or 3 separate FITS files (the latter only for
capture and subsequent use in other programs).
o Captured data saved in either lossless compressed FITS according to the
FITS standard or uncompressed FITS
o These same save formats available for any loaded image, making
Nebulosity serve to convert between many FITS formats (just select your
output format using the settings on the Preferences menu).
o Save current displayed image in BMP or JPG format (24-bit color) as
displayed
o Save current image in 16 bit/color (48-bit color) uncompressed TIFF,
compressed TIFF, or PNG (compressed) format
o Load 8/24 bit PNG, TIFF, JPG, and BMP (scaled to 16/48-bit) or 16/48-bit
PNG and TIFF.
o Batch convert from FITS to 16/48-bit PNG or compressed TIFF
o Batch convert from DSLR RAW and standard image formats (PNG, TIFF,
JPG, and BMP) to FITS
Extensive camera support (capture - Windows-only unless noted)
o Apogee Alta
o Atik 16/16IC/16HR series / Artemis 429/285 cameras
o Atik “Universal” (3xx, 4xx, 4000, 11000, etc)
o ASCOM 5 and ASCOM 6 compliant cameras
o Canon DIGIC II, III, and 4 DSLRs (Windows and OS X): EOS 1100D/T3,
1000D/Rebel XS, 450D/Rebel XSi, 400D/Digital Rebel XTi, 500D/T1i,
550D/T2i, 600D/T3i, 350D/Digital Rebel XT, 60D, 50D, 40D, 30D, 20D/
11
•
•
20Da, 5D Mark II, 5D, 7D, 1D Mark IIV, 1D Mark III, 1D Mark II N, 1D Mark
II, 1Ds Mark III and 1Ds Mark II. Captures are to FITS files with pure
Bayer-matrix data extracted on the fly (or ultra-fast color JPEGs - your
choice). Bulb-mode exposures via ShoeString DSUSB adapter, serial port
adapters, or parallel port adapters.
o CCD Labs Q8-HR and QHY8 (Windows & Intel/Leopard Macs)
o CCD Labs Q285M / QHY2Pro
o FLI
o Fishcamp Starfish (Windows and OS X)
o Meade DSI, DSI Pro, DSI II, and DSI II Pro (Windows and OS X).
o Moriavian G2/G3 (v3 or higher firmware)
o Opticstar DS-335 and DS-335 ICE
o Opticstar DS-336C
o Opticstar DS-142M and DS-142C
o Opticstar DS-145M and DS-145C
o Opticstar PL-130M and PL-130C
o Orion StarShoot Deep-Space Color Imager (original - others via ASCOM)
o QSI 500/600 series (Windows and Mac)
o QHY 8, 8Pro, and 9 (and CCD Labs variants)
o SAC10
o SAC7 / SC1 long-exposure modified webcams / Atik 1&2 - all via the
either a parallel port or via the ShoeString LXUSB adapter for all-USB (no
parallel port) long-exposure imaging.
o SBIG (Windows and OS X)
o Starlight Xpress SXV / SXVF / SXVR USB cameras (Windows and OS X)
o Simulated camera (Windows and OS X)
o Virtually any camera's images can be processed in Nebulosity.
Internal calculations
o All data stored internally in 32-bit floating point per color channel. For
B&W or RAW images, this equates to 32-bits and for color images, this
equates to 96-bits in all math routines. You will never have overflow
(saturation) or overflow or quantization issues as a result.
o Critical math routines computed in double-precision (64-bit per channel)
floating point.
o Routines are optimized for high-speed operation and most will make full
use of however many processing cores you have on your machine.
Image alignment, stacking and pre-processing
o Dark / flat / and bias frame pre-processing tool to let you pre-process
multiple sets of B&W, RAW one-shot color, or color images.
o Create and apply Bad Pixel Maps as an alternative way of removing hot
pixels.
o Align a series of images using simple translation (for equatorially mounted
telescopes) or using sub-pixel level accuracy and translation + rotation
and (optional) scaling (equatorial or alt-az telescopes)
o Drizzle alignment and resolution enhancement for either equatorial
(translation only) or alt-az (translation + rotation).
12
o Colors in Motion: Simultaneous over-sampling alignment and Debayer of
one-shot color images to significantly decrease color error and increase
resolution. For one-shot color imagers, this improves resolution and
reduces color error.
o Average a series of images without alignment (e.g., for combining darks,
flats, bias frames, etc.)
o Standard-deviation based stacking (aka “sigma clip”) of aligned frames to
reduce noise in final stack.
o Adaptive scaling of combined data (stacks) to use full 16-bit range (gives
you the best features of adding and averaging frames).
• Demosaicing, color cameras, and color synthesis
o De-mosaic a RAW one-shot color image using an array of very high
quality debayer routines (VNG, PPG, AHD, Bilinear, and color-binning).
Both interactive and batch-mode supported. Pixels become square in the
process if native pixels were not square.
o White balance on Canon DSLR settings for both stock and extended-IR
cameras
o Square pixels for images from B&W cameras.
o LRGB color synthesis (RGB, traditional LRGB, and color-ratio LRGB)
o Line filter reconstruction for one-shot cameras. Optimized reconstruction
of RAW images taken using line filters. General mode plus modes
optimized for H-alpha and O-III/H-beta on CMYG arrays.
o Convert color-format images to monochrome or monochrome into color
• A host of easy to use interactive image manipulation tools
o Versatile Levels / Power Stretch tool lets you apply not only simple linear
stretching of your images, but non-linear stretches as well. Pre- and poststretch histograms interactively displayed.
o Curves tool to let you customize just how you’ll stretch the image
o Digital Development Processing (DDP). A technique to make CCD images
look more like film images by using a hyperbolic scaling of the data. Here,
the basic technique is enhanced to allow easy darkening of the
background at the same time.
o Star Tightening. A technique to sharpen stars using an edge-detection
algorithm (does not leave the artifacts found in some other techniques).
o Unsharp Mask tool for image sharpening (Traditional and Laplacian image
sharpening also available).
o Color balance adjust (offset and scaling) with real-time 3-color histograms
for easy, accurate balancing. (Automatic color balancing available as
well.)
o Hue, saturation, and luminance adjustment
o Gaussian blurring tool
o Vertical smoothing / deinterlacing
o Adaptive median noise reduction
o Access to GREYCstoration noise reduction
o On any of these, use simple sliders or double-click on the value to set the
numbers directly.
13
• Geometric and mathematical image processing
o Pixel math tool to allow scaling / shifting the image intensities.
o 2x2 binning of images: addition, averaging, adaptive, and low-noise 2x2
for one-shot color sensors.
o Rotation / mirror imaging of images
o Resampling / resizing of images using a choice of 6 algorithms: Box,
Bilinear, B-Spline, Bicubic (Mitchell & Netravali), Catmull-Rom spline, &
Lanczos sinc
o Crop tool (interactive or direct specification)
• General tools
o Processing history continually logged in the History tool
o Notes tool for those bits you just don’t want to forget
o Blink tool to compare images (in Preview / Rename tool).
o Grade a series of images to determine the sharpest / best of the set
o Versitile Image Preview / Rename tool to quickly sift through large sets of
images.
o Image normalization and Histogram matching to balance intensity across
images.
o Measure Distance tool lets you measure the distance (CCD pixels, arcseconds, or arc-minutes) among up to 3 points.
o DSS Preview tool to let you use Digitized Sky Survey images to see what
you can expect to get on your chip.
• Scripting and Automation
o Can write / run scripts to automate captures (interactive and unattended)
using a simple tool anyone can work with.
o Scripts can be dynamically created and sent to Nebulosity either via the
clipboard or via TCP/IP sockets allowing other programs even to control
Nebulosity.
o Image manipulation and processing can be scripted by copying elements
from the History to the Macro tool
14
6 Main Screen
When you open Nebulosity, you are presented with a screen that looks like this
(Windows version is similar):
It has 4 main sections:
1. Image Window (Large black area)
2. Display Panel (Top portion on the right)
3. Capture Panel (Middle right)
4. Status Bar (Bottom)
6.1 Image Window
The image window is where your images will be displayed. It starts off at a default size
(optimized for 1024x768 displays), but is easily resized by simply resizing or
maximizing Nebulosity itself. If an image is too big to fit into the window, the scrollbars
will allow you to navigate around the image.
Alternatively, the Zoom button, located in the
Tip: Want to slide around in an image
Display panel, will resize the image to help
quickly? Try holding down the shift
make it fit your screen.
key and dragging around in the image
to do an accelerated “pan” through
6.2 Display Panel
the image.
Here, you have several controls that affect how
the image is displayed in the Image Window.
Keep in mind that your data are often in 16-bit (or 48-bit aka 16-bit/color) format. That
15
means that you can have 65,536 shades of grey in the image. But, your monitor can
most likely only display about 256 shades of grey (24-bit color). Thus, the data need to
be scaled to display well on your screen. That's the purpose of the first three controls
here.
These are the B and W sliders, and the Auto scale checkbox. The B and W sliders set
the level in your data to assign to black and white respectively. Slide the B slider to the
right and your image gets darker. You've told Nebulosity that a higher image intensity
equals black, meaning more of your data should be dark. Slide it to the left and the
image gets brighter. Likewise, slide the W slider left and the image gets brighter as
more of your data should be white. Put them closer and you have a higher contrast
image. Put them further apart and you have a lower contrast image. Flip sides (white
below black) and you'll invert the image. If you don't want to mess with any of this or if
the image gets way out of whack, select Auto
scale (it's set by default). The Auto scale
Keep in mind that these tools only affect
checkbox tries to set the B and W sliders
the way the image is displayed. They do
automatically by using data from the
not affect the actual data. If you save the
Histogram.
image, adjust the sliders or zoom
control and save it again under a new
If you want to manually set specific values for name, you'll have two identical copies of
B and W, you can enter them in the fields
the same data. (There is one exception
provided (that also read out the current values to this rule. The Save BMP As Displayed
of the sliders). To make the changes take
uses the values in the sliders to help get
effect, press Enter inside the edit box.
your data from 48-bits into 24-bits)
Below the sliders is a Histogram display.
When you first start Nebulosity it is black, but if you load an image or capture an image
(use the Camera Simulator if you don't have one) you'll see a red display in this
window. This box intentionally lines up with the sliders, for the left of the box
corresponds to intensities near zero in your image and the right corresponds to
intensities near the maximum (65,535 for 16-bits) in your image. So, if you see a small
area of red on the left side of the histogram and you're not seeing anything on the
screen, it means that you have a faint signal in the image. Slide the W slider to the left
to come near that small area of red and you'll see your faint image.
The Histogram is a very powerful tool in image capturing, for it tells you a lot about
your image. Are all of the data far to the left? If so, your entire image is faint and you
should increase your exposure or gain if possible (see below). Do you see a nice curve
that trials off to the right just before you get to the edge of the Histogram? If so, you've
got a nice exposure and are making the most of your data. Do you see that instead of
trailing off smoothly near the right edge, the curve ends abruptly at the right edge? If
so, you're saturating a lot of the pixels in your image and should likely use a shorter
exposure or less gain. Are you cutting of hard on the left edge? If so, use more gain,
more offset, or a greater exposure duration.
16
Finally, the panel has the Zoom button (marked "100%" by default). Repeated clicks
on the Zoom button will cycle through several zoom modes (20%, 25%, 33%, 50%,
100%, 200%, & 400%) to get a better view of your image. Next to this, you'll see + and
- buttons that let you zoom in and out respectively. Note again, this only affects how
you see your image, it does not change the underlying image
itself.
You can use Ctrl +
and Ctrl - (or Cmd
For a more detailed inspection of your image, try activating the
+ and Cmd - to
Pixel Stats pop-up window (under the Image menu).
zoom in and out.
6.3 Capture Panel
The main Capture Panel has several sub-sections. At the top, we have an area that
controls connection to the camera and advanced settings for the camera. Below this,
we have an area that lets you control details of the exposure and below this we have a
number of buttons that let you take various kinds of exposures.
6.3.1 Camera Sec5on
The Camera section contains a pull-down to select your camera model. When you pull
down your camera model, Nebulosity attempts to connect to the camera. Success of
failure will be noted in the left-hand panel of the Status
Bar.
Too many cameras listed there
to sort through each time? In
If you're new to CCD imaging and don't have a camera
the Edit menu, you’ll find an
yet or want to explore some of Nebulosity without
option to De-select cameras
attaching your camera, a Camera Simulator is
and remove them from the
provided as one of the camera choices. The camera is
list. Don’t worry - you can
always aimed at the same patch of sky (that happens
always add them back in later.
to have 20 stars of different brightness) but the mount
isn't perfect, so you'll notice the stars move a bit from
image to image. The camera has noise, and responds to all the controls in the
Exposure Panel, letting you get a feel for what to expect and how to use the program.
Here, you will also find an Advanced tab. Nebulosity picks default values of a number
of camera options that are optimal for most DSO imaging. However, if you want to
select any of these yourself, you can do so in the dialog box that appears when you
click this button.
6.3.2 Exposure Sec5on
Here, you have controls for all basic exposure options.
• Duration: How long per image (in seconds) should the exposure be? Note,
fractions like 1.5 allowed.
• Gain (optional): Some cameras let you adjust the gain and offset of the A/D
converter. This entry controls how much CCD amplifier gain should be used
during A/D conversion. (Think of gain as a volume knob for the signal coming off
the CCD). Numbers range from 0-63.
17
•
•
•
Offset (optional): What offset should be added to the signal during A/D
conversion? (The offset adds signal into every pixel to help you keep the pixels
from having zero values anywhere). Numbers range from 0-255. (See Automatic
Offset on p. 16)
# Exposures: How many images do you want to take?
Time lapse: How much time (seconds) should be inserted between each
image?
Most of these are fairly self-explanatory, but Gain and Offset deserve a bit of attention.
They get this in the Section Taking Good Images. For now, you can leave them at their
default values.
Want to set your CCD’s TEC?
The Duration and Time Lapse entries allow you to
The Advanced button and an
specify the exposure duration in seconds, but
entry in Preferences let you do
fractions are allowed. So, if you want an exposure of this (as does any camera-specific
a half a second, simply enter “0.5”. Remember that
tool). Want to know the current
a millisecond is a thousandth of a second (0.001). In temperature? In Preferences,
addition to allowing you to enter the time directly, the select CCD Temperature under
Duration control lets you pull down any of a number “Clock / TEC display”
of common times. The word “Duration” is actually a
button. Click and hold on it and a list of common
times will appear that you can quickly select without having to type numbers in while in
the dark.
6.3.3 Capture Sec5on
In this panel, you'll find the Preview button. This button takes a single image at
whatever duration, gain, and offset you've specified and shows it on the screen. It does
not save the image. This lets you fine-tune the composition of your image and hone in
on correct focus of your telescope. It also lets you determine the optimal duration,
gain, and offset. (Use the handy Frame and Focus button for rough focus and
composition).
The default directory is located in "My
There are three controls used in
Documents" (Windows) or “Documents” (OS X) in a
capturing a Series. A text entry folder called "Nebulosity". If you use the default
box near the bottom lets you
directory and it doesn't exist, Nebulosity will attempt to
set the default Name for the
create it. If you forget to set the directory you actually
series and a button lets you
want to use and capture a night's worth of data, this is
select the Directory the data
where it is. If you use a different directory and pull
will be saved in. Finally, at the
down Save Preferences from the Preferences menu, the
top of the panel is the Capture
current directory will be saved as the default
Series button. This starts the
sequence acquisition process.
For example, if you've setup for 10 exposures of 20 seconds to be stored in My
Documents\Nebulosity\August_20_2011 and called M51, Nebulosity will loop and take
all 10 exposures. The first will be called M51_1.fit, the second M51_2.fit, etc. At the end
18
of the capture, you'll hear the Windows Ta-Da! sound play. (To abort a sequence, press
the Abort button).
Three things to note concerning series captures:
1. If you provide a name that already exists (e.g., you hit Capture Series again
without changing the name), Nebulosity will create a new name to use in saving
the series. Here, it would be M51-1_1.fit, M51-1_2.fit, etc. Hit it again and you'll
get M51-2_1.fit, etc.
2. If you need to abort a series during the capture, press the Abort button in the
Camera panel (or click the mouse inside the Image Window and press the
ESCAPE key.)
3. The format the files are saved in is based on your choice in the Preferences
menu.
Finally, you will also see three buttons:
Frame and Focus, Fine Focus, and Abort.
The Abort button works in a number of
Frame and Focus is a useful tool for
places - during capturing, frame/focus,
composition of images and for obtaining
fine focus, alignment, etc. On several
rough focus. Press this button and the
cameras aborting can take several
camera will enter its most-sensitive, fastest
seconds to clear and reset the camera.
mode and continually loop exposures. This
gives something of a "live video" display,
showing you an image as quickly as possible (it may still take several seconds to
update, depending on the camera). Adjust your focus, move your telescope, etc. until
you have a reasonable image and then press Abort to cancel the automatic looping.
Once you have a basic focus and framing of your shot, you'll likely now want to use the
Fine Focus button to fine-tune your focus (not available on all cameras). When you
click on this button, you're asked to click on a star. This can be either from the last
Preview or from the last exposure in the Frame and Focus routine. When you do so, the
image will now continually display the area centered on that star in full resolution. Use
this to fine-tune your focus.
Focus can be achieved visually by looking
for the sharpest image while adjusting your
telescope's focus or by using the focus aids
provided. Three additional aids are given to
help you reach focus. The first of these is
also a visual aid. To the right of the star you
will see a profile of the star. When in sharpest
focus, this profile will be at its narrowest and
tallest.
The other two use calculated metrics to try
to determine how good the focus is. The first
During Frame and Focus and Fine Focus,
you can adjust a number of parameters
on the fly. You can alter the exposure
duration, gain and offset and you can also
turn on and off Auto-Ranging and adjust
your sliders. The effect of each won't be
seen until the next image appears,
though. You can also pause the looping
by hitting Ctrl-Space. Hitting this again
will restart the process.
19
uses the fact that when the star is in focus more light is hitting the center-most pixel
leading to a brighter value in that pixel. Therefore, the maximum value recorded in the
area should reach its peak when the image is in focus. The “Max” reading and the red
line in the graph in the lower left show the current value and history of this value.
The second metric calculated is the “Half Flux Radius”
or HFR. This is a metric devised by Larry Weber and
used in his popular Focus Max plug-in for several
packages. This is an excellent metric and is quite
possibly the most robust metric we have. In it, the best
star in the small region is first found and its center is
found. The total star flux is then found and the radius
of a circle around the star’s center that would contain
half the total flux is calculated. This is the HFR.
In the lower left the history of values for both the Max
and the HFR are plotted. The most recent 100 samples
are plotted so you can watch how the focus quality
changes as you adjust you telescope's focus knob.
This graph will auto-scale itself if the range is too large
or too small for the display. Finally, also shown on here
are the best values achieved during this Fine Focus run for both measures (horizontal
dotted lines).
6.4 Status Bar
At the very bottom of the screen is the Status Bar. Nebulosity gives you a lot of
information down there. The Status Bar is divided into 4 panels. The right-most panel
always tells you what Nebulosity is doing. It may read "Idle" (it's not doing anything),
"Capturing", "Processing", etc. Next to that, is a panel that shows you the X and Y
location of your cursor and the intensity of the image at that pixel (see the Pixel Stats
pop-up window under the Image menu to provide more detail).
The left two panels are used for information and instructions concerning what
Nebulosity is doing. Load an image and you'll see its dimensions and the name come
up here. Start an image alignment process and you'll get instructions and progress
here. Start an image capture and you'll also see your progress down here, along with
what file was just saved. When in doubt about what's going on, check the Status Bar.
6.5 CustomizaAon of the Interface
Nebulosity gives you the ability to customize the user interface to suit your needs. The
Display and Capture panels that come up by default can be closed or “torn off” by
simply dragging them off of the main window. Many components can be re-arranged
and additional components can be added. These other components are available in the
View menu. For example, here, we have replaced the normal Capture Panel with a
20
more
compact
version, the
Mini
Capture
Panel.
We’ve also
put the
Notes tool
above the
main image
area and
have a
specialized
control for
the QSI
cameras
there as
well. Not
everything
needs to
be
“docked”
to the main Nebulosity window. For example, the dialog that controls the link to PHD
Guiding is seen here floating above the main Nebulosity window.
6.5.1 Notes and History
Ever wish you could jot down some information about the series you’re capturing?
Something like “scope dewed up somewhere in the middle of the red frames” or
“forgot the right spacer for the reducer on these” or even just to record the more
mundane settings about the night’s progress? The Notes tool gives you handy place to
do this. Your text is saved as plain text so any program can read it.
Also, ever wish you could remember just what settings you used as you processed an
image for some step 12 steps ago? Open the History tool from the View menu and
you’ll see a running log of exactly what you did.
6.5.2 Camera-­‐specific Dialogs
The main interface of Nebulosity lets you control the basic features of all cameras with
the same interface, but some cameras have more features. For example, you may have
a filter wheel attached to the camera (also controllable via scripts) or the camera may
let you control the shutter (to make it easier to take dark frames). You’ll find cameraspecific control dialogs here in the View menu as well.
21
6.5.3 External Filter Wheel
If you’ve got an ASCOM-compliant filter wheel (Windows) or a Starlight Xpress filter
wheel (Mac OS X), Nebulosity can connect to it and control it in the same way it
controls filter wheels built into QSI, SBIG, or QHY cameras (see above). This control
will let you select the desired filter. Keep in mind, you can also script your captures,
telling Nebulosity to change filters as needed.
6.5.4 Pixel Stats
The Pixel Stats window will let you see the image intensity under the mouse pointer
(like you also see in the Status Bar), but it will also show you a lot more. For color
images, it splits this into the separate the R, G, and B values. For all images, it also
shows local statistics and statistics on the whole image. Mouse over a star and it’ll tell
you the HFR value for that star.
6.5.5 Link to PHD Guiding
PHD Guiding is the popular freeware guiding package by Stark Labs. With it, guiding
can be as simple as “Push Here Dummy.” With the PHD Guiding dialog, you can
establish a link to PHD Guiding so that the two packages can talk to each other. This
gives you two powerful features. First, it lets you pause guiding during the download of
your main image. Some cameras are sensitive to interruptions on the USB bus and if
the guide camera shares the USB bus, the image quality can be degraded. The “Pause
during download” option will let you enable this feature.
The second feature is to enable “dithering” of the image’s location across frames.
Between frames, Nebulosity can send a signal to PHD to tell it to move the “lock
position” (the position of the crosshairs in PHD) by a small, random amount. Once PHD
has moved the star and re-established stable guiding in the new location, a signal is
sent to Nebulosity to let it know it can continue with the next frame in the series.
To do this, you must:
1. Have PHD 1.8.6 or later running and tell it to “Enable Server” in the Tools menu
(PHD will remember if you last left the server on and restart it the next time you
start PHD). If Windows asks you whether it’s OK to do this, tell it yes.
2. In Nebulosity’s PHD Guiding dialog, click on the “Connect” button (Note: Shiftclick on the connect button and you can use a different TCP/IP port).
3. In the dialog, tell it how much “dither” to send. The dither sent will be a random
number of pixels in X and Y, scaled by the level you pull down here. In the lowest
level, the random numbers will vary from -0.5 to 0.5 pixels and in the highest
they will vary from -1.5 to 1.5 pixels in the guide frame. Since people typically
guide at shorter focal lengths than they image at, this will usually have a much
larger effect in your main images.
4. It may be worth changing the “Settle threshold”. This specifies how far off the
star can be from the lock position before PHD sends the message that it has
gotten the star back on target and it is OK to resume your series capture.
22
6.5.6 Macro Tool
If you’ve ever wanted to replay the processing you did on one image and apply it to
another image, the Macro Tool is for you. This kind of thing is very handy for
processing mosaics and can be useful if you want to re-create a state of an image (so
long as you have the output of the History tool). To use the Macro Tool, simply copy
and paste items from the History dialog into the Macro Tool window (or select them - a
triple-click works well here - and hit the History Tool’s “Append to Macro” button).
When you’ve got the steps you want in the order you want, hit Run and Nebulosity will
process the image accordingly.
You can, if you like, create the Macro Tool entries from scratch. Use previous entries in
the History tool to see what the commands are supposed to look like.
23
7 Capturing Images
Most of what you need to know to capture images was covered in the previous section
on the Exposure section of the Capture Panel. There are a few topics worth
considering on their own, however.
1. Monochrome vs. Color?
2. One-shot color: RAW vs. RGB?
3. File formats
4. Camera Gain and Offsets
7.1 Monochrome vs. Color Cameras
Monochrome cameras have CCD pixels that have no filter placed in front of them. Light
simply hits the CCD array and the intensity gets recorded and saved. The CCD and
Nebulosity don't care in the slightest whether you have no filter in place, an IR filter in
place, a red filter, an Ha filter, or any combination thereof. To the camera and to
Nebulosity, it's all black and white data that comes straight off of the CCD as every
pixel operates just as every other pixel.
One-shot color cameras are a different story altogether. Oneshot color cameras have tiny color filters placed over each
CCD pixel. Typically, red, green, and blue filters are used
(although other options for filter sets exist). For example, if one
looked at a small 4x4 pixel patch of the CCD, one might see
the arrangement shown on the right. Each pixel on the chip
codes for only one color. So, if you have 1 million pixels, you
have 500,000 green, 250,000 red and 250,000 blue pixels
(CCD makers over-emphasize the green since our eyes are most sensitive to green).
This is why you may hear people say that one-shot color imagers have less resolution
than monochrome imagers.
To some degree, this is true. Yet, when you look at a digital photograph from a digital
camera, you don't see this array of colors and you don't see a low-resolution shot.
Digital cameras use this same kind of one-shot color CCD but produce crisp, full-color
images with as many pixels in the output (each pixel having values for red, green, and
blue) as they have pixels on the chip.
Whether the way this works is black magic or math is up for you to judge, but there are
very good techniques for turning images from this "Bayer" matrix into a full-resolution,
full-color image. This conversion is called "De-Bayering" or "De-Mosaicing" the raw
CCD image. Depending upon the sophistication of the technique, the end result can be
as poor as having resolution of one fourth the pixel count or as good as having nearly
as good resolution as the full pixel count. In general, the "luminance" or "brightness"
resolution is almost as good as a monochrome CCD, while the color resolution (the
ability to rapidly change between red, green, and blue) is not as good, with techniques
differing in just how much is lost. Fortunately, while intensity in both daylight and
astronomical images can change very suddenly in an image (as we go from a black
24
background to a star), the hue (or color) changes much more gradually. Thus, we can
"get away" with having less color resolution than we have intensity resolution.
It is for this very reason that even when using monochrome CCDs, imagers often shoot
a luminance channel at full resolution and color channels at lower resolutions (by
"binning" their CCDs to increase the signal to noise ratio but decrease the resolution).
Thus, low color resolution but high intensity resolution is often chosen by monochrome
CCD imagers, narrowing the potential difference between the quality of the output
between the two CCD types.
7.2 One-­‐shot color cameras: Should I capture RAW or RGB?
Nebulosity lets you capture and save images from one-shot color cameras either in the
RAW format from the CCD (where pixels still follow the Bayer pattern or whatever
pattern is on your CCD) or in full-color RGB format. Here, by “RAW”, we don’t mean
CR2, NEF, CRW or any other of the formats used in DSLRs. By “RAW”, we simply
mean the pure, raw, unprocessed, unadulterated image data from the sensor of a oneshot color camera. All DSLRs are one-shot color cameras and all use the Bayer matrix
described above. The data from this matrix, as noted above, is monochrome
(grayscale) data. Before converting this into color, it is “RAW”. (This RAW data could
hypothetically be stored in CR2, NEF, etc. format or it could be stored in FITS, PNG,
TIFF, or any other format - it’s still the same data from the sensor.)
You have a choice in Nebulosity whether to keep the data in this pure RAW format on
capture or whether you want to convert it to color. Many people instinctively want to
convert it to color (since their camera is a color camera). This is an option and if you
select this in your Preferences (Capture, Acquisition mode), Nebulosity first captures
this raw data from the CCD and then applies a de-mosaic function to convert it into a
full-color image. This full-color image is then saved and the raw data are lost.
On-the-fly conversion to RGB is perhaps the simplest and most intuitive format for the
user. You ask for a full-color image and you get it. Many fine images are created this
way, but it does have a few drawbacks. First, each image is 3x as large as a RAW
image, taking up 3x as much space on your hard disk. Second, on-the-fly de-mosaic
takes some amount of time for each image. Thus, if your capture machine is a lowerend machine, you may want to capture in RAW and convert to RGB later.
Finally, RAW capture has one more advantage. Dark frame, bias frame, and light frame
pre-processing is somewhat more accurate at fixing images in RAW mode than in RGB
mode. In addition, if you capture in RAW format you can use the powerful Bad Pixel
Map tool, which must be used prior to the de-mosaic process (see 5.3 Bad Pixel
Mapping). For these reasons, it is better to capture your one-shot color data in
RAW format and convert it later.
If you do choose to save the data in RAW format and not convert on the fly, you will
pre-process your images in B&W / RAW mode and de-mosaic all of the pre-processed
25
images prior to stacking (otherwise, you'll put red pixels atop green pixels, etc. and
lose all hope of making a final color image).
7.3 File formats
Nebulosity can read just about any valid FITS image file out there (it makes extensive
use of NASA's FITS library) and can write images in a range of useful FITS formats. The
format it will write in is set by your choices in the Preferences menu. This is true not
only for captures but for any time you pull down "Save" from the File menu (thus letting
Nebulosity act as a FITS format converter).
For color images, you have several options. RGB FITS is the default. Here, a single file
holds the red, green, and blue data after the image has been converted into a full-color
image (de-mosaic). Unfortunately, there are two ways in which other programs have
chosen to implement RGB data in FITS files. The differences are esoteric to most (and
concern using 3 HDU's vs. using 3 axes) until one realizes that programs using one
standard don't generally like files written by the other standard. So, Nebulosity will not
only read both formats just fine, but it'll write either of them. They're labeled RGB FITS:
ImagesPlus and RGB FITS:
Maxim / AstroArt.
Right click on a .fit file in Windows and select "Open
With" and "Choose Program". Browse to Nebulosity
In addition to this, Nebulosity will
(c:\Program Files (x86)\Nebulosity3\Nebulosity3.exe)
write three separate FITS files for
and select "Always use the selected program". Or, on a
a full-color image if you so
Mac, right-click on it and pull down Get Info. Under
desire. One will have the red
Open With, select Nebulosity and then click the
data, one the green, and one the
Change All button. Now, double-clicking on .fit file
blue. This is a far more
will automatically start Nebulosity and load the image.
cumbersome way of dealing with
the data and unless you have a
very good reason to do this, odds are you shouldn’t do this.
Nebulosity can save in a compressed FITS format to save space. The compression
algorithm used is native to FITS and is a lossless one. You're doing no harm to your
data by using it. If you don't wish to use compression (e.g., you wish to use a program
that doesn't support it), simply uncheck this in the Preferences menu. (Note: Maxim DL
uses a "compressed FITS" format that is proprietary and not the standard FITS
compression. Nothing outside of Maxim DL can read this format and Maxim doesn't
seem to always like FITS' native compressed format.)
If space is not a concern and you want to absolutely maximize the quality of the saved
data, you can choose to save the data in 32-bit floating point format. This is the native
format used internally. Data files will be twice as large and, in truth, will likely show little
more than the default of saving in 16-bit integers.
Finally, you can choose to rescale your data to 15-bits rather than the full 16-bits
possible. Thus, your data will be scaled into the range of 0-32767 rather than 0-65535.
This is an option to support several programs.
26
Suggested settings if you plan to use other applications as well
AstroArt
16-bit, RGB FITS Maxim/AstroArt, uncompressed
ImagesPlus
16-bit, RGB FITS ImagesPlus, compressed
Iris
15-bit, 3-separate files, uncompressed or PNG/TIFF.
Maxim DL
16-bit, RGB FITS Maxim/AstroArt, no compression
Photoshop
File, Save 16-bit/color TIFF or PNG
FITS is used as a standard not only because it is so common in the astronomical
community, but also because it allows for arbitrary information to be stored along with
the image. So, Nebulosity stores information such as the time the image was captured,
what camera was used, what exposure duration, gain and offset were used, etc. along
with the image.
That said, many graphics programs do not support reading of FITS images. Here, you
have two options. First, you can save an image as displayed (i.e., taking into account
the B and W slider positions) in 24-bit BMP or JPEG format. If you do this, try to do
most of your processing beforehand as this format will allow for only 8-bits of
information for each color channel. Subtle gradations will be lost when you do this (but
remember, your monitor will only display 8-bits per color anyway).
Second, you can save in 16-bit/color (aka 48-bit color) TIFF or PNG format. Both
compressed (LZW) and uncompressed TIFF formats are supported (PNG format is
always compressed). These options all provide ways of saving your data without any
loss or degradation for use in other programs. These also are excellent ways to get
color images into programs like Iris v5.
Finally, you can load both 8-bit/color (24-bit) and 16-bit/color (48-bit) images from a
number of formats. 8-bit JPEG, BMP, TIFF, PNG, and TGA files can be loaded and will
be automatically stretched to 16-bits/color. 16-bit TIFF and PNG can be loaded as well.
7.4 Camera Gain and Offset
While many cameras have their gain and offset set by the manufacturer, some let you
adjust these parameters. For cameras that do, Nebulosity will include these fields in the
Capture panel. My advice is to set it once and forget it. To understand why, we can
first go over just what gain and offset do. If you want a good bit more on the details of
what these are, skip ahead to the section on this in Taking Good Images.
7.4.1 How should I set my gain and offset to set it and forget it? The best value for your camera may not be the best value for other cameras. In
particular, different makers set things up differently. For example, on a Meade DSI III
that I recently tested, running the gain full-out at 100% let it just hit full well at 65,535
27
ADU. Running below 100% and it hit full-well at 40,000 or 30,000, or 10,000 ADU.
There's no point in running this camera at anything less than 100% gain. On a CCD
Labs Q8-HR I have, even at gains of 0 and 1 (on its 0-63 scale), the camera would hit
65535 on bright objects (like the ceiling above my desk). There's no point in running
this camera at gains higher than 0 or 1.
Why is there no point? The camera only holds 25k e-. If a gain of 0 or 1 gets me to 0.38
e-/ADU (so that those 25k e- become 65535), running at 0.1 e-/ADU will only serve to
limit my dynamic range. Each single electron already comes out to more than 2 ADU.
So, to determine the gain and offset to use:
1) Take a bias frame and look for the minimum value in it. Is it at least, say 100 and less
than a thousand or a few thousand? If so, your offset is fine. If it's too low, boost the
offset. If it's high, drop it. Repeat until you have a bias frame with an offset in, roughly
100 - 1000. Don't worry about precision here as it won't matter at all in the end. You
now know your offset. Set it and forget it. Never change it.
2) Aim the camera at something bright or just put it on your desk with no lens or
lenscap on and take a picture. Look at the max value in the image. Is it well below 65k?
If so, boost the gain. Is it at 65k? If so drop the gain. Now, if you're on a real target
(daylight ones are great for this) you can look at the histogram and see the bunching up
at the top end as the camera is hitting full-well. Having that bunch-up roughly at 65,535
plus or minus a bit is where you want to be. If you pull up just shy, you'll get the "most
out of your chip" but you'll also have non-linearity up there. You've got more of a
chance of having odd color casts on saturated areas, for example, as a result. If you let
that just clip off, you've lost a touch but what you've lost is very non-linear data
anyway (all this assumes, BTW, an ABG chip which all of these cams in question are).
Record that gain and set it and forget it. Never change it.
By doing this simple, daytime, two-step process you've set things up perfectly. You'll
be sure to never hit the evil of zero and you'll be making your chip's dynamic range fit
best into the 16-bits of your ADC. Again, all the cameras in question have full-well
capacities below 65,535 so you are sure to have enough ADUs to fit every electron you
record into its own intensity value.
28
8 Overview of Image Processing
You've taken your images and are now comfortably inside. Now what? How do you get
all those raw frames to look like a nice pretty stack? Just what the heck is Bad Pixel
Mapping? Should I try Drizzle?
The rest of the manual provides answers to many individual questions and documents
each of the tools in more detail. The goal of this section is to let you see how all of
these fit together and to give you the necessary information to choose a path through
the initial processing of your data. This alone won't give you a full understanding of
how each tool works (see the individual section for each tool), but it should help put all
the pieces together. In addition to the coverage here, there are several tutorials and
walk-throughs available on the web.
The basic steps are as follows:
1. Decide on how you’ll deal with hot pixels (BPM vs. Dark subtraction)
2. Prepare and any sets of darks, flats or bias frames to remove hot pixels,
vignetting, etc.
3. Convert RAW images into color via Demosaic (if one-shot CCD used and
captured in RAW, which you really should do) and square-up your pixels (if
needed)
4. (optional) Normalize the images
5. (optional) Grading and Removing Frames
6. Stack the images (Align and Combine)
7. Crop the image to clean it up
8. (color only) Remove skyglow hue
9. Stretch the image (Levels, DDP, etc)
8.1.1 Deciding on Bad Pixel Mapping vs. Dark Subtrac5on
Both of these techniques are designed to deal with the thermal noise inherent in your
images and the resulting "hot pixels" that show up in the same spot on the image in
each frame. Dark subtraction is the traditional way of doing this. It works by simply
subtracting the value for each pixel in your "master dark" from the value of that pixel in
each light frame. If your light frames and dark frames were taken with the same
exposure duration and at the same temperature, dark subtraction will remove the hot
pixels (and "luke-warm" pixels as well - any thermal noise, not just the brightest). This
can work very well if you control the temperature, exposure duration, and take a lot of
dark frames. If you don't do these, you can end up with "holes" in the image (black
spots where the hot pixel used to be), incomplete hot pixel removal, and you can inject
noise into your light frames (see above).
Bad Pixel Mapping works differently. You first create a "Bad Pixel Map" (Batch, Bad
Pixels, Make Bad Pixel Map) using a dark frame or stack of dark frames. A slider
appears to let you set a threshold (feel free to use the default). Values in the dark frame
that are above the threshold say "this pixel is bad". Bad pixels, and only bad pixels are
fixed in your light frames by using surrounding good pixels to help fill in what this pixel
29
should have been. For many cameras (in my experience, the cooled cameras with Sony
sensors work wonderfully), this is an exceptionally powerful technique as the hot pixels
are removed effectively with no noise being injected. It's also very flexible as you can
use the same "master dark" from night to night and from exposure duration to
exposure duration just by adjusting the slider and making new maps as needed.
Note: If you use Bad Pixel Mapping you will not use Dark Subtraction and vice versa.
One or the other but no need for both. If you use Bad Pixel Mapping you can still use
flats and bias frames and it doesn't matter whether you apply BPM before or after your
other pre-processing. BPM just locally fixes the bad pixels and doesn’t touch any nonbad pixels. In fact, using both flats and bias frames is a good thing.
If you’re using Dark Subtraction, keep in mind that your dark frames already have the
bias current in them. You will not typically use both darks and biases as a result. If you
do, you’ll end up removing the bias twice (which injects the inverse of the bias back in).
So, don’t do that.
When using flats, it can be a good idea to let Nebulosity smooth the flats a bit for you.
At the very least, if using a one-shot color camera, use the 2x2 mean or CFA scaling
option in the pre-process tool. I often use a 7 pixel blur. This reduces the noise in the
flats and makes for smoother images.
8.1.2 Applying the darks, flats, and biases
Odds are you’ve taken some number of darks, flats, and/or bias frames (control
frames) in addition to your lights. What the heck do you do with them? Your goal is to
combine these control frames as needed (to reduce the noise in them) and use them to
fix the problems in your lights. Your lights will have not only the signal from your DSO
in them, but they will have dark current and bias current and will probably not be
perfectly evenly illuminated. Pre-processing is the step that applies these control
frames to your lights.
Before we apply them, though, we typically want to stack (or combine) multiple control
frames. Having more than one dark, flat, and/or bias frame is a good thing as each
individual frame has both the artifact you want to remove from your lights and random
noise. Stack a bunch of these together and the random noise goes away leaving you
with a clean image of the artifact you want to remove. Use just one and you remove the
artifact and whatever random noise that one frame had. Since it's random noise won't
be the same as the random noise in your image, using just one dark, flat, or bias will
actually inject noise into your light frame and make it noisier. This is why people take a
good number (20-100) of each of these.
In Nebulosity, you can stack them all first (to make “master” control frames), or you can
stack these on the fly while correcting your lights. This all happens in the Pre-process
30
image sets tool.
Prior versions of Nebulosity had a
separate tool to pre-process a single set
of lights with pre-stacked bias / dark /
flat frames. This has been removed as
the Pre-process image sets tool can do
anything the old one could do and more.
This tool is the same as the Multi-set
pre-process tool found in 2.3 and later.
This tool allows you to pre-process 1-5 sets
of lights using 0-3 sets of darks, 0-2 sets of
bias frames, and 0-5 sets of flats. If you pass
in more than one control frame (flat, dark, or
bias), it will stack it on the fly for you. If you
want to apply a bias frame or a dark frame to
your flats, it can do that too. Got data from a
mono camera with 5 filters each with its own
flat and with a few different exposure durations (and therefore a few different darks)?
No problem.
Using the tool is covered in more detail in the Multiple-Set Pre-Processing section, but
can be quickly summarized as follows:
1. Pull down Batch, Pre-process image sets
2. Click on the buttons to define your “control sets”. For example, click on “Dark
1” to select one or more dark frames that you can now refer to as “Dark 1”.
3. (optional) If you want to do things to your flats on the fly, tell it which bias and
dark to apply (both are entirely optional) and whether you want to blur the flats
at all. If you’re using a one-shot color camera, you will at least want to use the
2x2 mean here to remove the Bayer matrix. Blurring the flats will help reduce
the noise (grain) in them.
4. Click on the buttons to define your sets of light frames (“Light 1” through “Light
5”). You can have 1-5 sets of lights that you work on at once.
5. Select which biases, darks, and flats get applied to each of the sets of lights.
6. If this all made no sense, see the Multiple-Set Pre-Processing section where it is
laid out in more detail.
8.1.3 Using Bad Pixel Mapping
Let’s say you chose not to use traditional dark subtraction to take care of those hot
pixels and instead want to use bad pixel mapping. You still will probably apply flats
and/or biases (bias frames are good to use if you’re using BPM). To apply BPM to your
light frames:
1. Create a Bad Pixel Map if you don't already have one. Batch, Make Bad Pixel
Map. Select a dark frame or stack and start off by just hitting OK to use the
default threshold. If for some reason you don’t like that threshold, adjust the
slider to let you remove more or fewer pixels.
2. During pre-processing, click one of the Dark buttons and load the Bad Pixel
Map you created in step 1.
3. For that entry, go to where it says “Dark Subtract” and pull down one of the
BPM options. If you’re using a mono camera, use “BPM Mono”. If it’s a oneshot color camera but you’ve not demosaic’ed yet, use “BPM Color”. If you’ve
converted to color already, go back and read why I suggest you use RAW and
don’t convert to color yet.
31
8.1.4 Conver5ng RAW images to Color and/or Pixel Squaring (aka Reconstruc5on)
The last step before stacking your images is to convert them to color (if they are from a
one-shot color camera and you captured in RAW) and square them up as needed.
Some cameras have pixels that are not square and this will lead to oval rather than
round stars. The process of demosaic'ing (color reconstruction) and/or pixel squaring is
called Reconstruction in Nebulosity and the details of this can be found in the section
on Reconstruction: Demosaic’ing and Pixel Squaring.
Note, you can tell if your images need to be squared up by pulling down Image, Image
Info. Near the bottom you will see the pixel size and either a (0) or (1). If it is (1), the
pixels are square. Of course, the pixel dimensions will be the same in this case too.
To reconstruct all of your light frames, simply:
1. Pull down Batch, Batch Demosaic + Square (if images are from a one-shot color
camera) or Batch Square (if images are from a monochrome camera or you just
feel like squaring up a color cam's but keeping the image as monochrome for
some reason).
2. Select your frames
3. Ideally, Nebulosity will start loading and reconstructing the frames. If it pops up
a dialog asking for things like offsets, it means it did not recognize what camera
captured the image (or you have “manually override color reconstruction”
checked in the Preferences). If this happens, consult the Reconstruction:
Demosaic’ing and Pixel Squaring section.
4. In the end, you'll have a set of images named "recon_OriginalImage.fit"
8.1.5 Normalize Images (op5onal)
All things being equal, your 50 frames of M101 should all have the same intensity. They
were taken on the same night one right after the other and all had the same exposure
duration. So, they should be equally bright, right? Yes, but there's that nagging "all
things being equal" we supposed and, well, all things aren't always equal. For example
if you start with M101 high in the sky and image for a few hours it starts picking up
more skyglow as the session goes on, brightening the image up. That thin cloud that
passed over did a number on a frame that still looks good and sharp, but isn't the
same overall intensity as the others, etc. All things are not always equal.
If you're doing the Average/Default method of stacking, you need not worry about this
issue unless the changes are really quite severe. If you're using standard-deviation
based stacking, Drizzle, or Colors in Motion, it is a good idea to normalize your images
before stacking. What this will do is to get all of the frames to have roughly the same
brightness by removing differences in the background brightness and scaling across
frames.
There are two methods you can use to normalize images. The first (and original),
performs a purely linear stretch to put the black and white points in roughly the same
place. To normalize a set of images, simply:
32
1. Pull down Batch, Normalize images
2. Select the light frames you want to normalize
3. In the end, you'll have a set of images named "norm_OriginalName.fit"
The second, more advanced tool, attempts to equate the histograms of two or more
images. This is a more complex stretching procedure that can account for more kinds
of changes across images. To use this Match Histogram tool, simply:
1. Pull down Batch, Match Histograms
2. Select a reference frame and press OK. This will serve as the template image
that others will be matched to.
3. Select the set of frames you wish to normalize
4. In the end, you'll have a set of images named "histm_OriginalName.fit"
8.1.6 Grading and Removing Frames (op5onal)
Sometimes bad things happen. The tracking goes awry, a breeze blows, you trip over
the mount, etc. This is a good time to find those "bad" frames and pretend they never
happened. There are two tools to help you here, covered briefly here and in more depth
in Previewing and Grading Images.
8.1.6.1 Grade Image Quality
This will look at a set of frames and attempt to automatically grade them as to how
sharp they are relative to each other. The idea here being that you'll not use the least
sharp frames. Pull down Batch, Grade Image Quality and point it to your light frames. It
will rename them (or copy them with a new name) denoting how sharp each frame is.
8.1.6.2 Image Preview
This will let you easily go through your images one by one to examine them, (optionally)
rename them, and/or (optionally) delete them. File, Preview Files. If you've not tried
this, try it. It's quick, easy, and immensely useful.
8.1.7 Stacking: Align and Combine
It's now time to Align and Combine (stack) your light frames. Here, there are a large
number of options as to how to proceed. We'll start with the basic version first and
then detail the other paths you can take.
1. Pull down Batch, Align and Combine Images
2. If you're not on an alt-az mount, hit OK, keeping the defaults of saving the stack,
using Translation, and Average / Default stacking. If you're on an alt-az mount,
you'll need to include rotation, so change the Alignment Method to Translation +
Rotation.
3. Select your light frames
4. Find a star in your image that's not ultra faint and not big and bloated. Move
your mouse over it to make sure that the core of the star isn't all 65535 (the max
possible value). Click on that star and Nebulosity will advance to the next image.
If your mount's tracking is at all decent, the same star on the next frame should
be circled. If the circle is on the right star (don't worry about centering), just hit
33
Ctrl-click (or Command-Click on the Mac) to tell Nebuolsity "yes, that's the right
star and I want to use this frame". If it missed the star, just click on it (don't
worry about being precise). If the frame is a bad one and you'd like to skip it and
not include it, hit Shift-click.
5. If you're doing Translation + Rotation (or Drizzle), you'll need to find a second
star and run through each frame again. Try to pick one that's not very close to
the first star.
6. When you're done (the Status Bar will show you your progress), Nebuolsity will
align and combine all the images and pop up a dialog asking you for a filename
to save the resulting stack in.
There you have it! Basic stacking. There are some more advanced options you can try:
1. Translation + Rotation (+ Scale): The normal Translation alignment will only shift
images by whole pixels and does not account for any rotation across frames.
Running these will shift the images by fractional pixels (interpolating them as
needed), rotate them as needed and, if selected, scale them as needed to coregister the images.
2. Drizzle: Drizzle is a powerful technique that will align, combine, and increase the
resolution of your images during stacking. It is suitable for alt-az mounts as
rotation is included in the alignment. You will therefore need to select two stars
during alignment. Make sure you have Normalized your images at some point
first.
3. Colors in Motion: This tool is only available for images from one-shot color
cameras that have not been converted into color yet. It will align the images and
convert them into color at the same time. It is a translation-only based
alignment.
4. Standard Deviation (SD) stacking: Instead of taking the average value for each
pixel (across images), take the average but toss out "outliers" or values that are
atypical. Thus, if a hot pixel "crosses over" a pixel in the aligned image (the hot
pixel didn't move but the frame did when the stars were aligned), this bright hot
pixel will be an atypical sample and will be tossed out before averaging. To use
this technique, you must first do your alignment, saving each frame first and
then pass these aligned frames ("align_OriginalName.fit") into Align and
Combine again, selecting "None (fixed)" as the alignemnt method (and one of
the Std. Dev. thresholds in the Stacking Function). Make sure you have
Normalized your images at some point.
8.1.8 Crop off the edges
If you don’t select a region
After stacking, odds are you've got a dark border around to crop and pull down
your image as Nebulosity tried to make an output image
Image, Crop a dialog will
big enough to hold everything from every frame (an
appear asking you exactly
exception here is in rotation where you will have bits cut
how many pixels you want
off at times). Odds are you don't want this bit and it'll
to remove from each side.
just make the histograms look funky when you're
stretching. Use the mouse to define a rectangle that has
the good part of the image and pull down Image, Crop. Save this with a new name.
34
8.1.9 Remove the Skyglow Color
If you're shooting in color (one shot or having combined frames), odds are the
background sky is not a nice neutral gray, but rather something rather unpleasant
(green, pink, and orange are common). This comes from the color of your skyglow.
Fortunately, it's easy to remove. Simply pull down Image, Adjust Color Offset (unless
you've got a reason, accept the default values) or pull down Image, Auto Color
Balance. Save this with a new name.
8.1.10 Stretching
Now, the fun begins as it's time to see what you really have in that shot. Sitting atop
that skyglow should be the faint galaxy or nebula you were shooting and stretching is
how we bring this out. There are three main tools for stretching in Nebulosity. The first
is the Levels / Power Stretch, the second is Digital Development Processing (DDP), and
the third is Curves. For each of these, more detail is provide in the section on Image
Adjustment.
The goal in each of these is to pull your image's intensity profile (histogram) and stretch
it so that very low contrast differences are made more apparent. Thus, you are pulling
your faint galaxy arms away from the skyglow and doing things like sending the
skyglow down to a nice dark background. When doing this:
• Keep your eye on the histogram. The histogram is your friend.
• Until the very last steps of stretching, don't let the left edge of the histogram get
cut off and don't bang too much (e.g. the core of your galaxy) into the right edge
of the histogram. Once they hit the edges (0 and 65535), you'll never resolve
details in there again.
• Turn off auto-scaling (or let Nebulosity do this for you) so that what you're seeing
on the screen is the full 16-bit data in all its glory. This will help you use the full
range of intensities your image can take. Remember, the B and W sliders are just
there to make the image prettier on the screen (they do a stretch for display but
don't really affect the underlying image). So, have them at full left and full right
and then start to stretch. (If you're in auto-scale when you enter Levels, it will
turn it off and set these at the extremes for you).
• Don't try to do everything in one pass. Make several passes over the image to
slowly pull it into the condition you want it.
• Save often
8.1.10.1 Levels / Power Stretch
The Levels tool in Nebuolsity does the same math to your image as tools like
PhotoShop's Levels tool. You're setting a black point (top sider), a white point (middle
slider) and a midpoint or "power" (bottom slider). This is much like a “Curves” tool but
with a fixed basic shape to the curve (see Levels / Power Stretch). With several passes
over the data you can build up a complex “Curves” transformation to your data. In
general, for the first few passes, have the "power" slider be less than one (try values
like 0.6) as this will help accentuate the low-contrast details and pull them out. Start
getting the details to pull apart from the background before you work too hard on
35
pushing the background to being dark. You can always darken the background later.
8.1.10.2 Digital Development Processing
DDP is a tool that applies a special transformation designed to make linear CCDs
behave more like film. If you use DDP, do it first or without using the Levels tool much
beforehand as the math behind it expects you to have not altered the linear response
of your CCD's image. I find that DDP works best if the skyglow is not too bright to
begin with. Feel free to use the Levels tool and adjust the black-point (first) slider to
bring the histogram nearer to the left edge before running DDP. Just don't start
adjusting the Power (aka midpoint, aka 3rd slider) in the Levels tool before using DDP.
8.1.10.3 Curves
The Curves tool gives you a lot more flexibility than you would get with either DDP or
Levels and can be used both for initial stretching and for fine-tuning the results of
something like DDP. Feel free to use all of the tools and to mix the tools as you see fit.
To use the Curves tool, simply grab one of the two blue dots and move it around.
These two “control points” help you draw out the curve which is always shown in the
dialog. By having these two points and the two endpoints, you can draw a wide range
of very useful curves without getting into trouble by making a very odd transformation
of your image. As with Levels, you can do a lot more in several iterations than you can
in just one. As you re-use the tool, you’re effectively building up a more and more
complex curve.
36
9 Image Pre-­‐Processing: The Details
1.
2.
3.
4.
Pre-processing: Theory
Pre-processing one or more sets using the Multiple Set tool
Automatic dark frame scaling
Bad Pixel Mapping
9.1 Pre-­‐processing: Theory
Our CCD images have a number of artifacts in them. Typical artifacts include hot pixels
(bright dots that appear in the same place on the image and that get worse with
increased exposure duration or temperature), vignetting (un-even illumination of the
field), dust spots (large blobs or donuts that appear superimposed on the image, looking
a bit like a watermark), and noise (a static-like or grainy appearance to the image).
We can fix these by taking and applying a series of control frames that have the various
defects but not our image and using them to remove the defects. When thinking about
this, it is useful to consider the following:
LightImage = TargetImage × NonFlatness + DarkCurrent + BiasCurrent
DarkImage = DarkCurrent + BiasCurrent
BiasImage = BiasCurrent
FlatImage = NonFlatness + BiasCurrent + DarkCurrent_flat
Our goal here then is to apply the following formula to
our image:
RawImage − DarkImage − BiasImage
NewImage =
FlatImage − DarkImage_flat − BiasImage
Note, the dark current in a
flat image
(DarkCurrent_flat) is
typically far, far less than in
your light frame or dark
frames since the exposure
durations are typically a lot
less for your flats and can
typically be ignored.
Both Dark Frames and Bias Frames are taken with no
light hitting the camera. Dark frames are to be taken
under the same circumstances as your Light Frames (e.g., the RawImage pictures of
your DSO). Use the same duration of exposure and try to have the CCD at the same
temperature (e.g., if you use the TEC in your DSO shots, use it in the Dark frames.
Often, these are taken in the same imaging session or a collection of "master" dark
frames for various imaging situations is compiled. Always take a number of dark
frames (somewhere between 10 and the number of exposures used in your light
frames) and combine them (average or median) to create a suitable dark frame to be
used during pre-processing.
Bias frames can be taken at any time simply by covering the telescope or putting a
lenscap on the camera and taking a series of short exposures (e.g., 10 ms). Take a
good number of these some day when you're bored and combine them (average or
median) to create a master bias frame.
37
In contrast to Bias and Dark frames, Flat Frames are taken with light hitting the
camera, but with the light coming from an even field of illumination (e.g., aiming your
telescope at a white wall, defocused at the sky at dusk bouncing the scope around,
putting a diffuser over your telescope, etc). The exposure duration of Flat frames does
not matter per se, but should be long enough to ensure no pixels are at or near zero
and no pixels are near saturation (Nebulosity will automatically scale the intensity of the
image to have a mean of 1.0, so don't worry how bright it is overall). Again, take
several of these and combine them.
Nebulosity's Pre-process routine will subtract any Dark frame provided from each
image, subtract any Bias frame provided from each image, and divide the result by the
flat Frame. You may notice that this is leaving off part of the equation, as the
denominator does not include the part about subtracting the Dark frame or Bias frame
from the flat frame. This is because the Flat frame is typically taken at a different
duration (usually much shorter) than the Light frames, meaning a different Dark frame is
needed to remove the hot pixels from the Flat frame. What this means is that for best
results, you should pre-process your Flat frame by treating it like a Light frame and
applying a suitable Dark frame or Bias frame. The easiest way to do this is to tell
Nebulosity to apply the bias frame to your flats as well (or to smooth the flats).
9.2 Image Set Pre-­‐processing Tool
Some users will take a single set of lights, darks, and flats for a target. Others,
monochrome imagers in particular, will end up with multiple sets of lights, darks, flats,
etc. The Batch, Pre-process Image Sets tool lets either kind of user easily pre-process
images.
When you first fire it
up, it may not seem
as if it’s going to be
easy as the dialog
may seem a bit
daunting. What’s
shown on the right
here is a sample we’ll
walk through. It’s
really actually quite
straightforward,
though.
There are a few things
to keep in mind:
Define your control
frames (darks, biases, and flats) first. Here, you’re saying “these are the frames I’m
going to want to use to fix some lights.”
38
You can select individual images or multiple images. Nebulosity will stack the frames
on the fly if you select multiple. For the flats and lights, you can choose to apply the
control frames - the bias, dark, (and, for the light frames, flats). When you’re doing this,
you’re telling Nebulosity, which control set to apply.
Let’s walk through this example. First, I pressed the Dark 1 button and when the
dialog appeared, I selected an existing master dark frame (in this case it is called
“Master_Dark_30_fr_5m-10C.fit”). This is now Dark 1. I then pressed the Dark 2
button and grabbed three frames (these are actually from another camera with a
different sensor size even). Remember, you can select multiple images in those dialogs
by shift-clicking, control-clicking (or on the Mac, Command-clicking) the way you can
in other applications. Here, we’re telling Nebulosity to stack those three dark frames
and call that Dark 2. Note how next to the name of one of these dark frames
(“Dark_Stability_1m-1_052.fit”) you see the number three in parentheses. This is telling
you that there are three frames you selected here that will be stacked on the fly.
I did a similar thing to select an individual bias (Bias 1) and to stack several raw bias
frames (Bias 2). Again, these are actually from two different cameras (though they
needn’t be, of course).
I selected several sets of flats here as well. For Flat 1, I’ve got pre-stacked flats from
the first camera when the H-alpha filter was on it. For Flat 3, it’s the same camera with
the O-III filter on it. For Flat 2, it’s the other camera and I selected several flats here to
stack on the fly. For the flats, though, we want to do some pre-processing to clean
them up. First, the flats don’t have much dark current but they do have bias current.
So, I told Nebulosity to apply the appropriate bias frames to these flats. The first
camera’s bias frame was “Master_Bias.fit” and this was defined as Bias 1. For both
Flat 1 and Flat 3, we’ll apply this bias frame. The second camera’s bias frame was Bias
2 (the stack-on-the-fly bias frame of three individual images, including one called
“Bias_003.fit”.) So, Bias 2 is selected for Flat 2 and Bias 1 is selected for Flat 1 and
Flat 3. In addition, there will still be some noise here that I may want to reduce before
applying the flats to my lights. So, 7 pixel blur is selected here by default (the default
value is set in the Preferences dialog). You’ve got several options here to choose from.
If you’re using a one-shot color camera, though, you’ll want to do something here to at
least remove the Bayer matrix (any of them should remove this).
Now, we move onto the lights. Here, I selected some H-apha and O-III data from the
first camera and put them into Light 1 and Light 3. It’s important to note here that I
could have put them in any set of lights. I could have done Light 1 and Light 2 or even
started at the back and done Light 4 and Light 5. Which set you put them in doesn’t
matter. You don’t need to match up those numbers with your flats, biases, or
darks. Where you do the matching is in those pull-downs. So, for these two sets
(“(4) Ha_10m_021.fit” and “(5) O3_take2_E-1_005.fit”), we’re going to apply the same
dark frame - Dark 1. We’ll apply different flats, though, as there were different dust
bunnies on my two filters. So, the flat with the H-alpha filter (Flat 1) gets applied to the
light with the H-alpha filter (Light 1). Same deal for the flat with the O-III filter (Flat 3)
39
and the light with it (Light 3). For the lights from the second camera, Light 2, we’re
going to apply its dark (Dark 2), and it’s flat (Flat 2). Of course, since the bias current is
contained in the dark frame, we’re not going to apply any bias frames here (that would
double-subtract out the bias and inject the bias back in).
Once you’re all set, press OK and go away for awhile. Note that if you have the History
window open (View menu), you’ll get a full blow-by-blow of exactly what is going on.
9.3 Bad Pixel Mapping
Removal of hot pixels using the typical (or
automatically-scaled) dark frame subtraction
technique does have one real drawback.
Your dark frame contains not only
information about the hot pixels, but it
contains other kinds of noise as well. Stretch
the dark frame and you'll see the same kind
of "readout noise" you see in a bias frame
along with other noise components. If you
average many of these frames (some
suggest twice as many dark frames as light
frames), much of this noise will disappear,
but it may take a lot of dark frames to have it
go away. Therefore, while dark frame
subtraction will remove hot pixels, it can
actually add noise into your image!
An alternative method for removing hot
pixels is called Bad Pixel Mapping. In this
technique, you first identify those pixels on
your CCD that are prone to problems ñ your
hot pixels. Once you've identified those
pixels, only those pixels in your light frames
are touched. We don't really know what the
value in that pixel should be, but the
software can make an educated guess and
fill in for the bad value. This can work very
well as the images below show. First is a
zoomed-in frame from a one-shot camera that was de-mosaiced without any
correction. Next is the same area when the bad pixel map was applied prior to the demosaic process.
Using Bad Pixel Maps is quite easy but does require that you are either using a black
and white camera or, if using a one-shot color camera, that you capture in RAW mode
(Bad Pixel Mapping must be performed prior to converting the image into full-color).
The first step is to generate the map. For this, you need a dark frame or an average of
40
several dark frames. Ideally, this will be a combination of several frames taken at the
longest exposure duration you expect to use. Pull down Make bad pixel map from the
Batch menu and load the dark frame when prompted. A slider will now appear and the
display will show you your hot pixels. Nebulosity attempts to come up with a
reasonable position of the slider for you. Here, we have the default setting and display
for creating the Bad Pixel Map using a sample dark frame.
At this point, you can
adjust the slider and
you'll see the number of
bad pixels identified
change (here, showing
821 bad pixels). What
you’re seeing is actually
the dark frame after some brightening and stretching. Currently flagged bad pixels are
the black dots that are on the mostly gray background. The white dots shown here are
hot pixels that haven’t been marked as bad in this example. Move the threshold slider
to the left and more pixels get marked as bad and more black dots thus appear. Move
it to the right and fewer black dots appear (more white dots should appear). What you
are doing is moving a threshold - saying that anything above this intensity is a bad
pixel (shows up as black) and anything below it is a good pixel. When you like your
map, click on Done and you will be prompted for a name to give this Bad Pixel Map.
Give it a meaningful name, as you may well want to create several maps. If you used a
5-minute dark frame, you could use that dark frame to make several maps - one for
~5-minute exposures, one for ~1-minute exposures, and one for ~20-second
exposures for example by using different values of the threshold (letting fewer hot
pixels show for the shorter exposure maps). There is no "exact right value" here. You're
simply telling Nebulosity which pixels not to trust.
Once you have your map, you can now process your light frames. In the Pre-process
image tool change the entry that reads “Dark subtract” by default to the BPM option
corresponding to the kind of images you have (images from a one-shot color camera in
RAW format prior to the de-mosaic process OR images from a black and white
camera). It'll prompt you for the map to apply and then for the set of light frames you
want to process (shift-click or ctrl-click to select multiple frames).
When done, you can Batch De-mosaic the images if they were from a one-shot color
camera and then go on to Alignment and Stacking.
41
10 Reconstruc6on: Demosaic-­‐ing and Pixel Squaring
If you capture in RAW format from a one-shot color imager, you will at some point need
to "Debayer" or "Demosaic" your images. (If you capture doing color on the fly, this
happens immediately after image capture automatically). This converts the RAW image
from the camera into a full-color frame (see One-shot color: RAW vs. RGB). This can be
done either before or after pre-processing (although results will be best if you do it after
pre-processing - see Pre-Processing. In addition, whether you use a color or
monochrome imager, you will often need to convert the camera's native pixel
dimensions into square pixels prior to stacking your images as many cameras have
pixels that are natively not square.
Nebulosity provides tools to do this on both an individual image (on the Image menu)
and in a batch mode for a series of images (on the Batch menu). If Nebulosity is able to
determine the needed information about the image and where it came from, this will
happen automatically. If not (or if you have “Manually override color reconstruction”
selected in the Preferences), a dialog will appear and prompt you to enter in several
parameters about your camera so that the processing can go on accurately.
In this dialog, you must first tell Nebulosity if the sensor uses RGB color filters or
CMYG. Virtually all of the cameras in use by amateur astrophotographers are RGB
cameras. Next, consider the offsets. What this is doing is telling Nebulosity what color
the first pixel is (by saying how much of an offset - how many steps right and/or down there is in your image’s color coded array) versus what Nebulosity expects.
Unfortunately, there is really no way of knowing ahead of time what is correct here. For
RGB arrays, your choices are 0-1 for X and Y and for CMYG they’re 0-1 for X and 0-3
for Y. Try various settings until you get something that looks even (not striped) and is
close.
If your array uses pixels that aren’t the same size, enter in their size in the Pixel size
section (if they’re square, don’t worry about it
and just keep the default values of 1).
If the resulting image is close, but not quite
right, you can either adjust the color with the
various tools in the Image menu, or you can
use the Color control portion of the dialog.
Here, the original (input) color value is given in
the columns and the output color value is
given in the rows. By default, the diagonal has
1.0 and the other values are 0, saying all of the
red goes to red, green to green, etc. If you put
a 0.5 in so that the first row had values of 1.0, 0.5, and 0.0 this would mean that 50%
of the current green value for a pixel would be added into the red.
Typically, this level of control can be skipped, but if you’re working with an odd filter
42
setup or you’re working with CMYG arrays, this can be quite helpful. For example, if
you’re using a Sony CMYG chip and Nebulosity does not recognize the sensor, once
the offsets are in place, values of 1.06, 0.29, -0.41 in the first row, -0.4, 1.06, and 0.54
in the second row, and 0.50, -0.4, and 1.11 in the last row will give a reasonable color
rendition by compensating for the chip’s imperfect color filters.
It is important to note that Nebulosity will automatically square the pixels during the
debayer process for one-shot color images. Any color image is assumed to therefore
have square pixels.
It is also important to note that if you use a Canon DSLR, ideal color balance in
Nebulosity will be accomplished if you select the appropriate setting under "DSLR
White Balance / IR Filter" in the Preferences dialog.
43
11 Previewing and Grading Images
Let's face it. Some images just don't look good. Some of your shots may look great
and some may look horrible. Maybe the tracking failed, maybe dew started to form,
maybe you hit the scope or the wind blew. Who knows what, but some shots just
aren't as good as others.
When stacking images you have the chance to skip any frame you don't like but this
can be a bit too subjective at times. Nebulosity provides two ways to do this
beforehand.
The first is to use the Preview Files command in the File menu. You can use this to load
up a set of images and rename or delete ones you do not wish to use (you can also use
this to blink between images). The second is a way to automatically grade each image
in your set relative to each other image in the set to let you pick the best frames. In
the Batch menu, you'll find an entry for Grade Image Quality. Select it and you'll be
asked to choose the frames you want to grade.
Earlier versions of Nebulosity used an image-grading algorithm that graded the relative
quality of images. That is, which images were the sharpest and which were the least
sharp based on an assessment of what the edges looked like in the image. This has
now been replaced with an algorithm that grades the absolute quality of the images.
For each image that is graded, a set of stars is identified and the same half flux radius
(HFR) used during Fine Focus is calculated for these stars. The average HFR (times
100) is used as the filename post-grading. So, if you see an image called Q382_M51.fit
after grading, the average half flux radius was 3.82 in that image.
44
12 LRGB Color Synthesis
Nebulosity allows you to synthesize a color image from separate frames. For example,
users of monochrome cameras must take a set of images through red, green, and blue
filters to create a color image. Users might also want to combine a full-color image
from a one-shot color camera with a "luminance" frame taken from a monochrome
camera, or to combine images taken through separate line filters. Nebulosity provides a
tool for this.
To use the tool, you must first align all your images and save each frame (rather than
the whole stack) using the Align and Combine tool. Co-register the images using any
version of the Translation (+Rotation +(Scale)) tool you wish. You can do this on all the
raw frames at once or you can do this on the stacks you’ve made for each color type
(i.e., your processed red data, your processed green data, etc.)
Located in the Batch menu, the LRGB Color Synthesis tool can then be activated.
Select your mode (RGB, Traditional LRGB, and Color Ratio LRGB) and frames. If you
have a full-color file, you can load all 3 color planes at once using the "RGB frame"
button. This will overwrite any red, green, or blue data you have already loaded. (You
can load these first and then replace any color plane by loading another file into that
color plane directly.) If you know ahead of time you wish to scale the color channels
relative to each other, you can do so using the sliders.
12.1 RGB Mode
In RGB mode, 3 color channels are used and directly create a color image. It is the
simplest mode.
12.2 LRGB: TradiAonal HSI Mode
This implements the traditional LRGB technique. The red, green, and blue frames are
used to calculate a hue and saturation value at each pixel. The luminance (or intensity)
value is replaced by the value provided by the luminance frame. The resulting HSI (or
HSL) data are converted back into RGB in the output image. This is the traditional
method, but can lead to a loss of saturation.
12.3 LRGB: Color RaAo Mode
This technique gives an alternative method of luminance layering that avoids the loss
of color saturation by the traditional method. The RGB data are used to create R:L,
G:L, and B:L ratios (L derived from RGB). The L component is then replaced by the
value in the luminance frame and the image converted back to RGB.
45
13 Image Normaliza6on and Histogram Matching
Ideally, all frames taken under the same circumstances of the same target should all
have the same intensity. Often, this is not the case as changes in light level, cloud
cover, etc. can change the intensity from frame to frame. Further, if changes are made
in the capture settings (e.g., different exposure durations), you're certainly going to
have differences in overall image intensity across frames.
If you're doing the Average/Default method of stacking, you need not worry about this
issue unless the changes are really quite severe. If you're using standard-deviation
based stacking, Drizzle, or Colors in Motion, it is a good idea to normalize your images
before stacking. What this will do is to get all of the frames to have roughly the same
brightness by removing differences in the background brightness and scaling across
frames.
There are two methods to normalize the images in Nebulosity.
1) The Normalize entry in the Batch menu will go through all selected frames and
attempt to put them all in a common intensity range by taking care of offset and scaling
differences across frames. After normalization, all frames should have their minimum at
~100 and their maximum at ~65535. Do this before you do any alignment of the frames
(it can be done before or after pre-processing, but you don't want the black borders
surrounding the image that can come in during alignment to throw off the normalization
process.)
2) The Match Histogram entry in the Batch menu will ask you for a reference image and
attempt to match the histograms for a set of target images to this reference image. It’s
performing a more advanced stretch than you’d get in Normalize and can thereby be
more effective. Note, if you’re using a one-shot color camera, do your demosaic
before you try to use Match Histograms.
46
14 Stacking Images
Stacking multiple exposures is a fantastic thing to do for your images. If you can stack
your images, you don't need to hold perfect tracking as long (making life easier) and
you reduce any noise in your image that is not consistent from exposure to exposure
(much of the noise is not). Thus, a stack of images will look less grainy (less noisy) than
any one individual image. This lets you stretch and process the image more to bring
out fainter details. All in all, stacking is a very good thing.
In Nebulosity, stacking can be done with or without alignment. For light frames (where
stars are apt to move between each frame), you will want to align the images either
prior to stacking or during stacking (see below). For things like dark frames, bias
frames, and flat frames, you will not want to align the frames first.
14.1 Overview of TradiAonal Star-­‐based Alignment
Few of us have perfect mounts that track or guide so well that there is no drift
whatsoever across images. (In fact, it turns out to be better to have a bit of drift
between images, as your image isn't always aligned with whatever consistent noise is
in your camera, but this is a rather long topic not worth going into at the moment.) The
net result of this is that if we were to stack a series of images atop each other as they
came off the camera, we would end up with a blurred or streaky looking image. Each
star was not in the same place in each image (the whole field of stars moved between
images), so the result is quite poor.
In Nebulosity, this process is called "Align and Combine" and there are various options
available to you (see below). In each, you are asked to find a star (or sometimes two)
that is the same in each image. You will be asked to left-click on this common star (it is
best to use a somewhat isolated and non-saturated star) in each image. Don't worry
about being perfect in your clicking. Nebulosity will always search around the area
where you clicked to find the star's centroid (i.e., it will refine your click automatically).
In addition after the first image, Nebulosity will attempt to find the same star in each
image for you and place a circle around that star.
If you wish to keep Nebulosity's location, simply Ctrl-click (Command-click on the Mac)
anywhere in the image (if it gets it wrong, just click on the correct star). If you know you
want to keep all of the images and your tracking was good enough that Nebulosity can
find the star from frame to frame, an Alt-click will pretend you did a Ctrl-click on each
frame and accept all frames. Finally, if you want to skip an image (e.g., if it was blurred),
simply Shift-Click anywhere in the image and it will not be used in the stacking
process.
Nebulosity provides several ways to align a series of images prior to or during stacking
to take out this overall movement in the image. In the simplest method (Translation),
Nebulosity will take out shifts between your images (a.k.a. translations) and average the
aligned data. You do this by picking a common star in each image, and Nebulosity
takes care of the rest, shifting each frame (by whole pixels, without "resampling"). This
47
works very well for equatorially-mounted telescopes (including fork-mounted scopes
on a wedge). This does not work for Alt-Az mounted scopes. This style of mount
makes stars not only move left/right and up/down but the entire field rotates as well.
A more complex technique, Colors in Motion, is also used for stacks that have shifts
between images. Unlike the other techniques provided, Colors in Motion
simultaneously aligns RAW images, stacks them, and reconstructs color information
from one-shot color cameras. It cannot be used on RGB data or on black and white
data.
To align and combine images using alt-az mounted scopes (or equatorially- mounted
scopes), Nebulosity provides three other techniques. The first is similar to the above
but allows for rotation and sub-pixel alignment. It is called Translation + Rotation.
Related to this is Translation + Rotation + Scaling in which frames are allowed to be
resized to align atop each other. To let Nebulosity know about the possible rotation,
you must pick two stars in each image. Each image will be shifted and rotated to align
them all prior to averaging the data.
The final technique that works on both equatorial and alt-az mounted scopes is called
Drizzle (Align and Combine: Drizzle). Drizzle can not only combine images from alt-az
(or equatorial) scopes, but it also enhances the resolution very well. To do this, you will
again have to pick two stars in each image, and Nebulosity will do the rest.
14.2 CombinaAon methods
In all of these, you have the option of using either strict averaging or adaptive stacking.
In several (the Translation (+Rotation) (+Scaling)) you also have the option of saving
each individual file post-alignment rather than saving the stack. This can be very
useful, for example, in preparing frames taken through different filters for (L)RGB color
synthesis or for using the Standard Deviation method of stacking.
14.2.1 Averaging vs. Adding vs. Adap5ve Stacking
During any of the Align and Combine methods, Nebulosity can mathematically stack
images in one of two ways. By default, an Adaptive Stacking technique is used (see
Preferences menu). Some people worry a great deal about whether to add (sum) or
average their frames during the stacking process. Each technique has its ups and
downs. If you have 3 images in which the same pixel reads 100, 100, and 101,
summing gives you 301, whereas averaging gives you 100. Internally, in Nebulosity, the
average would be 100.33333 (as it should be), but when saved, it would become 100
as the images are saved in "integers" (aka whole numbers, not "floats" which let you
have fractional bits as well). This makes one think that adding is best, but another
example shows the problem there.
Let's now say that a bright pixel reads 32000, 32010, and 32100 in our three images.
The sum is 96110 here where the average is 32036.666. When saved, this would
become 65535 if summing were used and 32037 if averaging were used. Let's have
48
another pixel - even brighter - reading 64000, 64010, and 64100 in the image. Once
saved, the sum would make this 65535 and the average would make it 64037. Here,
we see the problem with simple summing. You can saturate the image pretty easily,
especially if you start with 16-bit images. Here, one pixel should be twice as bright as
the other and yet it ends up equally bright (65535) if adding is used, since this is the
highest possible value.
Nebulosity uses an Adaptive Stacking technique that avoids the weaknesses of both. It
can be viewed as always being somewhere in between adding and averaging your
data. The output (the stack) will always have a maximum value of ~65535 so that you
are always using the full range of your data. This is enabled by default and for most
uses will be optimal. (Note, it is not used when the Fixed Combine is selected as this
tool is often used for dark frames). Unless you have a real reason to, you should leave
this on (see Preferences menu). If you turn it off, Nebulosity will compute a straight
average when stacking.
14.2.2 Standard Devia5on Based Stacking (Sigma-­‐clip)
If you were to take a perfect image of a target, each pixel would have its "ideal" or
"true" value - how much intensity there is from that part of the target. The trouble is,
each time we sample the target (take an image of the target), we record both that true
value and some noise.
Averaging helps to get rid of this noise. It should tell us the central tendency and
therefore estimate the truth. The more samples we have, the better the estimate is. But,
if some samples are really abnormal ("outliers"), the average can get thrown off. For
example if a hot pixel drifts into a patch of background sky, the average value will go
up considerably and the hot pixel will show up in the stack (technically, of course, the
stars and sky moved, not the hot pixel).
Standard Deviation (SD) based stacking gives us a way
to identify these outliers and eliminate them prior to
computing the average. To do this, we calculate not only
the mean (average) of each pixel's value in the dataset
but also a second statistic that describes how much
variability there is around that mean. We calculate the
standard deviation (the square root of the variance) and
use this to filter out "bad" samples.
Note: Another term used
for this method is “sigma
clipping” or “sigma clipped
stacking”. Sigma is the
Greek symbol typically used
to refer to the standard
deviation.
If we assume the data are "normal", about 70% of all samples will lie within one
standard deviation of the mean (that is, 70% are less than one standard deviation
above or one standard deviation below the average). About 95% like within 2 SD of the
mean.
During SD stacking each (aligned) pixel, has the mean and standard deviation
calculated across all of the images in the stack. If your SD threshold is at 1.5, any
samples of that pixel that have an intensity beyond 1.5 SD from the mean are removed
49
and a new average, excluding these samples, is calculated. This is why hot pixels are
often eliminated using SD stacking - those hot pixel values are very abnormal and lie
far away from the mean.
With the filter set at 1.75, it takes a more extreme or "outlying" intensity value to be
counted as "bad" than at 1.5. At 2.0, it takes even more abnormal a value to be
excluded. Thus, more samples go into the final image using a higher threshold (and
more noise as well). Typically, filtering values at 1.5 or 1.75 will yield the best results.
To use Standard Deviation stacking, you must first Normalize your frames. Then, you
must align your images using any of the Translation (+Rotation (+Scaling)) routines. Do
this to all of your images and save each file rather than saving the average stack. Then,
select "None" for alignment method and the various SD stacking choices will appear.
Try 1.75 or 1.5 initially before using any more extreme values.
Note, that this requires a lot more memory than the more traditional techniques. If
you’ve not got much memory and you’ve got very large images, expect this to take
awhile.
14.3 Align and Combine OpAons
When aligning images to get them ready to stack, you’ve got a few options. Your
choice comes down to asking yourself things like whether you need to account for
rotation (e.g., your polar alignment isn’t great or the camera rotated between images).
Let’s go through the various options here.
14.3.1 Align and Combine: Fixed
Positive nothing moved or want to see just how much it did move? Fixed may be your
choice here. Running this will stack things without any alignment at all. At the very
least, it’s useful (mandatory) for stacking any control frames like darks, flats, and
biases.
14.3.2 Align and Combine: Transla5on
This is the most straightforward method of stacking. Each image is shifted up, down,
left, and/or right by whole pixels prior to align it with all the other images prior to
combining.
Once your images have been pre-processed, select Align and Combine: Translation
from the Batch menu. You will then be presented with a dialog asking you what frames
you wish to align. Select all of your pre-processed frames, even if you think a few you
may not want to keep and press OK.
At this point, you'll notice the cursor has changed to a cross-hair (see below). You may
also notice all of the menu items have gone grey, as you are now in alignment mode.
The status bar shows your progress, telling you how many frames you've selected to
align and how many you've done so far. Feel free to adjust the display levels or zoom
50
factor here to get a good view.
Your goal now is to identify the same star in each image. Pick a star that is fairly
isolated and that is not saturated. Big, bloated stars need not apply nor should ultrafaint ones near the level of the background. Opening up the Pixel Stats window can
help in selecting a suitable star. Put the cursor over this star (and remember which one
it is) and click the left mouse button. In so doing, you're saying "The star is here" to
Nebulosity.
If you want to abort the
In truth, you're actually saying, "The star is about here."
whole process, simply press
None of us can click perfectly all the time and doing so
the Abort button.
would be a very time consuming process as we would
obsess over whether it the click should be here or one
pixel over. So, Nebulosity never assumes you got it 100% on target. Instead, it looks in
a small area (+/- 5 pixels) to see if there's a better candidate for the center of that star.
That is, it refines your click. So, get close, but don't obsess over being perfect.
Nebulosity assumes you're not perfect and will try to fix it anyway. After the first image,
you will notice a circle appear around a star - hopefully, the same star you've been
using in prior images. To keep the current location (i.e. to say "yea, you got the right
star, there Nebulosity),
Ctrl-Click (CommandClick on the Mac).
If there is an image you
don't want to include in
the stack (e.g., a plane
flew through your DSO,
the mount mistracked,
you moved the scope to
re-center the target
during imaging, the
wind blew, a cop shined
a spotlight at your
scope - all of which
have happened to me),
just Shift-Left-Click
anywhere in the image.
That frame will be
ignored.
Once you've selected the same star in each image (at least for each image you plan to
use), Nebulosity then goes about aligning the images and combining them into one
composite image. Depending on how many images you're aligning and how big they
are, this could take some time. Nebulosity shows you its progress in the Status Bar.
After all images have been combined, Nebulosity prompts you for a name to save the
composite image as. Give it a name and press OK and you're done. The image now
51
displayed on the screen is this composite image.
14.3.3 Align and Combine: Transla5on + Rota5on (+ Scaling)
Using this method, images are shifted (translation), rotated, and optionally scaled by
sub-pixel amounts in order to optimally align images prior to combining them. The
process is very similar to that used in Align and Combine: Translation. The only
difference in what you do is to pick two stars in each image.
Once you have gone through and picked the first star (which Nebulosity uses to gauge
the translation), you will go back through all images and be asked to pick a second star
(used to gauge rotation). As before, feel free to keep the current best guess of the
star's location (Ctrl-Left-Click or Command-Left-Click on the Mac), to skip any image
you don't like by Shift-Left-Clicking or to abort by hitting the Abort button.
14.3.4 Drizzle alignment (Transla5on & Rota5on + Resolu5on enhancement)
Drizzle is a technique developed by Andy Fruchter (Space Telescope Science Institute)
and Richard Hook (Space Telescope European Coordinating Facility) to make the most
of images from the Hubble Deep Field's WFPC2 camera. Their insight stemmed from
the idea that images from the Hubble were undersampled. The optics were able to
resolve a lot more detail than the size of the CCD pixels would allow. So, a star that hit
the dead center of a pixel would have some spread over adjacent pixels but would still
look blocky. Small motions of the star on the CCD could keep it on the same pixels, but
the relative brightness of the pixels would alter slightly based on exactly where the star
was centered.
That effect can be seen here. On the top is a well-sampled shot
taken from the Digitized Sky Survey and below are several
undersampled versions in which the stars moved a bit prior to
being sampled (on a simulated camera). Drizzle uses this
information to not only align images, but also to create a higher
resolution stack than you find in any individual image.
Here is a diagram from Andy Fruchter's page on Drizzle showing
the technique. Each pixel in the original image is first reduced (from
the red pixel size to the blue pixel size - this is the "pixfrac" or
"pixel reduction factor" term). The pixels are then translated and
rotated so that they will all be in the same position in the output
image. You'll notice two things about the output image. First, the
pixels in the output image are smaller than the original pixels. This
is the "up-sample factor" - how much higher the output resolution
is than the input resolution.
52
Second, these pixels will
usually not line up perfectly
with a pixel in the output
image. Key to Drizzle is the
fact that these "drops" (the
blue pixels) fall onto the
output pixels to the degree
that they overlap. The value
dropped by a blue pixel will
fall a lot on one output pixel,
a little on another, less still on
a third, etc. Think of the
output grid as having little
wells (or spots on an icecube tray) for each output
pixel and the water drop
landing on a spot that hits multiple wells. As multiple images rain pixels down onto the
output image, the output pixels fill up to the degree that input pixels line up with the
output pixels.
That's the theory behind Drizzle. How does it work in Nebulosity? Images must be
black and white or full-color (but not RAW) and pre-processed. There should also be
some movement between images (e.g., if Fixed combine of images would make a
blurry average). You'll also want to have a minimum of 8-10 images to work with as well
and you should most likely have Normalized your images prior to stacking.
Drizzle requires two stars to be found in each image. First, find one star (left-clicking as
in Align and Combine: Translation) in all the images (shift-left-click to skip an image,
ctrl/command-left-click to keep the current guess). This first one will serve to let
Nebulosity know how much translation is in each image. After you have selected the
same star in each image, Nebulosity will loop back and ask you to find a second star in
each image. A red target will appear over the first star to let you know what you picked
the first time. Don't pick a star that's too close to the first one, as the second star lets
Nebulosity know how much rotation is present. The further away it is, the more
"leverage" you have.
Once the stars are picked, you'll be presented with a dialog asking you for a few
parameters to give to Drizzle. It asks for the "Pixel Reduction" factor (how much
smaller the pixels become before being transformed to the output grid) and the "Upsample" factor (how much bigger the output image is than the input images). Typical
pixel reduction factors range from 0.5-0.8 (0.6 is the default) and typical up-sample
factors range from 1.2-2.5 (1.5 is the default). If you try to use very small pixel drops
(e.g. a pixel reduction factor of 0.2) or very large up-sampling factors (e.g., 3.0), you
risk leaving "holes" in the output image where no pixels dropped from an input image
53
to the output image or other artifacts (right).
Finally, you'll be asked for an "Atomizer" value
(default value of 2). This parameter lets you trade off
speed and accuracy of the Drizzle process. Think of
it as how fine a "mist" is being made out of each
"drop". 1 is the fastest but is a bit less accurate and
should not be used if you have a small number of
images. 3 is the slowest but most accurate. 2
represents a good trade-off (and will rarely look any different than 3).
Be warned - Drizzle is
computationally intensive.
Be prepared to find
something else to do for a
while after you've picked
your stars.
Here is a shot of a small
portion of a test image
used to evaluate Drizzle.
On the left, we have a
single raw frame taken
from the DSS again. This
frame was shifted
randomly and
undersampled to create a stack of frames that where then aligned with either
translation (middle) or Drizzle (right). For display here, each image is shown at the same
scale. Note the separation of close stars recovered by Drizzle and its overall increase in
resolution.
14.3.5 Colors in Mo5on (Simultaneous transla5on align + Color reconstruc5on)
Colors In Motion (CIM) is a tool developed for Nebulosity that combines the process of
color reconstruction from one-shot color images, the alignment process, and the
stacking process in one step. This is done not as a convenience to the user but to
potentially achieve a more accurate and higher resolution final image by doing these
steps at the same time.
To use CIM you must have captured data in RAW format and you need to have a
sizable number of frames. In addition, you should most likely have Normalized your
images prior to stacking. Ten frames is the absolute minimum and it wants as many as
possible (30-50 frames would be reasonable starting points). Data should be preprocessed first but not de-mosaic'ed. Finally, your tracking should not be perfect.
Good or excellent is fine, but perfect is not. If you can make a sharp image by stacking
your frames without any alignment, don't bother trying CIM. If such a stack makes a
blurry image, you're in great shape for CIM.
54
Using CIM is much like using the normal alignment process. You will be asked to locate
the same star in each of the images you wish to align (Ctrl-click after the first image to
keep the star Nebulosity chose - Command-click on the Mac). Here too, you can skip
any individual image by simply Shift-Left-Clicking anywhere in the image. After picking
a common star in each, you are then prompted for a "CIM Threshold" (if you have at
least 10 images). This threshold marks the dividing line between "Full CIM" and "Partial
CIM" (Partial CIM blends a standard debayer version of the image with the CIM version
of the image). 10 is a reasonable value (lower numbers will force more pixels to be Full
CIM and higher numbers will force more pixels to be Partial CIM).
CIM is computationally intensive, so be prepared to wait awhile once you've told it
where the common star is in each image. When it's done, you'll be asked for a file to
save the results in.
14.4 AutomaAc Alignment (non-­‐stellar)
Finally, we have a fully automatic image registration option. It
can work well for star fields, but one of its main goals was to
be able to align images that don’t have clear reference points
(stars) and this is where it works best. The tool is found in the
Batch menu and when started, the dialog shown here appears
(the “Save stack” and “Diffeomorphic” options are currently
not available). The first thing to do is to press the Select
Master button and select a reference frame. If you hit OK, it
will proceed with the defaults and ask you for a set of target
frames to align. Note, this is not a fast process by any means.
You may want to keep the History window open (View menu)
to follow its progress. In the end, you will have a set of
images called “aa_OriginalName.fit” (or instead of “aa_”,
whatever you specified in the Prefix field).
If you decide to experiment with the parameters, here is what they mean:
Transformation Method
How should we stretch and squish the image to make it line up with the reference?
"Rigid" won't distort the image at all (translation + rotation only) and affine will
(translation, rotation, scale and shear).
Method
This controls how we are determining how well the images overlap. You have a pulldown with a number of options. By default, we have "MI 5%". Here, MI=Mutual
Information and MSE=Mean squared error. These are different "how much error is
there" metrics. For MI, you've got a few sampling rates (what percent of pixels or # of
pixels to work on). 5% seems to work well here and is a lot faster than if you go up to
higher rates. For MSE, there are two options: Full (use the whole frame) and Stars
55
(focus on things that look like stars).
Size
The amount of motion between frames can be accurately estimated without looking at
the whole frame or without looking at it at full resolution. By default, Nebulosity
downsamples the image to half size and runs on that to save time. You can run at fullsize or at 25% size as well or focus on just the central 512 pixels.
56
15 Image Adjustment
While Nebulosity is not designed to be an advanced image processing application like
PhotoShop or the GIMP, it does supply a number of purpose-built and very useful tools
for adjusting your images (usually the result of stacking). These tools are located under
the Image menu.
For any of these, if you decide you don't like the results, simply press Cancel or use
the Undo command in the Image menu (or Ctrl-Z). By default you have 3 steps of Undo
available, but with a quick trip to the Preferences menu, you can have unlimited Undos
and Redos.
15.1 Demosaic’ing and Pixel squaring
The tools in this section are primarily aimed at one-shot color cameras. They let you
convert the image into color (from a RAW Bayer-matrix grayscale image) or keep it as
monochrome, but reconstruct it in ways optimized for various filters.
15.1.1 Demosaic Image
If you captured a one-shot color image in RAW format and wish to see it in full-color,
use this tool. It will reconstruct the raw image and turn it into a full-color image for
viewing or saving. The algorithm used for this (there are several available) is set in your
Preferences. There is also a Batch version of this tool that will allow you to demosaic a
series of files. (Note, this should be done if you capture in RAW format after preprocessing but before alignment).
It is also important to note that if you use a Canon DSLR, ideal color balance in
Nebulosity will be accomplished if you select the appropriate setting under "DSLR
White Balance / IR Filter" in the Preferences dialog.
15.1.2 Square B&W If your sensor’s pixels are not square (i.e., they are a different height and width), you’ll
probably want to square them after pre-processing. For one-shot color cameras this
happens during the demosaic process. For black and white sensors, this happens
here. As with the demosaic process, you have tools to do this on the current image and
on a batch of images.
15.1.3 Reconstruc5ng Images from One-­‐shot Color Cameras and Line Filters
In a previous section (see Monochrome vs. Color?) a typical color filter array was
shown for a one-shot color camera that uses red, green, and blue filters. If we place a
"line filter" in front of such a camera, what happens? For example, suppose we placed
an H-alpha filter in front of this array? Such filters pass light in a very narrow range,
centered on 656nm of "Hydrogen alpha". Emission nebulae emit light specifically at
this wavelength (and several others), so passing this light and block all else can lead to
an excellent way to image nebulae amid light pollution and can lead to stunning images
of these nebulae. When multiple lines are imaged separately (e.g. one frame of Ha, one
57
of O-III, etc) they can be combined into beautiful "false color" images.
Typically, such imaging has been reserved for monochrome cameras. The reason can
be seen in that Bayer array. Photons that pass through the Ha filter are well into the red
area of the spectrum. As such, only the red pixels will get any light. The green and blue
pixels will be dark. Thus, we have only 25% of our pixels doing anything and the others
are merely contributing noise. So, when reconstructing RAW data, one could take the
RAW data and use the Low Noise 2x2 Bin or Adaptive 2x2 Bin tools. This would create
an image half the size but would remove all evidence of the Bayer pattern. The one
valid red pixel would be averaged with the three invalid pixels in a local 2x2 area and
the result would be dominated by the red signal. One could also use the normal
debayer routine and simply use the Discard Color tool. This is a common approach,
but one that may not be optimal (see image below).
While we cannot
escape the loss of
resolution entirely,
there are ways of
improving how
images are
reconstructed on
one-shot color
cameras when line
filters are used. For
example, when
CMYG color arrays
are used instead of
RGB arrays, more
pixels respond to Ha
light (as is shown by
this shot with an Ha
filter of the
Horsehead nebula,
courtesy of Michael
Garvin). Knowing
this, and knowing
how the pixels
respond to this light
can let us optimize
this reconstruction.
Nebulosity gives you several tools to do the reconstruction in addition to using the
binning tools or the Discard Color tool. One is a "Generic" method that will do a good
job on any line filter with any camera but is not optimized for any specific combination.
A second is a reconstruction optimized for "nebula" filters and O-III filters that leak light
in the Ha and beyond regions (e.g., Televue, Meade, and Lumicon filters). This is also a
58
rather generic reconstruction that will work well with a wide range of setups. Finally, for
CMYG arrays, there are optimized reconstructions for Ha and pure O-III filters (e.g.,
Astronomik, Orion, and Custom Scientific) that do not leak light in Ha or higher
wavelengths. A comparison of these techniques on the data from the Horsehead
nebula taken on an Orion StarShoot's CMYG array is shown below.
In addition, this menu has options for pulling out each of the color channels directly for
one-shot cameras with RGB sensors. Since the green pixels are twice as common as
the red and blue, you have options to grab either of the green fields or to average this
into a single green. Note, with these options, the image will be half the original’s size.
15.2 Geometric manipulaAons
The tools in this section let you change the size of the image and its orientation.
15.2.1 Crop
Usually, after stacking a series of images, you end up with a dark border in your
stacked frame. This is because Nebulosity had to move all the images around to get
them to line up and some needed to be moved further than others. To get rid of these
borders (or just to recompose your image), you can crop the image. One way to do so
is to simply drag a selection using the left mouse. Start in one corner and hold the
mouse button down to create a selection box and let go when the box is the desired
size. Once happy with the box, pull down Crop from the Image menu and you'll have
cropped in on just that area. A second way to do so is to pull down Crop from the
Image menu and enter the number of pixels to crop off directly
15.2.2 Mirror / Rotate
Need to rotate or mirror that image to get it to line up with something else (e.g.,
reality)? You can rotate 90 degrees clockwise or counter-clockwise, 180 degrees, or flip
your image up/down or left/right. Note, there are batch versions of these tools as well.
15.2.3 Binning
Images from some cameras are quite large and it can be useful to cut their size down
(e.g., to post to the Web). Nebulosity lets you bin images 2x2, thus cutting the image
size in half. An added benefit of binning is that by combining data from 4 pixels into
one, noise is reduced (much in the same way it is with stacking). Finally, one additional
use of binning 2x2 is to remove all color information from a RAW frame off a one-shot
color camera. This turns the one-shot color camera into something a lot closer to a
monochrome camera.
Nebulosity gives you 3 ways to bin your image. You can sum all 4 pixels (which will
brighten the image considerably), average all 4 pixels (which will keep the same
brightness) or perform an adaptive bin. The adaptive bin will combine the data in a way
between the summation and averaging, optimizing the combination so that the full
0-65535 range is used.
59
15.2.4 Resize
Want your image to be rescaled (aka “resized”, aka “resampled”)? The Resize tool in
the Image menu can make your image larger or smaller using a number of different
resampling algorithms. Simply enter in the scaling factor (e.g. “2” to double the size of
the image, “0.5” to make it half as big) and choose an algorithm.
Here, we have an
image that was first
reduced to half its size
by binning and then
restored to full size
with the different
algorithms available.
“Box” is fast but ugly
when increasing the
size of the image.
“Bilinear” is fast but
will smooth the image
a bit more than some
of the others. “BSpline” will lead to the
smoothest output.
“Lanczos sinc” can
lead to the sharpest
output but is more
prone to ringing than
“Catmull-Rom” or
“Bicubic”. The
Catmull-Rom usually
outperforms the
standard Bicubic.
15.3 Stretching and Intensity scaling
The tools in this section let you adjust the look of your image by stretching the contrast
and by performing other transformations to
the intensity of the image.
If you want even more control over the
values in these tools, double-click on the
15.3.1 Levels / Power Stretch
readout (Mac/Win) or on the slider (Mac)
The Levels / Power Stretch tool is a very
and you can specify values directly.
versatile and useful tool. Once you have
combined a number of images, you'll likely
notice that your sky is still bright and the DSO is quite possibly still dim. Overall, the
combined shot looks little like the wonderful shots you see posted on the Internet, even
by others with the same camera. The problem is that even though you've combined
60
many images, your skyglow and DSO are still at the same basic levels they were at
initially. Combining images does not make the result any brighter than the original.
What it does, though, is to make the result cleaner than the original. So, if your skyglow
had been at 10,000 ± 2,000 it may now be at 10,000 ± 100. If the faint bits of your DSO
were at 11,000 ± 2,000, they may now be at 11,000 ± 100. If we were to set the black
level of the image (just with the slider) up to 10,000 the sky would go quite black and
your DSO would remain.
The Levels / Power Stretch tool lets you
do this and quite a lot more. When run, it
The histogram shown here is based on the
presents you with three sliders and a
luminance values of the image. This may
window showing your image histogram (see lead to clipping of the data if you have not
above). One slider sets the black level and
already balanced the color channels.
another sets the white level. The third sets
Balance reasonably before stretching and/
the "power" (or middle slider in a "levels"
or keep an eye on the histogram shown in
tool - If you’ve used the Levels tool in
the main window (which is computed
Adobe PhotoShop© the only difference is
based on all color values).
that it reports 1/power). Leaving the power
at 1.0 performs a "linear stretch" of the
data. Setting the power below 1.0 will tend to brighten the fainter bits of the image.
Setting it above 1.0 will tend to darken the fainter bits and brighten the already brighter
bits. This is applying a specific kind of curve to your image (see below) by computing
for each pixel:
NewValue = OrigValue power
At the same time, it is stretching the data so that the output ranges between the values
set by the black and the white level sliders. You can see its effect in the histogram
window. Here, the initial histogram is shown in blue and the output, or resulting
histogram is shown as a dashed orange line. In the example shown here, you can
clearly see the histogram being stretched to pull out interesting bits in the data while
61
also resetting the black point.
To assist in using it this way, the blackpoint, midtone, and white-point lines are
superimposed on the histogram. These
lines show where in the original histogram
(blue line) the black (left line), midtone
(middle line) and white (right line) points
lie. You will note that as you move the
power slider, the midtone's position
relative to the white and black points will
move, but that it often won't be placed
directly under the slider. This is entirely
normal. If you wish to think in terms of
setting the midtone level of the image,
adjust the power slider until the middle
line (slightly darker than the other two) is at the desired place in the histogram.
There are a few things to note. First, if Auto-Scale is turned on prior to entering the
Levels tool, it will be turned off and the B and W sliders set at their full extent. This is to
show you how much of the full data range you are using and to encourage you to
stretch the image to use that full range. (If not in Auto-Scale mode, the sliders are not
moved). Second, the Levels tool can be quite computationally taxing on your computer,
especially if you are working with very large images. To make the adjustments more
responsive, try defining a region of interest (ROI) with the mouse (just as when
cropping) before entering Levels. You will preview the adjustments on this region only.
When you hit OK, the same adjustment will be done to the whole image.
Take your shot and try setting the power to 0.3 - 0.8 and then sliding the white and
black levels. You should notice that the faint bits of your DSO start to appear and
become a lot more prominent. Quite often, optimal results are obtained by using this
tool multiple times. Each time, make only moderate adjustments to the image and
don't worry about getting your background very dark the first time or so through.
Gradually hone in on your desired image. Don't worry about the fact that you're doing
this multiple times and that it might cause problems with the image range or values.
Remember, Nebulosity does everything in 32-bit floating-point numbers (96-bits per
pixel total) internally. Adjust and re-adjust as you see fit.
15.3.2 Curves The Curves tool dialog has two main areas. In the lower-right are some sample / default
curves you can try or use as a starting point. The “Keller” Curves come from Warren
Keller, author of the IP4AP tutorials. The main part of the dialog is on the left. There are
two control points (blue dots) that you can move to draw the curve. Grab one and
move it and you will see the red line (the curve you are making) move and the screen
will update to show the effect of the curve. These two points and the end points let
Nebulosity build a sensible curve (using a Bezier). While more points will give more
62
flexibility, they also can get you into more trouble. With two points many curves can be
drawn and with a second pass through, the combined effect of both curves gives
almost infinite control.
Note, in the Pre-set section,
there are entries for the last
curve you used and for a
saved-curve (created with
the “Save” button). These let
you come back and re-trace
out the curve used in the
past. Note also that the
Status bar and the History
dialog will record the
positions of both control
points so that you could recreate a curve another time.
15.3.3 Digital Development Processing (DDP)
Digital Development Processing (DDP) is a technique developed by K. Okano to make
CCD images look more like film images. Like the Power Stretch tool, it helps bring out
faint detail in the image and helps suppress skyglow. Okano's technique passes a
hyperbolic function to the data to make the linear CCD response much more like the Sshaped "gamma" response of film.
When you select Digital Development, four sliders are presented. Default parameters
for each are pre-selected.
The first slider, labeled Bkg sets a level for the background in the output image
(Okano's b parameter). The second slider, labeled Xover sets the cross-over point
(where the transformation shifts from a linear to a curved one - Okano's a parameter).
The third slider, labeled B-power provides a method for darkening the background
during the DDP process (not in Okano's description but may be set to 0 for a pure DDP
63
processing). Finally, the fourth slider, labeled Edge Enhancement controls the amount
of sharpening done during the DDP process (part of Okano's description).
15.3.4 Zero Min
If the lowest pixel value in your image is at 20,000 (resulting from a bright sky), why
would you want to let the rest of your data only use the remaining 44,535 possible
values when you save? You could have a lot more "room" for a range of intensities if
you shifted the whole image intensity down by 20,000. This tool will do that
automatically for you, by making subtracting the current minimum value from each
pixel.
15.3.5 Scale Intensity / Pixel Math
You may find times at which you wish to re-scale the brightness of an image in a very
controlled way. For example, if you use the Adaptive Stacking (see Preferences)
technique and have stacks taken with two different exposure durations, you might want
to rescale the intensity of one before combining them. Or, you might simply want to
brighten or darken the image, shift its intensity up or down, etc. The Scale Intensity tool
lets you do a number of things under the general heading of “Pixel Math”. You can
multiply or divide each pixel by a constant and you can add or subtract a constant
from each pixel. You can also transform the image by taking the log, square root, or
arcsine of each pixel’s value. (Using these, the image is normalized in the process). You
can also invert the image.
15.4 Color adjustment
Even if you go to great lengths to get your color balance “right” ahead of time, your
shots will almost always have color issues when you start to process them. Some of
this comes from camera issues, some from refractive effects in the atmosphere, and
some from your skyglow’s color. Tools in this section let you get the color just right.
15.4.1 Adjust Color Offset
When taking images in color, it is
often the case that the background
has a slight color hue rather than
being a neutral grey. This can be
the result of a color "bias" in the
image - the minimum value in each
color channel not being the same
(skyglow will give you this). This
can be fixed by subtracting a
constant number from one or two
of the color channels. The Adjust Color Offset tool lets you do this. A dialog box will
appear with sliders for red, green, and blue. Nebulosity will attempt to determine
reasonable values for the sliders when the dialog opens. The values you enter here will
be subtracted from the specified color channel(s). For example, sliding the Red slider
to 1100 will not affect the green and blue data but will make every red value 1100 less
64
than it was previously. A 3-color histogram is shown below the sliders to help in getting
the offset just right. Aim to have the left edge of the histograms similar for all three
colors. Pressing Cancel will revert back to the original image.
15.4.2 Adjust Color Scaling
In color imaging, it is often the case also that your color channels are not balanced in
their scaling either. This is particularly true in using separate color filters in a black and
white camera (if one does not perfectly adjust exposure duration to compensate for the
varying transmission of color filters and for the CCD's varying sensitivity to different
colors) but can occur in a one-shot color camera as well. The Adjust Color Scaling tool
lets you rescale the intensity of individual color channels much like the Scale Intensity
tool lets you rescale the intensity of the entire image. Setting the Green slider to 1.05
would, for example, multiply the intensity of each pixel's green component by 1.05. A
3-color histogram is shown below the sliders to help in getting the offset just right. Aim
to get the histograms similar in size to color balance the image. Pressing Cancel will
revert back to the original image.
15.4.3 Auto Color Balance
Rather than running the color offset and scaling tools (above) separately, you can have
Nebulosity automatically stretch and scale the channels to fix global color casts (it
does this by the same histogram matching routine used to normalize images’
histograms to a reference image).
15.4.4 Hue / Satura5on
While the color of your image can be adjusted by scaling each color channel separately
you can often have better control over it when making “by eye” adjustments using the
Hue, Saturation, and Luminance tool. Hue gives you control over the overall color
shade, saturation gives you control over how rich the colors appear, and luminance
gives you control over how bright the image is.
15.4.5 Discard Color and Convert to Color
At times, one may want to take a color image and strip it of all color information. This
can be useful, for example, if you want to make a luminance (L) channel for image
processing outside of Nebulosity or for using during pre-processing of monochrome
images. The Discard Color tool does just this, converting your color image into a
monochrome image (the average RGB value is used). The Convert to Color does the
converse. It takes a mono image and makes it into an RGB image (with every pixel
having R=G=B).
15.4.6 LRGB Color Synthesis Tool
This tool lets you combine separate L (optional), R, G, and B frames into a color image.
It is described in detail here.
15.5 Sharpening, Blurring, and Noise ReducAon
After processing our images, its usually necessary to perform some kind of sharpening
65
on them to bring out the details and make it crisper. At times, it’s also useful to blur
your image (to reduce noise or remove artifacts). Tools in this section assist with these
manipulations.
15.5.1 Sharpening and Tighten Star Edges
Often, lack of precise focus or seeing will end up making your images slightly blurry.
There are numerous techniques astro-imagers have used to fix this. Nebulosity
provides several tools for sharpening the image.
One very popular tool is an Unsharp mask. This algorithm actually makes a blurred
version of your image and subtracts this blur to make a sharper image. When you pull
down Sharpen Image, Unsharp Mask a dialog appears with two sliders. The “Amt”
slider is asking how much sharpening should be applied (the amount of sharpening).
The “r” slider is asking the amount of blur in the image - specifically, the radius of the
Gaussian blur that is in theory in the image. Thus, this controls the size of details
you’re working on sharpening.
Nebulosity also provides a tool that takes a different approach to the problem. The
Tighten Star Edges tool first examines your image and performs an edge detection
analysis using a modified version of the Sobel Edge Detector (modified to work better
with our round stars than the traditional Sobel). These edges are then subtracted from
your image to yield tighter stars and enhanced edges. Note, this is not the same kind
of edge enhancement done during DDP processing.
This is a shot of M57 as acquired (left) and after the Tighten Star Edge tool using the
default parameters. Using the slider that appears when you run this tool (located in the
Image menu), you can adjust the degree of edge enhancement applied.
15.5.2 Gaussian Blur
Nebulosity lets you apply a Gaussian blur to your image. This is another nice way of
reducing noise. For example, prior to applying a flat frame to your image it can be
66
useful to blur the flat first to reduce noise in the image. Small amounts of blur can also
be used after over-sharpening.
15.5.3 3x3 Pixel Median
One final way of smoothing the image is to apply a median filter to it. In this case, a
sliding window of 3x3 pixels is passed over the image and the median value is put in
place of the center pixel in that window. Thus, each pixel becomes the median of it
and its neighbors (the median being the middle-most number if their intensity values
were sorted).
15.5.4 Ver5cal Smoothing (deinterlace)
Some CCD sensors are “interlaced” while others (and CMOS sensors) are
“progressive” scan devices. Interlaced sensors read every other line in one pass and
then come back to read the skipped line in a second pass (e.g., 1, 3, 5, 7 … 2, 4, 6, 8,
…) while progressive sensors read the image out in order in a single pass (e.g., 1, 2, 3,
4, 5, …). Since it takes time to read out the image (or half the image) interlaced sensors
can be prone to a “Venetian Blind Effect” that makes the image appear to have fine
horizontal lines. In truth, there are no lines on the image but one set of lines (e.g., the
even lines) got a little bit longer exposure than the other set of lines (e.g., the odd lines).
Most cameras that use interlaced sensors have some means by which they try to
reduce or eliminate this effect, but at times it still comes through in the image. You can
remove it with the Vertical Smoothing tool. When started, a dialog will appear with a
slider that controls how much smoothing is applied. Adjust it until the artifact is gone.
Don’t overdo it as this is smoothing your image.
15.5.5 Noise ReducAon
Noise is in all of our images and nowhere is it more apparent than in the background
sky or the faint bits of the DSO. Here, the real signal we care about is so close to the
thermal noise, the read noise, and the skyglow’s shot noise that after stretching things
can look downright ugly. One solution is to not stretch so much. Another is to stretch
but then try to reduce this noise. Nebulosity has two tools to do this: Adaptive Median
Noise Reduction and GREYCstoration Noise Reduction.
15.5.5.1 AdapGve Median Noise ReducGon
The Adaptive Median Noise Reduction scheme works off of the premise that the noise
you want to reduce is all in the darker areas. A dialog appears with a slider when you
start the tool that asks for a threshold. Raising the threshold will have more of your
image subjected to the filter and lowering it will have less. What this is doing is
blending your original image with a median-filtered based image with the shift from one
to the other based on your slider position.
15.5.5.2 GREYCstoraGon Noise ReducGon
The GREYC Noise Reduction tool was authored by GREYC Labs and is an incredibly
powerful tool. One of its main components is a noise reduction tool that removes local
variations in intensity while respecting sharp boundaries in the image (e.g., our nice
67
high contrast stars, transitions from
sky to dust lanes, etc.) What is
included here in Nebulosity is their
command-line tool and an interface
to this tool. You don’t need to use the
command line tool at all as the
integration with it is designed to be
as seamless as possible and to make
it look as if it is one of Nebulosity’s
own tools.
GREYCstoration has a large number
of parameters you can set and the
dialog that appears when you start it
is more like the dialog you see in
many other programs. An array of
parameters with arcane names appears and you may have no idea just where to go
from there. At the very least, the defaults entered here should be a reasonable starting
point (and differ from GREYCstoration’s own defaults). That said, odds are you will be
trying a lot of different parameters as you hone in on what will do the best noise
reduction in your image without loosing vital details.
In this before and after example, there is clearly an improvement noise-wise. The
background is a lot smoother and overall the image is still quite sharp. It’s not perfect,
though, and still could benefit from some tuning of the parameters. There are a few
things that will help you do this:
• Before you enter GREYCstoration, select a box around something that has
stars, background, and your main object. You will be able to preview things on
just this region, which will make for much faster refreshes of the screen to see
the effect.
68
•
•
•
On entering GREYCstoration or after a change in parameters, hit the Preview
button to see the effect. You can hit the Show Original button to revert the
display back to the original image to see what was gained and what was lost.
You can change parameters as much as you like hitting Preview between each
iteration. Preview always goes back to the original data. Hitting Done will apply
this to the whole image (and can take a long time – watch the title bar).
Read the GREYCstoration documentation with examples.
The “Fast approximation” and nearest neighbor resample methods do speed
things up but don’t produce the best output.
Watch this space for more details of maximizing GREYCstoration’s effectiveness.
69
16 Other Tools
The File and Edit menus have several other tools that are worth noting:
16.1 Launch New Instance
Ever want to work on two images at once? Or, ever want to control two cameras at
once? Simply pull down Launch new instance from the File menu and you’ll have
another copy of Nebulosity running. Launch as many as you like!
16.2 Change Language
English not your cup of tea? (Sorry, I couldn’t resist.) Pull down Change language from
the File menu and you can transform Nebulosity’s menus and dialogs into several other
languages following a restart of Nebulosity. Please note, I do not speak any of these
other languages, so no help can be offered if you send me e-mails in German, Dutch,
Italian, French, etc.
16.3 Image Info and FITS Header
Want to know when you took an image? How about what the CCD temperature was
when you captured the frame? If the image is open, Image Info in the Edit menu will
tell you these sorts of things and more. For any FITS image, you can get this
information (and all information in the FITS header) by using the FITS Header tool in the
File menu.
16.4 DSS Loader
Ever wonder what a target will look like
on your camera with your telescope
before heading out to image? Of course
you have! The DSS Loader tool lets you
see what kind of framing you’ll have
using real images from the Digitized Sky
Survey. Start by entering the name of a
target (or coordinate position) in the top
field. Then select a desired Field of View
(FOV). Don’t worry about being precise
here yet, just pick something bigger than your rig will show. At this point, you can hit
Download Image and it’ll query the DSS (via Skyview) for an image and put it onscreen
(it may take a bit to download the image). Now, feel free to enter in values in the X-FOV
and Y-FOV field corresponding to what your scope + camera will show and you’ll see
selected on the image your exact FOV. Don’t know your current FOV? Click the X-FOV
or Y-FOV buttons and a calculator will appear to figure this out for you. Note, the
region is actually selected so if you like, you can now crop this out.
16.5 Measure Distance
Knowing how far apart several stars are can be quite useful. You can use it to calculate
the effects of focal reducers (measuring the distance between stars in prime-focus and
70
reduced images) or to say, "Hey is this pair of stars on my image really that pair of stars
in the atlas?" The Measure Distance tool lets you do this.
First, select up to 3 stars in your image by simply right-clicking on them in the image.
The first will get a red circle around it, the second a green, and the third a blue. If you
make an error, either Shift-R-Click to erase all points or keep selecting stars (you'll
cycle back to red after blue). Next, pull down Measure Distance from the Image menu.
You'll be asked for the resolution of the image in arc-seconds per pixel. (If you don't
know, simply use 1.0 as a value and all values will be in CCD pixels or see use the
calculator built into the DSS Loader tool or the formula shown in Your Telescope). A
window will then appear showing you the distance from red to green and, if three
points were selected, green to blue and red to blue.
16.6 Check / Update License
If, for some reason, you need to update your license code or see what your license
code is, this entry in the Edit menu will let you do so.
71
17 Supported Cameras
Nebulosity supports a wide range of cameras on both Windows and OS X. They are:
1. Canon DIGIC II, III, and 4 DSLRs (Windows and OS X): EOS 1100D/T3, 1000D/
Rebel XS, 450D/Rebel XSi, 400D/Digital Rebel XTi, 500D/T1i, 550D/T2i, 600D/
T3i, 350D/Digital Rebel XT, 60D, 50D, 40D, 30D, 20D/20Da, 5D Mark II, 5D, 7D,
1D Mark IIV, 1D Mark III, 1D Mark II N, 1D Mark II, 1Ds Mark III and 1Ds Mark II.
See below for more details.
2. Fishcamp Starfish
3. Meade DSI, DSI Pro, DSI II, DSI II Pro, DSI III, and DSI III Pro.
4. QSI 500 and 600 series
5. QHY8
6. SBIG (See below for more details)
7. Starlight Xpress SXV / SXVF / SXVR USB cameras (See below for more details)
In addition, on Windows, the following cameras are supported
1. Apogee Alta
2. Atik 16 series (all) / Artemis 429/285 cameras
3. Atik “Universal” (3xx, 4xx, 4000, 11000, etc).
4. CCD Labs Q285M / QHY2Pro
5. FLI
6. Moravian G2/G3 (v3 firmware or higher)
7. QHY8 Pro
8. QHY9
9. Opticstar DS-335, DS-336C, DS-145, DS-142, and PL-130
10. Orion StarShoot Deep-Space Color Imager (Other Starshoot cameras supported
via ASCOM)
11. SAC10
12. SAC7 / SC1 long-exposure modified webcams / Atik 1 and Atik 2 cameras. See
below for more details
13. Any camera with an ASCOM v5/v6 driver (See below for more details)
17.1 SBIG
Almost all SBIG cameras are supported by using SBIG's Universal Driver (some very
old ones are not supported by this driver and therefore not supported in Nebulosity).
This must be installed first (see SBIG's website if you have not already used SBIG's
software on your computer). On Windows and OS X, USB and Ethernet versions are
supported. On Windows, but not OS X, parallel port models are supported. Note,
however, that if you SBIG is a dual-chip model, the guide chip will not be available to
any other program when the camera is in use in Nebulosity.
17.2 Starlight Xpress
While older USB-1 based Starlight Xpress cameras are not supported, all USB-2 based
cameras (i.e., the "SXV" and later cameras) are supported.
72
The "Fast" mode available on the M (and MX) cameras is enganged whenever "High
Speed mode" is selected in the Advanced Panel. On these cameras, short exposures
are done via the "Interlaced" mode and longer exposures (generally ~1s or longer) are
done via the "Progressive" mode.
If you have an older parallel-port camera, you can convert it to an SXVF version via
Starlight Xpress’ USB2 converter box. Nebulosity will detect and use cameras that
have this device attached.
17.3 Canon DIGIC II, III, & 4 DSLRs
The Canon DIGIC II, III, and 4-based DLSRs are supported in Nebulosity on both the
Windows and Mac OS X platforms via Canon’s EDSDK (software development kit).
Note, older DIGIC I-based cameras (D30/60, 1D, 10D, and 300D) are not supported by
their EDSDK and therefore not supported in Nebulosity. Also note that Canon does not
support all of their cameras under Windows 7 and is phasing out support of older
cameras by the EDSDK (notably the DIGIC II based cameras like the Rebel 350).
The supported DSLRs act much as any one-shot color astronomical camera would in
Nebulosity. This can cause confusion for people when they first try to capture images.
Images will come up looking monochrome, the interface on the camera always says
“busy”, they get FITS files rather than CR2 files, and their white balance settings don’t
seem to have any effect. What the...?
Some of these things can be changed in the Preferences, but some cannot. What you
can’t change is setup this way for a reason, however. It will say “busy” as Nebulosity
has locked the camera and is preventing you from doing things like changing the
shutter speed or the ISO value. When you want to change these, you change them in
the Nebulosity interface (exposure time and gain / ISO) so Nebulosity knows these
values. Likewise, it will ignore your white balance settings (as skyglow will throw the
color balance off anyway - see below). It also won’t let you select JPEG or RAW
directly, but it will do so as needed based on what you’re doing. When you hit Preview
or Capture Series, it will take a RAW frame. When you hit Frame and Focus and Fine
Focus, it will shift into a JPEG mode for high speed (but lower quality) readouts - just
the kind of thing you want when trying to compose or focus.
17.3.1 Color images and FITS/CR2
Just like any one-shot color camera, you can opt to acquire images in pure RAW
(Bayer-matrix images that have not been converted to color yet), or in RGB via the
Preferences. Here, you can choose whether you wish to optimize speed (high-quality
JPEG images are downloaded from the camera) or quality (the RAW image is
downloaded, the Bayer matrix extracted, and Nebulosity's Demosiac routines are
applied). RAW Bayer-matrix images are the best choice when the highest quality is
desired as they allow you to pre-process the images before conversion into color, and
the color conversion (Demosaicing) is done using Nebulosity's best routines.
73
One thing to note here is what we mean by the term “RAW”. Some people feel that
“RAW” means “CR2” or “NEF” or “CRW” format. These formats are just containers for
image data. FITS is another container for image data as is TIFF or PNG or JPEG. All of
these except JPEG can save the full 16-bit, pure data off the sensor with no loss
whatsoever. “RAW” is just that - the raw sensor data. It has nothing to do with the
container format per se.
By default, Nebulosity saves images from your DSLR in FITS format. It saves the raw
sensor data in this container rather than in CR2, CRW, or whatever is the native format
for your camera. But, fear not, it is storing the raw sensor data. It just doesn’t store
things like the focal length setting of the zoom lens (that’s probably not hooked up to
the camera since you’re probably shooting through a telescope). (Aside - it’s not like
CR2 isn’t odd in it’s own right. It’s a lossless JPEG-compressed bit-packed image
stored inside a massive TIFF-like hierarchy.) If you want CR2 files, though, you do have
an option to save them in addition to the FITS files. Preferences, DSLR Save location
will let you choose Computer (FITS only), Computer FITS and CR2, and even to save
them on the CF card.
17.3.2 White balance
One cannot read a setting on the camera that lets you know if the camera uses a stock
IR filter or if it has been modified to have an extended-IR response (e.g. using a Baader
IR filter or a Huetech filter). Color reconstruction must be done differently if an
extended-IR filter is used, however. To ensure accurate color, make sure you have
made the appropriate selection in the Preferences, Processing, DSLR White balance /
IR filter section. If you’re not sure which one is correct, try taking a normal daytime shot
(a CR2 file shot of something outdoors with your DSLR lens is just fine) and try the
various options until you find one that gives a good color balance. Note, that if you see
pink cores to saturated areas, try the Straight Color Scale option. Keep in mind that
skyglow, atmospheric refraction, etc. will always muck with your colors, so there is no
need to get this spot-on. Every night will have a bit different color and every time you,
like all other imagers, will need to work the color balance.
17.3.3 Long exposures / bulb triggers
Canon DSLRs can be classified into several generations based on the DIGIC processor
used inside. With the advent of the DIGIC III and 4 DSLRs (circa 2007 and beyond), the
cameras can take exposures of any length via a special “bulb” command. By default,
Nebulosity will try to use these commands when you connect to a Canon DSLR
(“DIGIC III onboard” in the Preferences - see below). The only downside to this
arrangement is that you cannot use mirror lockup if you are using this method for long
exposures.
Older, DIGIC II DLSRs (e.g., 350 XT) are limited to 30s exposures when the camera's
internal timing is used. If you connect the USB cable to the camera and your computer
74
and this is your only connection, you will be limited to 30 seconds of exposure. To
achieve longer exposures, some form of "bulb adapter cable" is required . Various
forms of this exist to drive the camera's "bulb" setting from USB, serial, or parallel
ports on your computer. If you use one of these to allow you longer exposures you
must also keep the USB data cable connected (i.e. you will have two cables
connected).
In the Preferences menu, you will find options to let you choose your long-exposure
cable setup. Nebulosity supports the following:
• ShoeString DSUSB or DSUSB2 USB adapter (PC and Mac)
• Serial port adapters (COM1-8) such as Hap Griffen's (PC and Mac)
• Parallel port adapters (Pin 2 or Pin 3 on ports 0x378, 0x278, and 0x3BC) such as
Hap Griffen's (PC only)
• Onboard "blulb" timing for DIGIC III and 4 DSLRs (default)
Note: Not all USB->Serial adapters will work as we need direct control over several
data pins on the serial port. Generally, those
that require a driver disk will be more likely
Select your long-exposure adapter prior
to work (tested with inexpensive Prolificto connecting to the camera. Nebulosity
based adapters and Keyspan adapters).
attempts to connect to the long-exposure
adapter when it connects to the camera
itself.
17.3.4 Mirror lockup
The Canon cameras all support a mirrorlockup mode. When enabled, pressing the shutter button once lifts the mirror and
pressing it a second time triggers the exposure. Enabling this mode involves entering
your camera's Tools menu and finding the appropriate Custom Function (CFn) - usually
either 12 or 7, selecting and enabling this prior to connecting to the camera in
Nebulosity. Nebulosity has no way to set this mode itself, but it can detect if the
camera is in mirror-lockup mode and, if so, will send the appropriate shutter pulse prior
to the main exposure.
Unfortunately, if using the onboard "bulb" mode for DIGIC III or 4 DSLRs, you cannot
use the mirror lockup mode. There is no way to trigger a lockup over the USB line.
Thus, for mirror-lockup mode to be used, you must have a long-exposure cable
selected and attached. Another consequence of this is that mirror-lockup cannot be
used in USB-only mode.
17.3.5 Mode dials and lens seings
Nebulosity will attempt to read the current setting of your camera’s mode dial and
inform you if it is set improperly. For most cameras (those without a “B” setting on the
mode dial), the proper setting is “M” on the mode dial and if you have a DSLR lens
attached, it should be in the “MF” position. Nebulosity will program the camera and
use the onboard timing for short (<30 s) exposures and use whatever long-exposure
adapter (including the onboard DIGIC III/4) for longer exposures.
75
For cameras with a “B” setting on the mode dial (e.g., the 5D Mk II), things get a bit
more complex. On these cameras, if you want >30 s exposures, you must set the
camera to the “B” position on the mode dial. Unfortunately, we cannot program short
exposures in this mode. We can manually time short exposures, but you will never get
a precise 1/1000 s exposure this way. To get precise timing for short exposures, have
the camera in the “M” mode. As with the other cameras, any attached lens should be
set to “MF”.
Note also, that onboard noise reduction (automatic dark frame collection and
extraction) should be turned off. You are better off with a set of darks to combine and
subtract from your lights than a single dark. Your final SNR will be substantially better
if you do this and use that time imaging than if you take a single dark and subtract it
after each image.
17.3.6 Troubleshoo5ng Connec5ons
There are several steps to take to ensure Nebulosity will connect to your camera.
1. Ensure that the USB data cable is connected (in addition to any long-exposure
"bulb" cable) and that the camera did not decide to shut itself off from inactivity
(once connected, Nebulosity will keep it turned on)
2. Ensure that your camera drivers (from the Canon CD) have been installed on
your computer.
3. Verify that the EOS utility is installed and that it connects to the camera and can
be used to control the camera.
4. Ensure that the camera is set to communicate (Menu, Tools, Communication) to
the computer in "PC Connection" mode.
17.4 ASCOM Cameras
Nebulosity supports image capture through ASCOM v5/v6-compliant cameras under
Windows. This means that any camera with an ASCOM camera driver can be used
with Nebulosity, thus greatly expanding the number of cameras that can be used.
There are several things to know about ASCOM cameras, however.
1. For your ASCOM-compliant camera to work, you must install both the ASCOM
platform itself (www.ascom-standards.org) and the specific driver for your
camera.
2. When you attempt to connect to an ASCOM camera, an ASCOM dialog (the
“Chooser”) will appear. This is not a Nebulosity dialog, but one of ASCOM’s. If
your camera is not listed in there, your camera driver is not properly installed.
Consult your camera maker or the author of the driver for assistance. If the
ASCOM “Chooser” does not appear, ASCOM itself is not properly installed.
3. In Nebulosity v2, there were two styles of ASCOM connections corresponding to
the two ways ASOCM v5 let you talk to the camera: “early” and “late” binding.
The meaning of this is too arcane for users here and ASCOM dropped “early”
binding. Here in Nebulosity 3, all binding is “late”.
76
4. Not all of your camera’s features may be present. ASCOM’s camera driver
specification may not include features unique to your camera.
17.5 SAC7 and Long Exposure Webcams
The SAC7 and other cameras based on long-exposure modifications to webcams
according to the work of Steve Chambers (otherwise known as "SC-modified"
webcams) are supported in Nebulosity. There are a few things to note, however.
First, these cameras can either be used directly when attached to a parallel port (you'll
need to know the "port number") or via a USB port. Many modern computers lack a
parallel port or only include one that is derived from an internal USB-Parallel adapter.
Unfortunately, these adapters do not make the right signals for the cameras to operate.
ShoeString Astronomy has devised an excellent solution with their "LXUSB" product.
This plugs into a USB port and, when controlled by Nebulosity, will make signals that
can be used to fully control these cameras. Thus, cameras like the SAC7 can now be
used on machines that only have USB ports and that do not have a parallel port at all!
Second, these cameras typically have two modes - a "short exposure" mode and a
"long exposure" mode. In short exposure mode, the camera's own controls adjust the
exposure duration (typically 1/25th of a second or shorter) via a pop-up window. In
long exposure mode, the program (such as Nebulosity) controls the exposure duration.
In Nebulosity, short exposure mode is selected by setting the exposure duration to 0.
Anything greater than zero will put the camera into long exposure mode. In short
exposure mode, the shutter speed is controlled via a pop-up window. Press the
Advanced button in the Camera panel and you'll get a slightly different version of the
Advanced dialog described above. You'll find Setup and Format buttons that let you
configure the resolution, shutter speed, gain, frame rate, etc.
Note: For the best images in both short and long exposure modes, always set
the frame rate to a low setting such as 5 FPS. This minimizes the amount of
compression your images undergo. Do this in the Advanced Dialog using the
Setup and Format buttons
In addition to these buttons, the Advanced Dialog has one added section for these
cameras. A "Read delay" can be entered. The default value should work on most
systems but if you find you are dropping frames, try adjusting this value (5 ms
increments will be good). System speed and specifics of your camera may dictate a
slightly different value (10ms - 30ms for a typical range).
17.6 Camera Advanced Panel
If you click on the Advanced button in the Camera panel, a dialog box will appear that
lets you set various advanced controls on the camera. Some of the options may be
grayed out or not even present. If so, this means either that the current camera does
not support this feature or that some other feature is preventing it from being activated.
77
Below is a description of the available options.
• Amp off: Checking this box will have the CCD camera's amplifier turned off
during the exposure. This amplifier, when on causes "amp glow" - a brightening
usually in one corner of the image. Rarely would you ever want to uncheck this
box.
• Double Read: This option enables a feature designed to fix the "interlacing
artifact" found on interline transfer CCDs like the SAC10 and Orion StarShoot.
You'll notice each exposure takes twice as long but that you end up with a
smoother image. This option is particularly useful for shorter exposures where
the problem is worse (4 seconds of read time is a lot in comparison to a 2
second exposure but not much in comparison to a 5 minute exposure). If you're
working with bright objects or short exposures, you'll want to use this or the
VBE option.
• High speed read: If selected, the camera will read the image off the CCD more
rapidly but at the expense of increasing the noise. This is enabled by default
during Frame and Focus but is not to be used for DSO imaging. Selecting both
High speed read and Double Read is an excellent way to take good planetary or
lunar shots.
• Bin mode: Selecting this option will put the camera into a binned mode whereby
pixels are combined during the CCD readout itself. This increases sensitivity at
the cost of resolution and, at times, at the cost of color on one-shot color
cameras. The available bin modes depend on your particular camera.
• Enable TEC: You can set the cooler on your camera to a certain temperature (if
the camera supports this)
• Close shutter: If you want to take dark frames, setting this will keep the shutter
on your camera closed (if your camera supports this).
• Oversample: If selected the camera will sample and convert the information from
the CCD twice. The net result is a less noisy image, but one that takes a bit
longer to read and process.
• VBE balance color exp times: (SAC10 only) This feature attempts to fix the same
problem addressed by the Double Read option (the problem is sometimes called
the Venetian Blind Effect), but to do so with a single exposure. It intelligently
balances the intensity of the odd and even lines and can be quite useful for
shorter exposures.
78
18 Taking Good Images
You didn't buy a camera to take dim, noisy, fuzzy images, yet chances are quite decent
that's what you could get your first night out. How do those pros make such good
images? This guide won't make you a pro, but it will at least get you started in the right
direction. For help on any of these, consider joining the Stark Labs Yahoo Group. In
addition, head on over and have a look at the series on Signal to Noise I did for Cloudy
Nights.
18.1 Your Telescope
CCD cameras are not as forgiving as your eye and can be used to reveal any flaws you
have in your telescope. It's time to make sure it's in good shape by checking:
• Is it well collimated?
• Can you rigidly mount the camera to it or is there play in the focuser or
attachment?
• Have you got a good handle on dew prevention? (The author has more than
once taken a long series of exposures only to realize he was shooting through a
solid layer of due on the telescope objective.)
• Is it well-matched to the camera?
This last is, in and of itself, a rather lengthy topic with some disagreement as to what is
the absolute best match, but a few things can be agreed on. The most critical aspect
of this is to determine just how much sky each pixel covers using your telescope. That
can be done with the following simple formula:
Arcseconds per pixel =
206.265 × PixelSize
FocalLength
For maximum resolution, with perfect tracking (see below) and excellent seeing, a value
of 1"/pixel is a good target (some pros go to slightly smaller values still). For more
typical conditions with good seeing and good tracking, 1.5-2"/pixel is another fine
target. Larger amounts of sky covered per pixel will let you cover more sky and will not
stress your mount's guiding accuracy as much (see below), making values of 3-6"/pixel
quite reasonable for many situations. In so doing, you are trading off extreme resolution
for wider swaths of sky and less difficulty guiding.
From this formula, you can see that there are two ways to adjust the final resolution in
your image. You can either adjust the pixel size of the camera or you can adjust the
focal length of your telescope. Neither seem trivial at first glance and, while they can be
adjusted, it is only to a limited degree. (Telescope focal length can be shorted with a
focal reducer and lengthened with a Barlow. CCD pixel size can be effectively
increased by binning.) Thus, determining what telescope to use for a given camera or
vice versa is often best done before purchase.
79
18.2 Your Mount
A number of aspects about your mount will affect the quality of your images. Here,
we'll talk about accurate polar alignment and about periodic error and guiding.
Note: To see how much your mount is moving between images, right-click on a star to
lay down a "target" circle around it. This target will remain in the same place on the
image across captures, and let you see how far that star has moved.
18.2.1 Polar Alignment
If you've got an equatorial mount, aiming at Polaris with the RA and DEC zeroed will
get you somewhat close to polar alignment but not close enough for imaging. Using a
polar alignment scope on your equatorial mount or using your GOTO mount's
alignment procedure will get you closer. But, neither will get you spot-on enough for
long-exposure work. To do that, you've got two main options:
1) Drift alignment. In this technique, you watch how stars drift through the field and
adjust your mount accordingly. For a standard equatorial mount, this is your best bet.
It takes a bit of practice, but, once you know what you're doing, it'll take about 30
minutes to get a decent drift alignment - decent enough for the kind of exposures you'll
be able to do without guiding anyway. Others have done a good job describing the
technique, including a site by Pete Kennett or one by Scott Tucker of Starizona.
2) Iterative alignment. If you have a GOTO scope, you owe it to yourself to learn how
to do this. I can get a nice alignment that won't drift (well, periodic error of course, but
no overall drift) for an hour worth of imaging in about 5-10 minutes of work. There are
several sites on the web that go over the method (e.g., Michael Covington's), but the
basic idea is very simple.
First, do a "one star" alignment - the kind in which the scope aims itself towards where
Polaris should be (using the Kochab clock if that makes sense to you) and then asks
you to adjust the mount physically to center Polaris. After centering Polaris, the scope
slews over to a single star and asks you to use the keypad controls to center that other
star. At this point, you're close and have done the standard "one star" alignment.
Now, tell the scope to GOTO Polaris. Adjust the mount physically to remove about half
the error between where it ended up and where Polaris actually is (i.e., have it aim to
the spot about halfway between the GOTO and Polaris). Now, do a GOTO back to
your alignment star, center, and SYNC to that star. Repeat a few times until your
GOTOs on Polaris end up without any error and you're good to go!
18.2.2 Periodic Error and Guiding
Now that your scope is polar aligned, the stars won't drift across the field on average.
You'll still notice that they will rock back and forth a bit - sometimes very slowly and
sometimes in abrupt jumps. This is called periodic error and caused by minor
imperfections in your mount's worm gear - a cylindrical gear that actually turns the
80
telescope to counteract the earth's rotation. No worm gear is perfect, but some have
bigger problems than others. If you don't choose to guide your telescope during
imaging, these imperfections will limit how long you'll be able to expose each image.
Exactly how long you can go will depend on the size of the periodic error and the
amount of sky covered by each CCD pixel. Wide-angle shots with 10"/pixel are a lot
more tolerant of periodic error than zoomed-in shots at 1"/pixel.
Some amateurs run shots unguided and end up stacking many 15-40s long exposures
into one long image. With enough images, and with the right exposure settings (see
below), this can be used to make very nice images.
But, what can you do to lengthen this time or to fix the problem entirely? Several
mounts offer Periodic Error Correction (PEC). On these mounts, you train the telescope
to know what the error is like by following a single star and correcting the errors using
the telescope's controller. The mount then learns these corrections and applies them
automatically. This can reduce the error quite a bit.
A second technique, often used on its own or in conjunction with periodic error
correction is guiding. Here, an image of a star is sent to either an eyepiece (manual
guiding) or a second camera (autoguiding) while your main imaging camera collects
pictures. Two approaches are taken. In one, an Off-Axis Guider is used to split some of
the light away from the main camera and towards this eyepiece or second camera. A
small prism is placed so that the light split off is light that would not have fallen on your
main imaging camera anyway. In a second, another telescope (a guide telescope) is
attached to the imaging telescope. In both, this second view of a star is used to
determine when the telescope is drifting slightly off target and to correct this problem
by sending very small movement commands to the mount.
Many packages are out there to help you autoguide your mount. A free one from Stark
Labs, PHD Guiding, works well on a wide range of cameras and mounts and is
designed to be "Push Here Dummy" simple. Its goal is to make it so that you have little
excuse for not trying autoguiding.
18.3 Focus
Getting your camera sharply focused is critical to taking good pictures. The Frame and
Focus routine will get you close, but will often not get you to as sharp a focus as you
could get. For this, you'll want to make sure you're using the full Preview mode or the
Fine Focus mode, making only small adjustments to your focus between each shot.
You can evaluate your focus by simple visual inspection or by calculating several
statistics about a star. In particular, when a star is in focus, it will get more of its light on
a central CCD pixel than when out of focus. The Fine Focus tool offers an excellent
focus aid that will help you achieve critical focus.
In addition to these techniques, there are a few others you can try. One technique is to
build or buy a Hartman Mask, a diffraction mask, or a Bahtinov Mask. They're not
81
tough to build - many consist of cutouts in pieces of cardboard and one is assembled
out of TinkerToys (no, really). All work by having you place something in front of the
scope during focusing. When the star is nice and sharp, the artifacts induced by each
disappear or form a particular pattern. The Bahtinov mask is an exceptionally easy to
use version and is very effective.
A second technique to try is to use the fact that in focus stars get more of their light on
the CCD than out of focus stars. When in focus, you'll be able to see stars in the
Preview or even Frame and Focus that would disappear when out of focus. Adjust the
exposure duration or gain until you can just barely see a star. Adjust the focus to see if
you can make it brighter or if it disappears on either side of where you are right now (or,
if you know you're a bit out, make the star disappear with the duration or gain and
reappear with the focus knob).
18.4 Exposure sefngs
When taking images, there are a few simple rules to follow that will let you collect
frames that can be used to make a nice final picture.
18.4.1 Rule #1: Use the Histogram to keep your background above the floor and bright bits below the ceiling.
First, you should always try to expose images so that the background sky is "off the
floor" and the stars (or at least the cores of the DSOs) are "off the ceiling". What this
means is that you don't want large parts of your image to have values of zero or of
65535 (the minimum and maximum possible values). Any time a pixel has either of
these values, we've lost information. For example, let's say a star is at 65535 and one
next to it is really twice as bright. Both get recorded at 65535 and the final image
doesn't show a difference between the two. Once we've reached this maximum, we
simply can't go any higher and so important details (such as the difference between
these stars) are lost.
The same holds true on the dim end. Let's say a faint arm of a galaxy is just barely
brighter than the skyglow around it (a very common situation). If your background sky
is recorded as zero, quite possibly the faint bit of the galaxy is at zero as well. No
matter how many images you stack, if they all have zero in them, you'll never be able to
find that dim galaxy arm in your image.
How do you do this? The exposure duration is the most obvious method. Longer
exposures will brighten the image (moving the histogram to the right). In addition, the
increasing the gain and offset controls will also brighten the image. Both will add more
noise into the image, but a little bit more noise is a lot better to have than ultra-black
backgrounds. If you're running unguided images, you'll likely use higher values of gain
and offset than those running guided.
82
18.4.2 Rule #2: Take lots of images
Every image you take has noise in it. So, adding images together adds noise into the
image, right? Yes and no. If you compare a stack of 20 exposures of 30-second each
to one exposure of 10 minutes, the single longer exposure will quite probably be a
cleaner looking shot. But, if you compare one of the 30-second long images to the
combination of all 20 exposures, the combination or "stack" of images will have a lot
less noise in it than the single frame.
Why is this? Much of the noise in our images is uncorrelated or white noise. What this
means is that each time we sample something (e.g., each time we take an image), we
get some noise added into the image that has nothing to do with the amount of noise
added in the last time we took the image. (Hot pixels and readout noise are examples
of correlated noise and are addressed in dark frames and bias frames respectively).
When we combine multiple images, this uncorrelated noise starts to disappear. Four
30-second exposures will have half the noise of a single 30-second exposure (noise
follows a 1/sqrt(N) function where N is the number of images you combine). One
hundred such frames will have one tenth as much noise (and therefore 10x the SNR).
Reducing the noise allows one to "stretch" the image to make a very fine distinction
between dim portions of a DSO and the skyglow that is just the tiniest bit darker than
the DSO (this will always be the case - its just a question of how small that difference
is.)
18.4.3 Rule #3: Don't over-­‐tax your mount
If your mount can only take exposures of 30-seconds before showing tracking errors
on most exposures, don't try going any longer than 30-seconds until you can guide
your mount (PHD Guiding from Stark Labs is free and tries to make this as painless as
possible). Take Rule #1 and Rule #2 to heart and gather many noisier shots. Each one
may look pretty bad and it may look like you'll never get a good image out of your
efforts. Don't despair. I've had many nice shots come out of raw frames that look like
noise with barely a hint of any DSO in there much less a nice smooth one.
18.4.4 What do gain and offset do? After coming off the CCD and before hitting the actual analog-digital-converter (ADC)
to turn the signal into a number, there is typically a small pre-amplifier (this may be
inside the ADC chip itself). What this preamp does is allow you to boost the signal by
some variable amount and to shift the signal up by some variable amount. The
boosting is called gain and the shift is called offset.
So, let's say that you have pixels that would correspond to 0.1, 0.2, 1.1, and 1.0 ADU
were the ADC able to deal with fractional numbers. Now, given that it's not, this would
turn into 0, 0, 1, and 1 ADU. Two bad things have happened. First, the 0.1 and 0.2 have
become the same number and the 1.1 and 1.0 have become the same number. We've
distorted the truth and failed to accurately represent subtle changes in intensity. This
failure is called quantization error. Second, the first two have become 0 and, as noted
83
above, 0 is an evil black hole of information.
Well, what if we scaled these up by 10x before converting them into numbers (i.e., we
introduce some gain)? We'd get 1, 2, 11, and 10. Hey, now we're getting somewhere!
With gain alone, we've actually fixed both problems. In reality, the situation is often
different and the ADC's threshold for moving from 0 to 1 might be high enough so that
it takes a good number of electrons to move from 0 to 1. This is where injecting an
offset (a DC voltage) into the signal comes in to make sure that all signals you could
possibly have coming off the CCD turn into a number other than zero.
18.4.5 Gain's downside: Bit depth and dynamic range From the above example, it would seem like we should all run with lots of gain. The
more the better! Heck, it makes the picture brighter too! I often get questions about this
with the assumption that gain is making the camera more sensitive. It's not. Gain does
not make your camera more sensitive. It boosts the noise as well as the signal and
does not help the signal to noise ratio (SNR) in and of itself. Gain trades off dynamic
range and quantization error.
We saw above how it reduces quantization error. By boosting the signal we can have
fractional differences become whole-number differences. What's this about dynamic
range?
Let's come up with another example. Let's have one camera with a gain of 1. So, 1 e-/
ADU. Let's have another run at 0.5 e-/ADU. Now, let's have a pixel with 1k e-, another
with 10k e-, another at 30k e-, and another at 50k e-. In our 1 e-/ADU cam, we of
course have intensities of 1000, 10000, 30000, and 50000. In our 0.5 e-/ADU cam, we
have intensities of 2000, 20000, 60000, and 65535. What? Why not 100000? Well, our
16-bit camera has a fixed limit of 65535. Anything above that gets clipped off. So while
the 1 e-/ADU camera can faithfully preserve this whole range, the 0.5 e-/ADU camera
can't. Its dynamic range is limited now.
18.4.6 How do manufacturers determine gain and offset for cameras that don't allow the user to adjust them? Let's pretend we're making a real-world camera now and put in some real numbers
and see how these play out. Let's look at a Kodak KAI-2020 sensor, for example. The
chip has a well-depth specified at 45k e-. So, if we want to stick 45,000 intensity
values into a range of 0-65,535, one easy way to do it is to set the gain at 45,000 /
65535 or at 0.69 e-/ADU. Guess what the SBIG ST-2000 (which uses this chip) has the
gain fixed at... 0.6 e-/ADU. How about the QSI 520ci? 0.8 e-/ADU. As 45k e- is a target
value with actual chips varying a bit, the two makers have chosen to set things up a bit
differently to deal with this variation (SBIG's will clip the top end off as it's going nonlinear a bit more readily), but both are in the same range and both fix the value.
Why? There's no real point in letting users adjust this. Let's say we let users control the
gain and they set it to 5 e-/ADU. Well, with 45k e- for a maximum electron count at 5
84
e-/ADU, we end up with a max of 9,000 ADU and we induce strong quantization error.
10, 11, 12, 13 and 14 e- would all become the same value of 2 ADU in the image,
loosing the detail you so desperately want. What if the user set it the other way to 0.1
e-/ADU? Well, you'd turn those electron counts into 100, 110, 120, 130, and 140 ADU
and wonder just what's the point of skipping 10 ADU per electron. You'd also make
6553 e- be the effective full-well capacity of the chip. So, 6535:1 would be the
maximum dynamic range rather than 45000:1. Oops. That nice detail in the core of the
galaxy will have been blown out and saturated. You could have kept it preserved and
not lost a darn thing (since each electron counts for > 1 ADU) if you'd left the gain at
~0.7 e-/ADU.
What about offset? Well, it's easy enough to figure out the minimum value a chip is
going to produce and add enough offset in the ADC process to keep it such that this is
never going to hit 0.
85
19 Menu Quick Reference
19.1 File Menu
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Open File: Loads any FITS (color or B&W, compressed or not, 8-64 bits,
integers, floating points, you name it), PNG, TIFF, JPG, BMP, or DSLR RAW
(CR2, CRW, NEF, etc.) file into memory and display. 8-bit/color files are
automatically stretched to full range.
Preview Files: Opens a dialog that lets you preview a set of files, deleting and
renaming them as desired. Useful for filtering images and for quick looks at files.
DSS Loader: Download an image from the Digitized Sky Survey and overlay an
FOV indicator to help you plan your shots.
FITS Header Tool: Lets you view the contents of the header of a FITS file.
Save current file (FITS): Saves the currently displayed image in FITS format
using 16-bit integers (0-65,535). Compression set by Preferences, Save as
compressed FITS.
Save BMP file as displayed: Saves the currently displayed image in Windows
BMP (bitmap) format. The values of the black and white sliders set the black and
white levels in this, since BMP format is only 8-bits / color. How it looks is how it
will save.
Save JPG file as displayed: Like Save BMP, but in JPEG format. Any JPEG
quality / compression (0-100) factor possible.
Save 16-bit/color TIFF: Saves the current image in TIFF format (lossless
compressed or uncompressed) at full 16-bit/color (aka 48-bit color) bit depth.
This preserves all information in your image for use in graphics programs
Save 16-bit/color PNG: Saves the current image in PNG format (always lossless
compression) at full 16-bit/color (aka 48-bit color) bit depth. This preserves all
information in your image for use in graphics programs
Save 16-bit/color PPM/PGM/PNM: Saves the current image in the appropriate
variant of these “portable pixel map” UNIX-based standard formats.
Save Color Components: Saves the current color frame as three separate FITS
files corresponding to the the red, green, and blue components of the image.
Launch new instance: Launches a separate running copy of Nebulosity to let
you work on another image or control another camera.
Change language: Changes the language used in the user-interface. You must
restart Nebulosity after this.
Exit: Exit the program (on Macs, in the Nebulosity menu)
19.2 Edit Menu
•
•
•
Undo: Undo the last change to your image. Undo will let you step back from any
changes made by tools in the Image menu. By default, you can take 3 steps
back. You can opt to disable Undo in the Preferences menu (to run a bit faster)
or to have virtually unlimited undo capability.
Redo: Think you liked it better with that processing you just undid? Redo.
Image Info: Shows information about the current image including its size and the
various capture parameters that either were stored in the FITS header or will be
86
•
•
•
•
•
•
stored when the image is saved.
Measure Distance: Measure the distance in CCD pixels, arc-seconds, and arcminutes among up to 3 points in the image (right-click to set points first).
Edit / Create Script: Open a window that allows you to create a capture script
and load / edit an existing script.
Run Script: Run a capture script, automating the image capture process
De-select cameras: Camera pull-down getting a bit cumbersome? Use this to
remove entries from the list (you can always add them back in later).
Check / Update License: Verify your current license code and status and/or
update your license code.
Preferences: Set various preferences. See Preferences. (On Macs, in the
Nebulosity menu)
19.3 Batch Menu (see also the Pre-­‐Processing secAon)
•
•
•
•
•
•
•
•
•
•
•
Pre-process image sets: Apply traditional dark frame or bad pixel maps, flat
frame, and bias frame corrections to correct for typical CCD artifacts. Apply
these corrections to a series of images.
Make Bad Pixel Map: Create a map of the bad pixels on your CCD.
Batch Demosaic + Square RAW Color and Batch Square BW: Batch versions of
the tools found in the Image menu.
Grade Image Quality: Grade a set of images to determine the sharpest (and
fuzziest) of the set.
Normalize Intensities: Normalize all images in a set to remove offset and scaling
differences.
Match Histograms: Equate a set of images’ histograms to match that of a target
image.
Align and Combine: Align and (optionally) combine a series of images. A dialog
will appear to let you control the method. Methods include: Fixed (no alignment),
Translation ("one star", full-pixel shifts), Translation + Rotation (subpixel,
including rotation such as with an alt-az mount), Translation + Rotation + Scaling
(same, but including a scaling term), Drizzle, and Colors in Motion. For Fixed
alignment, Standard Deviation based stacking is an option.
Automatic Alignment (non-stellar): Automatically align frames without picking
reference stars.
Batch Geometry: Batch versions of rotation, binning, resampling and cropping.
Batch Conversion: Tools to convert a set of images from FITS to various
graphics formats or vice versa.
Batch One-shot Color with Line Filters: Batch versions of the tools in the Image
menu that extract portions of the color filter array.
19.4 Image Menu (see also the Image Adjustment secAon)
•
•
De-Mosaic RAW + Square: Convert a single RAW CCD image currently
displayed from a one-shot color camera into a full-color image. Faster and
better quality modes available.
Square B&W pixels: Squares pixels from black and white images.
87
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
One-shot color with line filters: Tools for reconstructing a RAW image taken with
line filters (e.g., Ha, Hb, OIII) from a one-shot color camera are provided along
with a special Low Noise 2x2 bin optimized for these cameras.
Crop: Resize the image by removing or trimming unwanted edges.
Mirror/Rotate Image: Tools are provided for 90 and 180 degree rotation and for
mirroring an image horizontally or vertically.
Resize Image: Resample the image to change its size using any one of 6
different resampling algorithms (Box, Bilinear, B-Spline, Bicubic, Mitchell,
Catmull-Rom)
Bin Image: Shrink an image by binning (combining 2x2 patches of an image into
individual pixels). Perform 2x2 binning using simple summation, simple
averaging, or an adaptive algorithm. These reduce your image size by 2x.
Levels / Power Stretch: Apply a user-controlled stretch routine to the current
image. You can use this much in the same way a Levels tool is used to bring out
details in the image.
Digital Development: Apply a user-controlled stretch routine to the current image
designed to make CCD images look more like film images. An excellent way to
bring out faint detail in your images.
Curves: Create a curve to transform the intensity of your image. Very powerful
stretching tool.
Zero Min: Add or subtract a constant from the current image so that its minimum
will be zero.
Scale Intensity: Pixel math to add, subtract, multiply, etc. each pixel.
Adjust Color Background (Offset): Subtract user-specified values from the red,
green, and blue color channels (e.g., from skyglow) to balance the color of the
background in the image.
Adjust Color Scaling: Apply a user-controlled scaling to the red, green, and blue
color channels separately to help balance the image.
Auto Color Balance: Automatically balance the color (both offset and scaling) to
remove a color-cast.
Adjust Hue / Saturation: Tool to adjust the hue, saturation, and luminance of the
image.
Discard Color: Remove all color information from an image (extract the
luminance data).
Convert to Color: Change a monochrome image into a color-format image.
LRGB Color Synthesis: Create a color image from separate files using RGB,
traditional HSI-based LRGB, or Color Ratio based LRGB
Gaussian Blur Image: Blur (smooth) your image by using a Gaussian kernel.
Sharpen Image: Four tools are provided. Traditional and Laplacian sharpen tools
based on 3x3 kernels are provided along with an Unsharp mask and the Tighten
Star Edge tool. This applies an edge-detection routine (not a typical "sharpen"
or "unsharp mask") to tighten stars and enhance edges in your image.
Vertical smoothing (deinterlace): Smooth the image vertically to remove effects
from interlaced sensors.
Adaptive median noise reduction: Blend a median-based denoised image with
your original image to remove noise in the background
88
•
GREYCstoration noise reduction: Use the powerful tool from GRYEC Labs to
reduce noise while preserving details and important features in your image.
19.5 View Menu
Toggle the visibility of any of the following tools
• Main image: Your actual image. Why are you turning it off?
• Display Control: The B and W sliders and historam
• Capture Control: The capture control panel that lets you connect to the camera
and control exposures
• Notes: A text box that lets you jot down notes and save them to a basic text file
• History: A text box that keeps track of your processing history (what you’ve
done at each step to your image). You can save this for future use or reflection
upon in your golden years.
• Macro: Image processing steps that were logged in the History window can be
replayed to the same or a different image using this tool. Copy the items from
the History window, put them in the order you like, and hit Run.
• SBIG Control: A small dialog that lets you control the TEC, filter wheel, shutter
state, etc. for your SBIG camera
• QSI Control: A small dialog that lets you control the TEC, filter wheel, shutter
state, etc. for your QSI camera.
• Mini Capture: A shrunken version of the main Capture Control panel. This small
version lets you keep more panels active while capturing images.
• PHD Link: Control the link to PHD Guiding
• Ext. Filter Wheel: Control an external ASCOM-compliant (Windows) or Starlight
Xpress (Mac) filter wheel.
• Pixel Stats: Opens a pop-up window that shows the intensity (I), red (R), green
(G), and blue (B) values under the current pixel, the min, mean, and max in a box
21x21 pixels big (+/- 10 pixels) around the current pixel, and the min, mean, and
max of the entire image. If there is a star near the cursor, it will also report the
HFR of the star (how wide it is) and any shift between the peak value of the star
and the center based on the HFR. You can keep this dialog up as long as you
like and continue to work in Nebulosity. As the mouse moves around or as new
images are acquired, the window will update itself.
• Reset View: You GUI look like a tornado hit it? Use this to reset to the default
layout.
19.6 Help Menu
•
•
•
About: Display program and version information
Show Help: Display this manual
Check web for updates: Connect to the Stark Labs website and check for
updates. If an update is available, the Release Notes will be shown and you will
be given the opportunity to visit the Stark Labs website to download the new
version.
89
20 Preferences
20.1 Capture
•
•
•
•
•
•
•
•
•
•
•
DSLR Long Exposure Adapter: Without a "bulb" adapter cable ("USB only, 30s
max"), DIGIC II cameras will be limited to 30 second exposures. Here, select
which long-exposure adapter you have. Please make this selection before
connecting to the camera.
Mirror lockup delay (ms): Number of msec to wait after the mirror-lockup signal
has been sent on DSLRs before starting the exposure.
DSLR Save location: Should the images be downloaded to the computer, saved
on the compact flash card, or both? Should CR2 files be saved locally as well
as FITS?
DSLR LiveView use: Should LiveView be used in Fine Focus, Frame and Focus,
both, or never?
Acquisition mode: When taking images with a one-shot color camera, what
should be done about converting them to full-color?
o RAW CCD data: Do no reconstruction and keep the data as RAW CCD
data. When saved, one FITS file with the raw data from the CCD
(effectively a black and white image that contains the color information)
will be saved. You will likely want to De-Mosaic the image prior to
alignment and stacking or use Colors in Motion.
o RGB Optimize speed: Do color reconstruction on the fly during image
acquisition and try to go for the fastest good color reconstruction at the
expense of a bit of quality.
o RGB Optimize quality: Do color reconstruction on the fly during image
acquisition and try to go for the highest quality color reconstruction at the
expense of a bit of speed.
Capture alert sound: Give an audible alert at the end of each image or the end or
the entire series.
Use max binning in Frame & Focus: Use the highest binning level or restrict to
2x2 binning for Frame and Focus.
Show crosshairs in Frame & Focus: Enable or disable the crosshairs (disable is
useful for a real-time video view).
Save Fine Focus info: Enable this and during fine focusing, an image will be
created of the full fine focus info. The file is in the current save directory and has
the name "Focus_#_#_#_#.bmp". The #'s are CurrentMax, Historical max of
CurrentMax, CurrentHDR, Historical max of CurrentHDR. By doing it this way,
all one needs to do is to scan the directory for a file that matches this pattern
and look at that filename to get the info. Of course, there can be a short period
of time in which no file exists (last one erased before new one written).
Enable Big Status Display during capture: During series captures, the progress
will be displayed in a pop-up dialog for easy viewing if you've left the computer
unattended.
TEC / CCD Temperature set point: For cameras that can regulate the cooling,
this is the desired temperature (degrees Centigrade).
90
20.2 Output
•
Save as compressed FITS: FITS files are saved in lossless compressed FITS
format to save space with no loss of data integrity (default). Note, however, that
some applications do not support this aspect of the FITS standard.
• Save in 32-bit floating point: FITS files are saved in the 32-bit floating point
format used internally to ensure no possible loss of data resolution at a cost of
files being twice as large
• Use 15-bits (0-32767) instead of 16-bits: FITS files are saved in data ranging
from 0-32767 rather than 0-65,535 if this is selected. Some programs (e.g., Iris)
require this format.
• Color file format: When saving full-color data from a one-shot color CCD camera
(e.g., the SAC-10), this preference controls how the color data are to be saved.
o RGB FITS - ImagesPlus: One FITS file with red, green, and blue
components of a reconstructed (de-mosaic'ed) full-color image stored
inside in the style expected by ImagesPlus (separate "HDU" per color)
(default).
o RGB FITS - Maxim / AstroArt: One FITS file with red, green, and blue
components of a reconstructed (de-mosaic'ed) full-color image stored
inside in the style expected by Maxim DL and AstroArt (a "3-axis" or "3D"
image with color along the third axis).
o 3 FITS files: Reconstruct the full color image and save the red, green, and
blue data in three separate files. This should only be used if Nebulosity is
not to be the primary pre-processing application and if the application to
be used does not support RGB FITS (e.g., Iris). Seriously - don’t use this
unless you really, really know what you’re doing and know you need this.
I estimate there is one person in the world that needs this as the default.
• Series naming convention: Choose to have images in a series named by a 3digit code or with a UTC date code (DDD_HHMMSS).
20.3 Processing
•
•
•
Undo / Redo settings: You can opt for either no undo capability (to run faster
and save hard disk space), 3 steps worth of undo (default), or virtually unlimited
undo capacity.
Use adaptive stacking: For the stacking techniques that you use on your light
frames (Translation, Drizzle, Colors in Motion), the image will automatically have
the intensity scaled to use the full range of the 16-bit file format used. Adding
images and averaging images each have their strengths and weaknesses. The
Adaptive stacking technique side-steps the weaknesses of each and lets you
get the most out of your data. The only downside is that a stack of 30s images
and a stack of 3m images would appear equally "bright" after stacking this way.
Flat processing: What should be done to your flats when you apply them by
default in the pre-processing dialog? You can choose to have nothing done or
to have several filters applied. If you’re using a one-shot color camera, you’ll at
least want to apply the 2x2 mean filter (this removes the Bayer matrix).
91
•
•
•
Demosaic (debayer) method: What algorithm should be used for converting oneshot color images to color? VNG (Variable Number of Gradients by Ting Chen)
is the default and an excellent choice. PPG (Patterned Pixel Grouping by Alain
Desboilles) and AHD (Adaptive Homogeneity-Directed by Keigo Hirakawa,
Thomas Parks, and Paul Lee) are also popular advanced debayers. Bilinear is
quite fast if a bit soft. Color binning will reduce your image size by a factor of 2
but does no interpolation.
Manually override color reconstruction: Typically, Nebulosity will attempt to
determine what kind of camera a one-shot color file comes from and set the
various demosaic options automatically. At times, you may wish to override this
automatic behavior and specify offsets, array types, color mixing, etc. manually.
Enabling this preference will bring the manual color reconstruction dialog up
each time so that you can override any automatic behavior.
DSLR White Balance / IR filter: Ideally, the pixels are white balanced prior to
actually implementing the demosaic of a RAW image. For most cameras, this
white balance is known a priori, but DSLRs can be stock or modified. Choose
the setting here that best corresponds to your camera setup. Note, that at times,
if there are severely saturated areas, this may lead to a pink area in the saturated
zones. If this occurs, the Straight Color Scale option can be used.
20.4 Colors
If you want to customize the interface’s coloring, you’ve got a few options here. Have
fun!
20.5 Misc
•
•
•
Display orientation: Should Nebulosity flip or rotate the display?
Clock / TEC display: In the control panel, Nebulosity can display a small clock
that will show the current time in a range of time formats or show the CCD's
current temperature. The time formats all use your computer's internal clock as
the starting point and convert that into other times. Note that local sidereal time
and Polaris RA depend on Nebulosity knowing your longitude.
o No clock: Hide the clock
o Local time: The current local time
o UT/GMT time: The current Universal Time (or Greenwich Mean Time)
o GMT Sidereal: GMST or Greenwich Mean Sidereal Time
o Local sidereal: The current local sidereal time (useful in finding objects
with setting circles)
o Polaris RA: Polaris' current right ascension (useful in using polar
alignment scopes)
o CCD Temperature: Current temperature of the CCD in centigrade.
Longitude: Local sidereal and Polaris RA require knowing your current longitude.
Enter it in decimal notation (e.g., -77.1 not H:M:S) with west (e.g., USA locations)
being negative.
92
21 Scripts
Nebulosity provides you with the ability to automate your capture process by using
scripts. Scripts are simple text files that list a series of commands for Nebulosity to
perform in sequence. For example, the script shown here would set the output
directory to be \ccd\Oct22_11 on your "C" drive (usually the letter associated with your
hard drive). If the directory didn't exist, Nebulosity would attempt to create it. It would
set the output file name to be "m27", the duration to be 2 minutes and then capture 10
images in a series (m27_1.fit, m27_2.fit, etc). It would then pause and alert the user to
"Setup for darks" (i.e., place the lenscap over the telescope). After the user hits OK, it
would then capture 10 dark frames (dark_1.fit, etc.)
Nebulosity's scripts can be created dynamically using the clipboard's operating
system. If commands are placed on the clipboard and Nebulosity is in a special
"Listen" mode, it will suspend reading commands from the script file and instead read
them from the clipboard. In addition, with a “Listenport #” command, commands can
be sent via TCP/IP sockets. This allows other programs to dynamically control
Nebulosity's actions. Full list of Commands
21.1 Script Editor
Commands act just as if you were
You can write scripts in any text editor (save in
to do them in the GUI. So, if you've
"ASCII text" format) or in Nebulosity's built-in
already set something in the GUI
editor. Simply pull-down Create / Edit Script from
or if it is the default, there is no
the Edit menu. Here, you can start typing
need to enter it in the script.
commands, load an existing script, or, most
simply, use the list of commands on the left. The
left side of the dialog shows all the available script commands (the most common ones
are grouped at the top of the list and repeated below in the appropriate sections). Find
a command you want and double-click on it (or drag/drop it) and it’ll be added to the
script on the right. Don’t know what you want? On the right, you’ll also see what each
command does and what the parameter means. Here, for example, you see that the
Capture command takes a number value that is the # of images in the sequence.
93
When done, you'll likely want to save your script (Save button) and then press Done.
When you're ready to execute the script, either
press Run Script while the Script Editor is open or
Script files can contain extra spaces
simply pull down Run Script from the Edit menu.
or blank lines if you want to make
Nebulosity will then first verify that it's a valid
them look cleaner when writing
script. Then, it will go through line by line,
them. Nebulosity will simply skip
executing each command until it reaches the end
any extra spaces or lines it finds. In
of the file. As it does so, the Status bar will keep
addition, if you start a line with the
you apprised of what Nebulosity is currently
“#” symbol, the line will be ignored.
doing. Pressing the Abort button will cancel the
This is a useful way to put
script at any time.
comments in your scripts.
One thing of note - when you’re specifying times,
the units are seconds (in Nebulosity 1 and 2 times
were all in milliseconds). You can use fractions if you like (e.g., 10.5 would be ten and
a half seconds and 0.001 would be one millisecond). You can also put an “m” suffix on
the number to indicate minutes. So, if you wanted a minute and a half (90 seconds)
exposure you could use “90”, “90.0”, or “1.5m”.
21.2 Full Command List
The following is the list of commands recognized
by Nebulosity. They are presented here capitalized
to help show you the meaning of the command,
but Nebulosity ignores the case of all commands.
So, "SetName" is the same as "setname" and
"SeTNAme".
You can execute scripts at startup
by passing the script name as a
command-line argument. For
example "nebulosity script.neb" will
automatically execute script.neb
Each command must be placed on a separate line and each line must have a
command and a parameter with at least one space between the command and the
parameter. When the list shows the parameter to be N, it means a number should be
provided as the parameter. When the list shows the parameter to be S, it means a
string (aka text) should be provided as the parameter.
21.2.1 Capture Setup Commands
These commands mirror the controls present in the Control Panel and several of the
settings available in the Preferences menu.
• SetName S - Sets the base filename to be S
• SetDirectory S - Sets the capture directory to be S
• SetDuration N - Sets the exposure duration to N seconds (fractions allowed use “m” as a suffix to denote minutes)
• SetTimelapse N - Sets the time lapse to be N seconds (fractions allowed - use
“m” as a suffix to denote minutes)
• SetBinning N – Sets the bin mode. 0:No binning, 1 or 2:2x2, 3:3x3, 4:4x4
94
•
•
•
•
•
•
•
SetShutter N – Sets the shutter state. 0:Open (light frame), 1:Closed (dark
frame)
SetFilter N – Selects filter #N from your camera’s filter wheel if connected (1indexed)
SetExtFilter N – Selects filter #N on the external filter wheel if connected (1indexed)
SetGain N - Sets the camera gain to be N
SetOffset N - Sets the camera offset to be N
SetColorFormat N - Sets the color file format used when (and if) full-color
images are written. 0: RGB FITS in ImagesPlus format. 1: RGB FITS in Maxim /
AstroArt format. 2: 3 separate FITS files.
SetAcqMode N - Sets the color acquisition mode. 0: RAW or BW images. 1:
RGB Optimize speed. 2: RGB Optimize quality.
21.2.2 Control Commands
These commands control the capturing process itself and let you interact with the user.
• Capture N - Captures a series with N images according to the current settings.
• CaptureSingle S - This (e.g., “CaptureSingle foo”) will capture a single frame
and save it with the given name (e.g. “foo.fit”) even if it already exists. The file
will be in the current save directory. In addition, if the filename specified is
"metric" (or "metric.fit"), it will run the whole-image HFR routine used in quality
grading and the file will be "metric_###.fit". The ### is 100x the HFR. Note,
since Nebulosity doesn't know when this file has been used and when you're
done with it, it will only erase / write a new one if you run CaptureSingle again. If
you quit Nebulosity though, this file will still linger in the directory.
• Connect N - Connect to camera N where N is where you would find the camera
in the pull-down list in Nebulosity's main window (0=indexed). So, “No
camera” (aka disconnect) is “Connect 0”.
• ConnectName S - Connect to camera named S, where S is the name as-listed
in the pull-down list in the main window. For example, “Connect Meade DSI”
• PromptOK S - Displays S on the screen and prompts the user to hit OK or
Cancel. If OK is hit, the script continues. If Cancel is hit, the script is stopped.
• Delay N - Pause execution for N seconds (fractions allowed - use “m” as a
suffix to denote minutes)
• SetPHDDither N - Sets the dither level in the link to PHD guiding (0=none,
1=small, ... 5=extreme)
• SetBLevel N and SetWLevel N - Sets the B and W slider levels to N. If N = -1,
auto-scaling is turned on.
• Listen N - Enable (1) or disable (0) listening to commands from the clipboard.
Each command on the clipboard must start with "/NEB". For example, "/NEB
Listen 0" on the clipboard will return processing to the script file.
• Listenport N - Enable (port number) or disable (0) listening to commands via
sockets. For example, you might have a script with one line saying: Listenport
4301 and when run, Nebulosity would begin waiting for a connection on port
95
4301. •
o Nebulosity first waits for the connection to be formed (if you wish to
bail, the Capture Abort will break out) and after that, it enters into a
loop. It will wait for any data to come in using "blocking sockets". There is a timeout of 1s per read though, giving you the ability to
break out. Many commands can be sent at once and if there are
commands still to be processed, Nebulosity will do them without
waiting (currently, a 2kb buffer is used for each read). Thus, it's not
like you can only do one command per second. You can only abort
once a second.
o To leave the mode and return control to Nebulosity itself either a)
break the connection (Nebulosity will pop up a "connection lost"
dialog) or b) send the command "Listenport 0".
o During all this, a small dialog will pop up that will show what
Nebulosity is getting over the port, the status of the connection, etc. -all useful info for debugging. At the moment, said dialog is not
functioning on the Mac but you will see things flash by on the status
bar.
Exit N - Wait N milliseconds and then exit the program
21.2.3 Advanced Camera Control Commands
These commands have the same function as the Advanced camera dialog box,
allowing you to override the current settings shown in the dialog. If the camera is not
capable of the command given in the script, the command is ignored.
• SetAmpControl N - 1: Amplifier control is enabled and the CCD amp is off
during exposures. 0: Amplifier control is disabled and the CCD amp is on during
exposures.
• SetHighSpeed N - 1: Enable high speed readout mode. 0: Disable high speed
readout mode.
• SetOversample N - 1: Enable CCD oversampling. 0: Disable CCD
oversampling.
• SetDoubleRead N - 1: Enable double-read mode. 0: Disable double-read mode
96
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