Digital imaging guidelines (UPDIG)

Image Receiver Guidelines Table of Contents
Acknowledgments
6. Sharpening
Stock image distributors
Magazine publishers
Publication designers
Web designers
Forword
Quick Guide
1. Getting on the Same Page with Color
Displaying Color Correctly
2. Getting on the Same Page with File
Size and Resolution
File Dimensions
Digital Capture Quality Parameters:
Sensor Size and Image Dimensions
7. Other File Quality Parameters
Aggressive pixel editing
Scanning
Bad processing examples
Limiting output levels for print
8. File Naming
Stock image distributors
Magazine publishers
Publication and web designers
Changing file names
3. Delivery File Formats
TIFF Files
JPEG Files
raw Files
DNG Files
9. Metadata
Overview
Types of Metadata
Magazines and publishers
Publication Designers
Web designers
Photographers
METAmachine
4. File Delivery Options
About these digital image files
CD-R or DVD-R
Hard drives
Electronic delivery
Electronic Delivery Methods
5. ICC Profiles
10. Fine Art Reproduction – What
Museums Need
File Delivery Color Profile Options
What flavor of RGB?
Maybe CMYK?
Image Workflow
File Delivery
File Request Template
Appendix
Page 1 Acknowledgments
Managing Editor: Richard Anderson
Technical Editor:
Michael Stewart
Contributing Editor: David Riecks
Contributing Editor: Peter Krogh
Copy Editor:
Greg Smith
UPDIG Contributors:
Andre Cornellier
Bob Croxford
Graeme Cookson
Kevin DePalmer
Dennis Dunbar
Robert Edwards
Ken Fleisher
George Fulton
Judy Herrmann
Reed Hoffmann
Bob Marchant
Rick McCleary
Sam Merrell
Alan Newman
Michelle Alvarado Novak
Betsy Reid
Stanley Rowin
Sarah Saunders
Jeff Sedlik
Ulrik Södergren
Tamra Stallings
Eddie Tapp
Mike Upstone
Page 2 UPDIG Coalition Member Groups
Advertising Photographers of America (APA)
Advertising and Illustrative Photographers Association (AIPA)
Australian Institute of Professional Photography (AIPP)
American Society of Media Photographers (ASMP)
American Society of Picture Professionals (ASPP)
Associação Brasileira dos Fotógrafos de Publicidade (ABRAFOTO)
Association of Photographers [UK] (AOP)
Australian Commercial + Media Photographers (ACMP)
Australian Institute of Professional Photographers (AIPP)
Canadian Association of Photographers and Illustrators in Communications (CAPIC)
Editorial Photographers (EP)
Hong Kong Institute of Professional Photographers (HKIPP)
International Digital Enterprise Alliance (IDEA)
Museum Computer Network (MCN)
National Association of Professional Photographers of Italy - TAU Visual (FOTOGRAFI)
National Press Photographers Association (NPPA)
Picture Archive Council of America (PACA)
PLUS (PLUS)
Professional Photo Companies and Photo Agencies Association [Sweden] (BLF)
Professional Photographers of America (PPA)
Stock Artists Alliance (SAA)
Page 3 Foreword
Vision Statement: The photography industry is undergoing great change. New ways of making
and working with photographs are replacing older methods. The broad capabilities and
efficiencies of digital capture and delivery have revolutionized the image marketplace. But in
doing so they have exposed gaps in creative and quality control that frequently lead to confusion,
inefficiencies, loss of quality and unnecessary expense. We recognize the need for worldwide
imaging standards to facilitate reliable and repeatable image reproduction.
Film-based photography workflows relied on transparencies or prints to establish and
share color and tonal information. Digital imaging lacks such tangible references. Integrated
color management with accepted ICC workflows fills that void. After more than a decade of
regular use and adjustment, ICC (also known as ICM) has proved itself a mature technology used
by thousands of professionals. As the first link in the image supply chain, photographers must
establish and adhere to the best practices of color management. And those who work with their
photographs must follow the same standards in order to accurately reproduce those photographs.
The media for delivery and storage of images have also changed. Digital Asset
Management offers a set of standards to replace film-based workflows. Our task is to identify
terminology and workflow for image creation, image delivery, accurate reproduction and
archiving. Implementing the standards and practices in these guidelines will enhance creative
control and economic efficiency.
Mission Statement: The UPDIG Coalition is a global initiative of industry organizations working to
identify, define and share best technical practices for reproducing, exchanging and archiving digital
images.
The UPDIG Coalition brings together industry experts representing creators, distributors, users,
technology providers and others dedicated to this mission.
The UPDIG Coalition collectively publishes and promotes the Universal Photographic Digital
Imaging Guidelines.
Objectives:
1. Make digital imaging practices clearer and more reliable by developing standards for quality and
terminology.
2. Develop web resources for imaging professionals that establish common digital imaging terms.
3. Demonstrate the creative and economic benefits of these guidelines.
4. Develop industry guidelines and suggested workflows for various types of image reproduction.
5. Establish consensus on these issues among professional photography trade organizations, image
distributors, archives, printers, software developers and manufacturers.
Standardization Goals:
1. Color Management: ICC Workflow, Hardware Calibration, Guide Prints and Proofs
2. File Delivery and Archiving: Naming Conventions, Resolution, Sharpening, Digital Workflow
3. Metadata: embedded IPTC and XMP data, including PLUS Licensing Language
4. Industry Terminology
Page 4 About the Guidelines for Image Receivers
The purpose of this document is to help you create Submission Guidelines that
effectively communicate your requirements to your image suppliers.
The Goal
The UPDIG Image Receiver Guidelines recommend standards to improve the “hand-off” of
digital image files from photographers to end users of all types. This diverse community includes
stock image distributors, magazine and book publishers, publication designers, web designers,
art directors, museums, printers, fine-art publishers and more.
UPDIG’s Digital Imaging Submissions Guidelines working group reviewed the submission
practices of various end-user communities, identified best practices and standards that meet their
needs, and produced this document to help users develop their own Submission Guidelines.
Members of the DISG working group represent both digital image suppliers and user
communities, including magazine publishers, stock image distributors, graphic and web
designers, museums, and others.
The Problem
While end users usually have specific criteria for image submission, they often lack clear,
consistent terminology for communicating those guidelines. As a result, photographers submit
digital images in a variety of image formats, with various resolutions, camera types, color
profiles, metadata schemas, sharpening and tonal correction.
That is a problem, not just for the end users, but also for the photographers. It inhibits creativity
and results in inefficiencies for all who work with digital images.
Recommendation
All digital image users should consistently present a Submission Guidelines “checklist” when
requesting digital image files. Base such checklists on best practices outlined in the UPDIG
Universal Digital Imaging Guidelines and summarized in this report.
Page 5 Image Receivers Quick Guide
Before developing your Submissions Guidelines, carefully evaluate your needs and decide on a
consistent policy for instructing photographers how to prepare digital files.
The following are key elements to address. After choosing the options that fit your business, you
can use our Guideline Generator to prepare a document that contains your own custom Digital
Image Submission Guidelines.
Main Principles
!
Digital images should look the same as they transfer between devices, platforms and vendors. !
Digital images should be prepared in the correct resolution, size and sharpness for the device(s) on which they will be viewed or printed. !
Digital images should have embedded metadata that conform to the IPTC and PLUS standards, making them searchable while providing relevant rights and usage information – including creator’s name, contact information and a description of licensed uses. !
Digital images should be protected from accidental erasure or corruption and stored carefully to ensure their availability to future generations. 1. Color Management
Both suppliers and receivers of a digital image file need to see the same image. This requires
setting systems to display standardized, consistent color, and embedding in images color profiles
that describe their color characteristics.
2. File Size and Resolution
File size depends on intended use. As a rule, stock image distributors require the largest file
sizes, since final use is unknown. Magazines can be more specific, although they may request
Double Page (11 x 17 inches, A3 or 420mm x 297mm) for maximum flexibility in placement
and cropping. Graphic designers and advertising art directors should be very specific about file
sizes, since it’s best for the photographer to deliver files within 10 percent of final size – unless
the designer will handle resizing and output sharpening. Designers who do resize images should
start with a larger size and reduce it, rather than attempting to increase the resolution of smaller
image files. Web use and digital projection require much smaller file sizes. Widescreen HDTV
requires 1920 x 1080 pixels for uncropped images.
Page 6 3. File Delivery Formats
TIFF and JPEG (quality 10 or higher in the Photoshop “Save As” dialog) are usually the
preferred file formats for delivery. Be aware that JPEGS saved at anything less than level 12
quality may have compression artifacts that could show up under certain printing conditions,
such as the use of stochastic screens. Both are standard formats, universally supported by imageediting and image-cataloging software. And both can safely hold metadata. Be careful about
requesting or submitting raw file formats, including DNG files. While they allow a wide range of
adjustments, this can lead to final color and tone that differ considerably from what the
photographer intended – or the client expects. complete guideline
4. File Delivery Options
Regardless of format, file extensions are essential. If receiving images on CD-R or DVD-R discs,
request them recorded in ISO 9660, cross-platform format. A portable hard drive may be the best
tool for delivering large numbers of files, but beware of compatibility issues between Macintosh
and Windows file systems. Online delivery via email, FTP or a commercial web delivery service
offers an increasingly attractive and viable option. Compressing folders of files in digital
wrappers, such as ZIP or SIT archives, helps avoid file corruption. UPDIG recommends
requesting “ReadMe” files on all delivery media, listing file characteristics, any special concerns
and usage rights. complete guideline
5. ICC Profiles
There are two primary color modes in use : RGB and CMYK. RGB (Red, Green, Blue) is the
standard archival and delivery color space, and holds more color information than CMYK (Cyan,
Magenta, Yellow, Black) which is used for print output.
The conversion of an RGB file to CMYK involves a loss of color information tat is not
reversible, so in general it is best to retain images in RGB as far as possible in the workflow.
CMYK’s image files with four channels are 25 percent larger than the equivalent RGB with
three channels. Adobe RGB (1998) is the most common delivery standard, with ECI RGB being
adopted by some publishers in Europe. Unfortunately, the sRGB color space is probably used
more often than either, although its smaller color gamut clips some colors that fall within the
gamut of an offset press rendering those otherwise printable colors un-printable.
Color profiles specify which color space the image is in, and should always be embedded in the
image when saving, and preserved when opening, for correct viewing and handling. It is
impossible to interpret the color numbers in an image correctly and with certainty if the color
space is not known. complete guideline
6. Sharpening
After resolution and color, sharpening can have the most impact on perceived image quality.
Most digital cameras deliver slightly soft images (unless the capture format is set to JPEG or
TIFF with in-camera sharpening turned up). This results from the almost universal use of antialiasing filters to combat moiré. Scanning camera backs and some medium-format backs –
Page 7 especially those with multi-shot mode may not have this feature. If we assume single-shot raw
file capture, it is entirely appropriate to request a small amount of default “capture” sharpening.
Without it, digital image files may look too soft.
How and when to add process sharpening and sharpening for output requires clear
communication between those requesting digital files, and those creating and delivering digital
image files.
7. File Quality
All images, whether originating from scanner or from digital camera files, need to be processed,
but unskilled processing and correction, as well as over-sharpening and over-compression, lead
to degradation of image quality. High quality scanners should be used, but image quality will
depend as much on scanning and image processing skills as on the technology employed.
Image submissions guidelines are always important, and photographers should work to generally
accepted standards if their customer does not supply their own guidelines (link to Digital Image
Spec) along with proper quality control. Guides to common image faults may also be useful for
suppliers.
8. File Naming
To avoid problems with files transferred across computing platforms, request file names that use
only the letters of the Latin alphabet (A-Z, a-z), the numerals 0 through 9, hyphens and
underscores. Avoid other punctuation marks, accented vowels, non-Latin letters or other nonstandard characters, such as \:/*<> or brackets. On a local network or with rewritable media,
limit the file name to 31 characters or less (including the three-letter file extension. Limit file
names to 11 characters or less (including the three-letter file extension) when burning to optical
media, since some computers may not support long file names. Use a single period (“.”) between
the file name and extension. Specify unique file names. Multiple files with the same name will
cause problems for computers and people alike, and a newer file might automatically overwrite
an older with the same name, or vice-versa. You may want to specify including the numeric date
and/or the name of the photographer as part of the file name as a way to avoid duplicate names.
For a complete guide to file-naming protocol, see the Controlled Vocabulary website.
Stock image distributors have the greatest need to implement their own file naming conventions,
followed by magazines and publishers.
For image-file tracking, it’s a good idea to ask image creators to include their original file names
in Document Title fields (of images’ metadata), since many end users change the names
associated with files.
9. Metadata
Digital image metadata is information about the image. Information like picture number, caption,
credit and copyright that used to be carried on the hard copy, should now be found within the
digital file so that images are captioned and can be traced back to their supplier and copyright
holder.
Page 8 Metadata come in many forms and schemas, from EXIF, to ICC profiles, IPTC, XMP and PLUS.
Metadata needs vary for stock image distributors, magazines, publishers, print and web
designers, and photographers maintaining their own libraries. When requesting or delivering
digital image files, understand and use metadata to streamline the production workflow. It’s
increasingly common for stock-image distributors and magazine publishers to develop custom
XMP metadata panels that conform to their particular needs.complete guideline
10. Fine Art Reproduction
Creating and preparing digital reproductions of fine artworks requires all the careful attention to
details already presented, but it also poses significant additional challenges. Unlike typical photo
reproduction, the colors within digital files of artwork are not subject to interpretation or
individual preference. The reproduction must produce the same perceived color (within physical
limitations) as the original artwork. complete guideline
Page 9 IMAGE RECIEVERS GUIDELINES
1.
Getting on the Same Page with Color
Both suppliers and receivers of a digital image file need to see the same picture on their
monitors. This requires setting up systems to display standardized consistent color in a way, and
using color profiles in images to describe their color characteristics. The first step is to set
computers and monitors to display accurate color.
How to set up your system to display correct color:
Step 1. Configure Adobe Photoshop’s “Color Settings” to Preserve Embedded Profiles. In
addition, check the boxes that will cause Photoshop to warn of profile mismatches or missing
profiles when opening or pasting files.
Page 10 If Photoshop presents the Embedded Profile Mismatch warning, choose Use the embedded
profile.
Step 2. Calibrate and profile your monitor.
Calibrate and profile monitors with a hardware device. Visual calibration is inadequate for
professional image editing. Accurate visual representation of the image file is extremely
important to the imaging process. Calibration standards range from:
Gamma — 1.8 to 2.2 (or L*).
White point — 5000K to 6500K.
Brightness levels — 80cd/m² to 140 cd/m².
As a general calibration guideline for both Windows and Macintosh systems, use gamma 2.2 and
choose an appropriate color temperature between 5000K and 6500K. If preparing files for offset
printing, you will find most printers and prepress professionals use 5000K to 5500K. When
preparing files for the web, 6500K is probably the best choice, since most consumer-grade
monitors are set to higher color temperatures. Choose an appropriate monitor brightness level
according to the room’s ambient lighting before beginning your calibration.
Paired with accurate, embedded color profiles, a calibrated and profiled monitor allows you to
“soft-proof” the intended output in Adobe Photoshop. A daylight (D50) light source at correct
brightness helps visually match monitor to print. To achieve correct brightness, lower the
luminance of the light source (viewing booth) to match the luminance of the monitor.
Page 11 2.
Getting on the Same Page with File
Size and Resolution
File Dimensions
File size needs depend on intended uses. As a rule, stock image distributors require the largest
file sizes, since final usage is unknown. Magazines can be more specific, although they may
request Double Page (11 x 17 inches, A3 or 420mm x 297mm) for maximum flexibility in
placement and cropping. Graphic designers and advertising art directors should be very specific
about file sizes, since it’s best for the photographer to deliver files within 10 percent of final size
– unless the designer will handle resizing and output sharpening. Designers who do resize
images should start with a larger size and reduce it, rather than attempting to increase the
resolution of smaller image files. Web use and digital projection require much smaller file sizes,
measured in pixels, and usually saved at screen resolution of 72 ppi. Widescreen HDTV requires
a minimum of 1920 x 1080 pixels for uncropped images, though producers often ask for larger
files to allow for panning and zooming.
Uncompressed file size requirements for print RGB TIFF at 300 ppi (default for 150‐line screen) Double page (A3) 50 MB Full page (A4) 25 MB Half page (A5) 12 MB Quarter page 6 MB Eighth page 3 MB Billboard 48 MB at 600 ppi (but 300 ppi is acceptable, i.e. 24 MB) The current, uncompressed file-size standard for commercial stock images is 48 to 50 MB. A 50MB file works for larger than A3 reproduction, because posters, banners, point-of-purchase
displays and billboards print with coarser halftone screens, measured in lines per inch (lpi). As
the final size for such uses gets larger, so does the intended viewing distance, obscuring coarse
screen patterns. Hence, a 50 MB image file at 200 ppi, which is plenty of resolution for a 100-lpi
halftone or color separation, can then fill a 17x22 sheet.
Note, too, that software RIPs driving wide-format printing devices have sophisticated algorithms
that interpolate image files on the fly. If reproducing images at larger sizes than an image file’s
Page 12 pixel count allows, it is usually better to allow the RIP to resize large images – rather than using
Photoshop, and then dealing with larger files during delivery, processing and storage.
Digital Capture Quality Parameters: Sensor Size and File Dimensions
Camera sensor size and file dimensions offer only a rough guide to final printed quality. Digital
image files may be made larger or smaller by interpolation in imaging software, which adds or
removes pixels. RIP software and hardware can also interpolate during printing. Careful
interpolation can increase digital image files in size by as much as 400 percent. However, even
with the best software and a skilled operator, success depends on the qualities of the original file
and the final output required. Note that increasing resolution through interpolation does not add
detail to a file. It may, however, prevent obvious pixelation for a smoother look.
Sensor size Pixel Approx. Doc. Size Approx. 8‐bit (megapixels) Dimensions at 300 ppi File Size 17.5 MP 3413 x 5120 11 x 17 inches 50 MB 11.2 MP 2731 x 4096 9 x 13 inches 32 MB 6.3 MP 2048 x 3072 7 x 10 inches 18 MB Stock-image distributors, magazine publishers and others requesting files may specify a certain
minimum size for camera sensors used to generate image files. However, pixel quality trumps
pixel numbers, so a high-quality 6-MP sensor will often be sufficient for 95 percent of intended
uses, while pictures from a 10-MP sensor in a point-and-shoot camera may not look good beyond
8x10 inches at 300 ppi, especially if used at an ISO exposure setting above 200. Although many
publishers have standardized on 300 ppi for 150-line screens, the actual requirement is 1.3- to 2times the halftone-screen resolution, so as a practical matter, smaller files can successfully work
for a given final size.
3.
Delivery File Formats
In most cases, the choice for delivered file formats is between uncompressed TIFF and JPEG
(quality 8 or higher in the Photoshop “Save As” dialog). Both are standard formats, universally
supported by image-editing and image-cataloging software. And both can safely hold metadata.
TIFF Files
TIFF has the advantage of higher quality due to the absence of compression artifacts, but JPEGs
saved with low compression (Photoshop’s level 10 or better) will not have detectable artifacts.
Page 13 TIFF files and low-compression JPEG files may be saved and resaved without image
deterioration, but saving and re-saving a high-compression JPEG file will degrade image quality.
TIFF files can also have greater bit depth, which makes them better candidates for significant
pixel-level editing, such as color and tone adjustment. However, a TIFF file is at least five times
larger than even the least-compressed JPEG version of the same image, or 10 times larger in 16bit mode. This means TIFF files will take up five to 10 times more storage space when archived
or written to removable media for delivery. They will also require five to 10 times the bandwidth
or transmission time to send electronically. TIFF files do have several compression options:
LZW, ZIP, and JPEG. Since TIFF compression is not universally supported, it is important to
communicate which compression method you intend to use with the receiver of the files. In
general, once an image has image processing and correction by a skilled photographer or digital
imaging specialist, there is little advantage to requesting or delivering 16-bit files unless the end
user plans to make significant creative tonal or color edits to the image.
There is also a peculiarity with saving TIFF files: The option to save using PC byte order or
Macintosh byte order. The TIFF specifications require all applications to read either byte order,
so either option is correct.
JPEG Files
JPEG offers smaller sizes due to the format’s sliding compression scale. But JPEG uses “lossy”
compression algorithms that toss out some data. As the compression level increases, so does the
amount of “lost” data, decreasing both image quality and file size. The trade-off is more
compression artifacts with more compression in order to achieve smaller file sizes. Although
Photoshop’s JPEG quality level 8 is the minimum recommended value, a setting of 10 to 12 is
preferred. Other programs have different compression scales, so test to determine the
corresponding settings. For instance, Microsoft Expression Media (formerly iView Media Pro)
offers a JPEG setting of “high” that is roughly equivalent to Photoshop’s setting of 10.
Remember: Uncompressed TIFF files without generating any artifacts. JPEG files may be saved
and resaved several times at level 12 without obvious image deterioration, but saving and resaving a high-compression JPEG files will degrade image quality.
Other potential drawbacks to the JPEG format are its maximum 8-bit color depth, compared to
TIFF’s 16-bit limit. JPEG image files also cannot include layers or alpha channels.
For most people, most of the time, the JPEG format will be easier to work with than TIFF. With
proper handling, a JPEG image at quality 12 would appear equal to a TIFF, except the latter can
include a greater bit depth, layers and channels. Stock-image distributors can choose between
JPEG and TIFF, depending on their client base and preferred method of delivery. Because stock
distributors now focus on electronic delivery, JPEG files are likely the best choice. With
sufficient infrastructure, stock distributors could maintain both formats: a TIFF original, and a
method to make JPEG copies as needed.
TIFF and JPEG at a glance:
TIFF
Standard JPEG
Page 14 Uncompressed format (except TIFF
compressed)
Compressed format
No compression artifacts
Compression artifacts not usually visible
at Photoshop quality 10 and above.
JPEGS saved at quality 12 are visually
indistinguishable from TIFF.
Re-compression takes place each time an
image is resaved. Repeated saving of an
image leads to deterioration
Can be opened and saved without
damage to the image
Can be 16 bit, so better for image editing
Only 8 bit. Editing will cause more
damage to the image
Larger file. Use more storage space, but
are standard for quality archiving.
Smaller file, Good for emailing,
downloading, bulk storage of news
images.
Files are ‘flattened’ and cannot include
these.
Can include layers and alpha channels
Raw Files
When we speak of raw file delivery, we must carefully define terms. Some people consider as
“raw” an unretouched or “out-of-camera” TIFF file. Actual raw files may be of two varieties:
proprietary raw (such as .CR2, .NEF, .FFF and more than 100 others) or Adobe’s DNG format.
As a general principle, you should neither request nor deliver proprietary raw files. Proprietary
raw files are by their very nature unprocessed. They may not convey the color, tone or other
visual information the photographer intended to share. One imperfect solution would be
including a Macbeth color checker or similar tool in every frame, or in a reference image created
with the same lighting. Another way, also imperfect, would be for the photographer to adjust the
color in raw-file processing software, and then specify the receiver use an identical version of the
software to open the files. The photographer would need to carefully include reference data, such
as XMP “sidecar” files from Adobe Camera Raw or the cached color settings folder from Phase
One Capture One. Even when following these procedures, people requesting proprietary raw
files need to understand they are, in effect, preventing the photographer from creating a finished
product. There is also a risk of losing metadata, since in many cases, custom metadata
identifying the creator, copyright information, status and usage terms is not embedded in a
proprietary raw file, but is attached as a sidecar (usually XMP) file. A potential compromise
might be requesting proprietary raw files AND TIFF or JPEG files processed by the
photographer, offering a comparison to insure against inaccurate raw conversion.
DNG FILES
Some standards bodies (DISC, for instance) list DNG as a possible delivery file format. Since
DNG images are in a raw file format, they are also subject to interpretation. However, a DNG
file does contain processing information, along with metadata and a rendered JPEG file. The
rendered JPEG offers a visual representation of the photographer’s processing. This JPEG can be
viewed in Adobe Bridge, Photoshop or most cataloging software, such as Microsoft Expression
Media. While DNG is a better delivery option than proprietary raw, it stills gives the receiver
wide latitude to interpret the image. This can be a double-edged sword – good if the
photographer has poor technique, and bad if the photographer has a processing style or finished
look that is overridden by the end user. Page 15 JPEG 2000 (.jp2, .jpx, .jpf)
Created by the Joint Photographic Experts Group committee in 2000, JPEG2000 is a waveletbased, image-compression standard. JPEG2000 offers a compression performance gain of
around 20 percent compared to the standard JPEG. For lower or higher compression rates, the
improvement can be somewhat greater. It has, however, notably higher computational and
memory demands, which for most photographers means opening and closing JPEG2000 files
takes more time. Unlike the standard, baseline JPEG, JPEG2000 does offer a fully lossless
option. Improved compression options and performance, however, are not the main benefit of
JPEG2000. The real achievement of JPEG2000 is “smart decoding.” This feature enables a
decoding application (or plug-in) to access and decode only the required portion of the code
stream. This means a single JPEG2000 image can supply multiple, reduced-resolution versions
of the original. These might include specific file sizes, and/or a high-quality, high-resolution
view of a specific portion of the image. This makes JPEG2000 an excellent format if you require
the ability to smoothly zoom, pan and rotate images. Creating compressed images that contain
different quality levels allows master images in an archive to supply multiple derivatives, saving
time and bandwidth. This makes an image archive much more efficient. In addition to this array
of output options, JPEG2000 can handle very large images, at least up to a terabyte. JPEG2000,
unlike standard, baseline JPEG, supports high-bit-depth (up to 16 bits per channel vs. 8 for
standard JPEG) and high-dynamic-range images. JPEG2000 also underpins the MJ2 and JPM
formats for motion images (each frame is a JPEG2000-compressed image) and compound
images (images, graphics and text). These additional features make the JPEG2000 format a
potentially valuable option for archiving film, video and historical materials.
JPEG2000 has not been widely adopted by photographers. Raw and DNG are viewed as better
solutions for their specific needs. And JPEG2000 is not a good candidate to replace standard
JPEG as a digital capture format, since the former requires far greater compression time. Still,
photographers may want to become more familiar with JPEG2000 and be prepared to deliver in
one of its variants, since many cultural heritage and digital preservation communities use it as the
basis for their collections. Unlike photographers, these institutions are less concerned with
improved rendering options over time. Their focus is preserving a specific rendering intent in the
best way possible, with the most efficient storage and delivery options. JPEG2000 does these
things well. The unanswered question is whether this “niche” adoption of JPEG2000 will ensure
its long-term viability as an archival format. Some of the cultural institutions finding JPEG2000
to be the best current solution for their archiving needs include The Library of Congress, the
Harvard University Library, Library and Archives Canada, Chronicling America website and the
Google Library Project. For more information, see the Digital Preservation Coalition Technology
Watch Report.
HD Photo (JPEG XR)
Although this may become an important format in the future, especially if the JPEG Committee
approves this as an additional JPEG standard, lack of native support in Photoshop CS3 means
that requesting files in this format would be problematic. Originally called Windows Media
Photo, then renamed HD Photo, Microsoft is now working with the Joint Photographic Experts
Group to make the format a JPEG standard, JPEG XR (for extended dynamic range). HD Photo
has shown no compelling advantages over JPEG2000 for archival purposes. But HD Photo does
have a faster compression algorithm compared to standard JPEG, and this offers some potential
advantages as a digital capture format. Since the main drawbacks to the standard JPEG format
are the 8-bit tonal depth limit and limited dynamic range, JPEG XR could provide a welcome
step. Covered by Microsoft’s Open Specification Promise, JPEG XR avoids problems associated
with proprietary raw formats. The future of JPEG XR depends on camera makers offering it as a
Page 16 capture format. If and when this happens, then JPEG XR would become a potential archival
format as well.
HD Photo is also the basis for the Microsoft Seadragon project and the related Photosynth photo
viewing technology. These new technologies could also drive the adoption of HD Photo, since
they both use HD Photo’s ability to decode only the needed portions of an image, allowing for
rapid screen draws as images are combined, zoomed, and panned. The latest Windows operating
systems include full support for HD Photo. Updated version of HD Photo Plug-ins are available
from Bill Crow’s HD Photo Blog http://blogs.msdn.com/billcrow/archive/2007/08/17/updatedhd-photo-plug-ins-for-adobe-photoshop-software.aspx Although OS X doesn't currently support
HD Photo as a native format, it is supported through installing the HD Photo Plug-ins for Adobe
Photoshop CS2 and CS3. Whenever Photoshop writes an HD Photo file, it also creates a
thumbnail on the resource fork in a compatible format for OS X. The result is that thumbnails are
created that show in the OS X finder.
4.
File Delivery Options
File extensions. Regardless of file format, always add file extensions, especially for crossplatform transmission or delivery. (Multiple extensions such as “.CR2.TIF” can sometimes
confuse both systems and operators. Use only a single, correct extension for the file type.)
ReadMe files. UPDIG recommends including ReadMe files on all delivered media and with
transmitted image files. The ReadMe should list and describe all delivered files, their
characteristics and usage rights. ReadMe files must be specific. Here is a sample ReadMe file for
digital image files delivered to a graphic designer for use in an annual report:
ABOUT THESE DIGITAL IMAGE FILES: These files are copyright XYZ.
RIGHTS GRANTED: One-time, exclusive use by The ABC Company for its 2008 annual report.
All rights are reserved except those specifically granted by the invoice.
Page 17 These files are optimized for print applications. They are in Adobe RGB(1998) color space. For
accurate color, preserve the embedded profile when opening.
These files have been sharpened for output. If they are reduced in size by more than 20 percent,
they may need to be sharpened slightly.
For more information about digital file handling: http://www.UPDIG.org.
Cross-rendered CMYK guide prints can offer a very general reference point for digital image
files. But photographers who include CMYK guide prints with their digital files must profile
their printers. This is particularly true with CMYK files submitted to offset printers. Unless using
a professional proofing system, photographers should include disclaimers stating guide prints are
for color reference only and are not “verifiable” proofs such as contract proofs. Contract proofs
can only be supplied by the offset printer hired to produce the job.
Provision of CMYK guide prints cannot be a key requirement in the stock industry or in many
areas of the publishing industry. The numbers of images are too great and/or file delivery is
increasingly done electronically. CMYK Guide prints may be requested for color critical work.
However even in this area, soft proofing of the delivered files is becoming the norm. In many
cases, instead of relying on guide prints, photographers should routinely soft proof in the output
color space.
Since inkjet prints made from RGB editing spaces may have wider color gamuts than available
from an offset press, a CMYK guide print will more accurately reflect what is possible to
achieve on an offset press if the print is “cross rendered.” Cross rendering involves printing from
your current output color space through an intermediate output color space to simulate the
appearance of the final output space. For example, in the print dialogue, the “source” space
would be either the CMYK file or the CMYK proof space. Choosing Relative Colorimetric as
the rendering intent will limit a desktop printer’s color to the gamut of the CMYK file. If you use
Absolute Colorimetric (or check the “simulate paper color” box in Photoshop CS2 or CS3), you
may more closely simulate the actual press sheet, since the whites will more closely match the
duller white of the actual press stock.
For more detailed information about creating cross-rendered guide prints, see the UPDIG Image
Creators Guideline for CMYK Guide Prints and Verifiable Proofs.
CD-R or DVD-R delivery requires recording in ISO 9660 cross-platform format. It is helpful
to use inkjet-writable media or LightScribe media, on which certain optical disc burners can
produce laser-etched labels. Do not accept or send media with adhesive labels, since they may
separate from the disc and damage optical drives.
Hard drives can be the most efficient media for delivering large or large numbers of files. If
you request hard-drive delivery, note compatibility issues between the Macintosh and Windows
file systems. Since Macs have common USB ports and can read and write to Windows-formatted
media, it’s best to use an external USB or USB 2.0 hard drive, initialized with the standard
“Windows FAT” format. Note the FAT system limits individual file sizes to four gigabytes.
Another way to avoid compatibility issues is to use network drives over a local area network,
which is increasingly practical as common network speeds increase.
Page 18 Electronic delivery
Electronic delivery is increasingly attractive and viable. Digital wrappers, such as ZIP and SIT
archives, can simplify transmitting multiple files, incorporate a delivery ReadMe file and be
secured with a password. The “Compress” function in Macintosh OS X 10.5.x (“Create Archive
of” in OS X 10.4 and 10.3) will append a period and the ZIP extension to the resulting,
compressed file or folder. For example, a folder called “Images” becomes the archive
“Images.zip.” Note, however, if you choose a single file with an existing extension, Mac’s
“Compress” will create an archive file name with two periods and two extensions. For example,
“image1.tif will become “image1.tif.zip.” Mac systems have no problem with the double
extension, but web servers aren’t always so tolerant. Because of security concerns, many ISP’s
will automatically delete or quarantine such files.
Electronic Delivery Methods
Internet Download. Web file delivery is convenient. The file can be transferred onto a web
server system by browser upload, FTP or other transfer method. The person delivering the file
will require the specific address and path of the delivery directory on the receiving web server, as
well as any needed login information.
Email Attachment. So long as the file being transferred is not too large, email attachment can
be a dependable way to receive an image file. Attachments as large as 1 or 2 megabytes work
reliably with most services. Some email services will accept huge attachments, but most do not.
Check before requesting large files. Note that with the huge volume of email coursing through
the Internet at any given moment, valuable files are hard for a thief to discern, so reasonably
sized files are relatively secure when sent as email attachments.
Supplier and client must both have access to email services and email software. But many email
service providers offer a web browser interface to email accounts, usually without security. Most
browser-based web email services include file upload capabilities, but more than a few still do
not.
5.
ICC Profiles
File Delivery Color Profile Options
There are two primary color modes in use : RGB and CMYK. RGB (Red, Green, Blue) is the
standard archival and delivery color space, and holds more color information than CMYK (Cyan,
Magenta, Yellow, Black) which is used for print output.
The conversion of an RGB file to CMYK involves a loss of color information that is not
reversible, so in general it is best to retain images in RGB as far as possible in the workflow.
CMYK’s image files with 4 channels are 25 percent larger than the equivalent RGB with 3
channels. Adobe RGB (1998) is the most common delivery standard, with ECI RGB being
adopted by some publishers in Europe. Unfortunately, the sRGB color space is probably used
more often than either, although its smaller color gamut clips some colors that fall within the
gamut of an offset press rendering those otherwise printable colors un-printable.
Page 19 Color profiles specify which color space the image is in, and should always be attached to the
image when saving, and preserved when opening, for correct viewing and handling. It is
impossible to interpret the color numbers in an image correctly and with certainty if the color
space is not known.
RGB or CMYK for File Delivery?
The two color modes offer a choice of RGB or CMYK file delivery by photographers. Which is
best depends on a number of factors:
a) General publishing or high-end advertising/printing.
b) The skill and pre-press knowledge of the photographer.
c) Whether the printer/press/press profile is known in advance.
d) Whether the photographer is delivering a final finished image, or whether there will be
other color edits or retouching that will be done to the image by a graphic designer,
retouch artist or other.
e) The skill and pre-press knowledge of the graphic designer and/or the printer.
f) The likelihood that embedded RGB profiles will be honored.
In general, the industry trend is for most photographers to deliver RGB files, and for graphic
designers to convert them to CMYK, except for cases where the printer makes the conversion.
Many high-end photographers prefer to deliver CMYK files, particularly those who handle
retouching or other special processes in the workflow. This requires good communication and
buy-in by the client/designer/printer so that the photographer has all the required information.
Knowing the exact press profile is not always necessary as many printers use device link profiles
to optimize CMYK files to their specific press conditions. The advantage of device link profiles
over CMYK-CMYK conversions using Photoshop, is that the black (K) channel is preserved.
CMYK delivery has sometimes been referred to as an “early binding” workflow. The main
advantage of early binding workflow is simplicity, and less reliance that others down the
workflow chain will do something wrong (strip or ignore and embedded RGB profile), or not do
something (forget to sharpen image files for output after resizing).
RGB delivery, and maintaining the files in RGB until just before inclusion in the page
layout/.PDF/or even by means of the platesetter’s or imagesetter’s RIP is referred to as “late
binding” workflow. The advantage of this workflow is flexibility. Image files can be worked on
by people down the workflow, and the image files can be easily repurposed for different types of
print output, or even used for display or web use. The caveat is that everyone in the workflow
needs to understand and use color management and correct file handling and output sharpening.
Another potential disadvantage- but one many clients have some familiarity with- is that the
quality of the final output is not known until you are at the proof, or even the press stage. With
that as an introduction, here is a chart showing the main differences between the two workflows,
followed by information about the RGB workflow, followed by information about the CMYK
workflow. Page 20 RGB
CMYK
Image will pass through a number of
hands on its way to output.
Shorter supply chain to design, or print
Image for repurposing – print, web,
presentation etc.
Use in print only
The default for general publishing use.
An option in controlled high-end colour
critical situations
Always use when the image is to be
archived for further use
Not suitable for archiving
Color range restricted for print output.
Larger gamut than CMYK for most
colors
Smaller file size than CMYK – 3
channels, R,G, and B.
Larger file size due to 4 channels, C,M,Y
and K.
Page 21 The RGB workflow:
The most commonly used RGB profiles include:
!
!
!
sRGB, a narrow-gamut but common space.
Adobe RGB (1998), a wider-gamut color space.
ECI-RGB, a color space increasingly popular in Europe.
Adobe RGB (1998) and ECI-RGB come much closer than sRGB to encompassing the color
gamut of standard CMYK profiles. Those involved in high-end, fine-art inkjet printing often use
ProPhoto RGB, which has an even wider gamut. But the wider gamut leaves more room for
error, making it problematic for standard delivery. Note, too, that proper display of colors in
Adobe RGB (1998) files requires calibrated monitors and users who preserve embedded profiles.
Under ideal circumstances, it’s best to work with the Adobe RGB (1998) or ECI color space, but
the additional color gamut can become a liability if the printer strips out or ignores the embedded
profile, resulting in flat and toneless images. Despite these problems, it is important to
communicate that the default setting in Europe and US Pre-press defaults in Photoshop is
currently Adobe RGB (1998). Everyone in the workflow should be encouraged to implement
color-managed workflows for best results.
General principles for RGB file delivery:
!
!
!
!
!
If no submission guidelines are provided, the standard submission format is Adobe RGB
for print or general use, and sRGB for web only use.
Photographers should always embed the RGB profile.
Photographers state clearly what they are providing in the read-me file.
Receivers of RGB files should always preserve embedded RGB profiles when opening
files.
If the client asks for sRGB for print, then that is what the photographer has to provide.
The results are the responsibility of the client.
For web and digital projection, sRGB is the clear choice. The web is a lowest common
denominator medium, where you must assume monitors are not calibrated and web browsers are
unable to read color profile tags.
For more information about web browser color management, visit these web pages:
http://www.northlight-images.co.uk/article_pages/web-browser-color-management.html*
http://regex.info/blog/photo-tech/color-spaces-page2/*
The CMYK Workflow:
The most commonly used CMYK profiles include:
!
!
!
US Web Coated (SWOP) v2, ships with Photoshop as the North American Prepress 2
default.
Coated FOGRA27 (ISO 12647-2-2004), ships with Photoshop as the Europe Prepress 2
default
Japan Color 2001 Coated, the Japan Prepress 2 default.
Page 22 !
Many printers in the US are starting to use the G7 methodology and GRACol CMYK
profiles for their sheetfed presses. These new CMYK profiles will likely ship in the new
versions of Photoshop.
Although the industry trend is for the RGB workflow to extend further towards the print output,
the vast majority of small- to medium-size offset printers still prefer receiving CMYK files, even
if they often can’t specify exactly which CMYK profile they need. As result, many graphic
designers use Photoshop’s Image>Mode>CMYK to convert RGB files to CMYK. By doing so,
they convert to the default CMYK profile in Photoshop’s Color Settings. The initial default
setting is U.S. Web Coated (SWOP) v2, or Coated FOGRA27 in Europe, even though these may
not be optimal profiles. Graphic designers need to assess the choice of CMYK delivery based on
the capabilities (or lack of them) of the photographer or the printer to convert files. In some highend situations, the photographer may be skilled at CMYK conversion and will achieve better
results than the designer. In these instances, the photographer will need access to the printer
settings. Our testing and other anecdotal evidence leads us to believe that U.S. Web Coated
(SWOP) v2 in the U.S. at least, is much preferred over the U.S. Sheetfed v2 profiles even for
sheetfed presses. This is because most printers have standardized on the SWOP profile and not
the Sheetfed profile. Printers that have gone over to G7/GRACoL printing are finding that while
the GRACoL profiles are optimal, the SWOP profiles still work quite well, which was an
intended result of the G7 method. One additional, but little known fact is that many printers
routinely re-separate CMYK files to better match their press conditions, or even to save ink by
trimming total ink coverage and/or increasing GCR (grey component replacement). Some
printers use Photoshop to do this, which while not optimal, works, while others use device link
profiles, a better method.
6.
Sharpening
After resolution and color, sharpening can have the most impact on perceived image quality.
Most digital cameras deliver slightly soft images (unless the capture format is set to JPEG or
TIFF with in-camera sharpening turned up). This results from the almost universal use of antialiasing filters to combat moiré patterns. Scanning camera backs do not need AA filters Some
medium-format backs – especially those with multi-shot mode – also may not have an AA filter.
If we assume single-shot raw file capture, it is entirely appropriate to request a small amount of
default “capture” sharpening. Without it, digital image files may look too soft.
Stock image distributors often request “unsharpened” digital files. It might be better to
request no sharpening beyond capture sharpening, since online catalog images need to look
reasonably sharp. Most raw processing software applies a certain amount of sharpening as a
default, although you can usually adjust this or turn it off. Some photographers, worried there
will be no output sharpening, increase this somewhat when processing raw files. Every stock
distributor should educate clients the best ways to sharpen images from their collection, since
digital images will only look their best when sharpened specifically for the type and final size of
output.
Magazine publishers usually specify unsharpened files, as well. But a small amount of
capture sharpening will do no harm, while neglecting to sharpen will definitely affect printed
quality.
Page 23 Publication designers and web designers vary in their understanding of sharpening. One
possible strategy is to deliver layered TIFF or PSD files at or near final size, with sharpening
applied to a duplicate layer, and a ReadMe file and/or delivery memo explaining the layer. Name
such a layer “sharpening.” Fuzzy pictures on the web often result from extreme down-sampling
of large camera files. One strategy is to deliver smaller files to web designers, typically in the
1024- to 768-pixel range for height or width, and apply a degree of sharpening that looks right at
that size. Better web designers will request no sharpening beyond capture (or raw conversion),
but will indicate they plan to sharpen appropriately as they resize.
Sharpening is a very complex topic. For a more complete discussion, see Bruce Fraser’s Real
World Image Sharpening, Peachpit Press, 2006.
7.
Other File Quality Parameters
Excessive unskilled pixel editing can create quality problems
Too much Photoshop-style pixel editing of color and tone, as well as oversharpening, can easily
degrade image quality. Keep an image file in 16-bit mode if planning major adjustments, since
the greater bit depth allows wider adjustments without loss of detail, posterization or banding.
Editing in a raw converter or in 16-bit mode helps avoid some of the problems associated with
aggressive adjustments. Editing images encoded with a greater bit depth than standard 8-bit
images offers more intermediate “slots” where the pixel values can move, eliminating many gaps
and spikes in the histogram of the finished image.
Result of excessive tone adjustments in an 8-bit file; the comb-like gaps in the graph indicate
unused tonal values. Shading gradations that should be imperceptibly smooth will become
noticeable, often called “banding.”
Page 24 Scanning
Use high-quality scanners for film originals. But note the quality of the scan depends as much on
operator skill as on the technology employed. Inexpensive flatbed scanners designed for
reflective art and documents are not recommended for film. Scans should be viewed at 100
percent on screen to detect dust, hairs, fingerprints and any other problems. Make and store the
scans in a wide-gamut color space. Ideally, profile the scanner with a test target and profiling
software. Delivered images should be free of sensor dust spots, scan lines, digital aberrations,
artifacts, banding, excessive grain, moiré, chromatic aberrations and excessive noise.
Bad processing examples
Michelle Alvarado, photography director at Panoramic Images, has provided the following
illustrations of common digital image quality problems. Some of these examples are from film
scans, others from digital capture.
Blocked color. Image shows blocked black and bluish color areas. Also, the shadows on the
hippo skin appear to be blue in some areas, suggesting a chromatic aberration.
Photo © Panoramic Images
Page 25 File corruption. Three examples show some different types of visual file corruption. Note the
heavy green lines at the bottom of the left image. The middle image has a line that changes
position each time this file is opened. The right image has a fixed line that could be mistaken for
a power line in this skyline image.
Photos © Panoramic Images
Page 26 Dirty scan or sensor. This needs spotting to remove specks of dirt, dust and hairs. A dirty digital
camera sensor can create a similar effect. If your camera exhibits evidence of dust on the sensor,
clean the sensor.
Photo © Panoramic Images
Page 27 Oversharpened grain. Example of a scanned image with more than usual grain. It also appears
oversharpened. If an image is grainy (or a digital image is noisy), avoid oversharpening. There
are Photoshop techniques (adding blur) and plug-ins, as well as stand-alone programs, designed
to reduce noise and grain.
Photo © Panoramic Images
Page 28 Moiré. This striped shirt has a moiré pattern. Moiré can result from interference between the
sensor grid, and a grid occurring in the original scene. Moving closer or farther away, or tilting
the camera slightly, may alleviate moiré at the time of capture. Many cameras have anti-aliasing
filters in front to their sensors to counteract moiré, but some do not. If you notice moiré in a
captured image, some raw processing software and Photoshop plug-ins can remove or reduce it.
Photo © Panoramic Images
Page 29 Noise and scan lines. Green scan lines, along with green and red noise, appear in the shadow
areas. Digital cameras set at high ISO sensitivities often generate luminance or color noise. If
high ISO settings are unavoidable, you can reduce this noise with controls in raw processing
software, with Photoshop filters and plug-ins, or with standalone programs.
Photo © Panoramic Images
Page 30 Poor stitching technique. Two digital images are combined below, leaving visible stitch lines
and repeated picture elements. High-quality stitching software and careful technique can help
reduce this. Carefully evaluate stitched images at 100 percent or larger on your monitor.
Photo © Panoramic Images
Limiting output levels for print
Although we have recommended in the past that digital images prepared for commercial printing
should have output levels set between 3 and 253 (RGB mode).
The restriction of levels to between 3 and 253 is a) unnecessary
http://photoshopnews.com/2008/01/14/camera-raw-and-the-shadow-output-levels/ and b) a
reflection of incorrect handling of shadows and highlights. (For example: lifting shadow detail
Page 31 will do more for the image.) The amount of tone restriction depends on the output and should
properly be done close to print output. The printer may pull the range in much further. Many
publishers will repurpose images for web use the future. It is a mistake to handle all images as if
they were being output to print.
8.
File Naming
To avoid problems with files transferred across computing platforms, request file names that use
only the letters of the Latin alphabet (A-Z, a-z), the numerals 0-9, hyphens and underscores.
Avoid other punctuation marks, accented letters, non-Latin letters, and other non-standard
characters, such as \:/*<> or brackets. On a local network or with rewritable media, limit the file
name to 31 characters or less (including the three-letter file extension). Limit file names to 11
characters or less (including the three-letter file extension) when burning to optical media, since
some computers don’t support long file names. Use a single period (.) between the file name and
extension.
Specify unique file names. Multiple files with the same name will cause problems for computers
and people alike, and a newer file might automatically overwrite an older with the same name, or
vice-versa. You may want to specify including the numeric date and/or the photographers name
as part of the file name as a way to avoid duplicate names.
For a complete guide to file-naming protocol, see the Controlled Vocabulary web site.
Stock image distributors have the greatest need to implement their own file naming
conventions, since they have the largest collections, along with the greatest need to search them
and link images to their creators. Ideally, stock image distributors would append their naming
convention to a creator’s file name, but this can become unwieldy. Future versions of the IPTC
Core metadata schema should allow storage of multiple filenames. In the interim, putting both
the stock image distributor’s file name and the creator’s file name in the Document Title field
may help reduce confusion.
Magazine and newspaper publishers may have similar needs if they maintain picture archives.
Publication and web designers generally require only a user-friendly naming convention.
Naming files sequentially and grouping them in a way that relates to the needs of the project
usually accomplishes this. Designers should never rename files, but it is usually not a problem if
they add to the file name, inserting their modification before the file extension. For instance, if
the original file name is 07055397.TIFF, it could be modified to 07055397CMYK.TIFF, or
07055_397RedTruck.TIFF.
Until the IPTC Photo Metadata working groups have standardized fields that can accommodate
original file names, place the file name in the Document Title field, so that if the file is renamed,
it can still be located with its original or delivered name. There are batch methods available for
this in Bridge, Expression Media (formerly iView Media Pro), BreezeBrowser and
PhotoMechanic.
Page 32 What happens if someone changes the file name? Putting the photographer’s file name in the
Document Title field preserves it as a searchable criterion should someone change the name of
the image file.
9.
Metadata
Overview
Digital image metadata is information about the image. Information like picture number, caption,
credit and copyright that used to be carried on the hard copy, should now be found within the
digital file so that images are captioned and can be traced back to their supplier and copyright
holder.
The use of metadata means that image workflows can be more efficiently automated. If
information is added to the image at an early stage, it may not need re-keying, saving time and
money throughout the supply chain. All organizations handling images need metadata policies to
ensure retention of standard critical information and creation and onward retention of other job
specific data. It’s increasingly common for stock-image distributors and magazine publishers to
develop custom XMP metadata panels that conform to their particular needs.
Technical metadata is created at image capture or creation, and is carried in the image as EXIF
data and ICC profiles. This information should be retained by software systems handling images.
Removal should be possible for exceptional cases where its retention is in contravention of data,
personal or state security. (e.g. The GPS of an image may be hazardous to photographer
or subject or to a nation's security).
Rights, captioning and licensing metadata is carried in XMP format in the image file. The
various standard schemes in use overlap in places. The information can appear as Adobe XMP
fields, IPTC fields, and PLUS fields and is found in the image under File/fileinfo in Photoshop.
It can also be viewed by other image viewing and archiving programs, though they vary in the
fields they display. It’s increasingly common for stock-image distributors and magazine
publishers to develop custom XMP metadata panels that conform to their particular needs.A
good example is the BAPLA metadata template.
The ability to readily identify and contact the rights holder for an image will likely become
critically important if (when) proposed “Orphan Works” copyright exemptions become law. Bills
introduced in the U.S. Congress, if passed, would affect all photographers by allowing anyone
possessing an image to copy, modify, distribute, publish and otherwise use it for any purpose,
after a reasonably diligent search fails to identify the rights holder. All images previously
distributed without rights metadata may soon be vulnerable to unlicensed use as “orphaned”
images. Embedding complete, accurate rights metadata will help protect images from unlicensed
use. Many other countries are considering similar legislation.
Metadata can be embedded in standard file formats, such as TIFF and JPEG. You can also
embed metadata in raw files, although this is only recommended for the Adobe DNG file format,
since proprietary raw formats are neither standardized nor publicly documented. For now, with
proprietary formats, it’s best to attach metadata in a sidecar file. You can also save metadata in
Page 33 an image database, making sure the image file and its metadata reside in the same volume, folder
or program, in order for the metadata to connect to the image file.
A variety of programs can read and write metadata. Microsoft Expression Media (iView Media
Pro), PhotoMechanic and ImageIngester can add metadata as files transfer from camera card to
computer. Image editing programs (such as Adobe Photoshop), Image browsers, (such as Adobe
Bridge), cataloging programs (such as Expression Media/iView, Extensis Portfolio and Canto
Cumulus), and all-in-one programs (such as Adobe Lightroom and Apple’s Aperture) can read,
write and edit metadata. Some camera makers’ processing software, such as Nikon’s Capture
NX, provide some basic (usually only legacy IPTC and not IPTC Core) metadata writing, as do
some raw processors, such as Capture One.
, Content and Status panels that appear under the File > File Info menu in Photoshop.
Types of Metadata:
EXIF: All digital cameras include technical metadata, called EXIF data. This provides a host of
information, such as the camera make and model, its serial number, the date and time of image
capture, the shutter speed, lens used, the ISO speed setting, and often, other technical details,
such as white balance and distance to the subject. Raw file processing software can use this
information to more accurately render images.
ICC Color Profile Tags: Without a color profile tag, the person receiving an image file can
only guess the color space used to create and edit it. Always embed an ICC profile in digital
images.
IPTC: This includes user-supplied information. The earliest schema was the IPTC IIM
(Information Interchange Module) model created by the newspaper industry in 1991 and
incorporated into Photoshop in 1995. While considered a legacy format, it remains widely used
and readable by most software that accesses metadata. The IIM format stores information
separately but shares many fields with the XMP format. Referred to as the IPTC Core , or IPTC
Core Schema for XMP, it includes the Description, IPTC Contact, Imagee Extensible Metadata
Platform or XMP is a specific type of extensible markup language used in PDF, photography and
photo editing applications. Adobe introduced XMP in 1991. Adobe, IPTC and IDEAlliance
collaborated to introduce in 2005 the IPTC Core Schema for XMP, which transfers metadata
values from IPTC headers to the more modern and flexible XMP. Stock image distributors and
other organizations have since created custom metadata panels that enhance metadata usefulness.
PLUS: The Picture Licensing Universal System is an integrated set of standards for
communicating rights metadata associated with commissioned and stock images. The PLUS
standards are developed, approved and maintained by the PLUS Coalition, an international, nonprofit umbrella association that includes publishers, designers, advertising agencies,
photographers, illustrators, stock image distributors, artist representatives, museums, libraries,
and standards bodies, such as UPDIG, IPTC, IDEAlliance and others.
Metadata in Use
The type and amount of metadata photographers should embed in their images depends on who will receive the image files. Page 34 Stock image distributors can’t function without metadata. They depend particularly on
keywords. Stock distributors that ask photographers to add keywords should offer detailed
guidance. Captions, which are titles or explanations of images, should be foundations for the
keywords that follow. Good keywords explain the who, what, where, when and how of a picture
– important tools for finding specific image files.
You’ll find an excellent guide to keywording on the Controlled Vocabulary web site.
In addition to the standard IPTC templates, stock image distributors may elect to create their own
custom metadata panels – not only to capture additional data but also to organize it differently
from the IPTC defaults. It’s easy to install these XMP-based panels in Photoshop (see
http://www.adobe.com/products/xmp/custompanel.html), and they offer a good way for stock
image distributors to insert custom metadata into image files. Although often handled separately,
photo releases can be stored in metadata by using a custom template. However, this custom
metadata is not yet easily available or interoperable with most image database or cataloging
software.
Magazines and publishers also depend on metadata, particularly caption, headline, author,
contact information and usage rights. Image files with complete information can save magazines
valuable time on deadline. Picture metadata can even provide additional reporting information,
especially details such as GPS data and capture time.
Many publishers have rights-tracking software to verify they have all necessary licenses, have
made all necessary payments, and conversely, can track licenses to others for using images they
own. Good metadata reduces the cost of operating such systems.
Publishers who lack such software still may need to contact rights holders, perhaps for many
years to come. They may need to defend themselves from lawsuits (libel, invasion of privacy,
plagiarism, etc.) or undertake a regulatory compliance audit. If a publisher is sold, part of the
buyer’s “due diligence” includes ensuring its assets are untainted by legal clouds. Metadata
reduces the need to rely on paper records.
For all of these reasons, photographers should determine magazine and publisher metadata
needs, and when appropriate, install and populate custom metadata templates.
Publication designers can improve their workflow by organizing and categorizing pictures
for metadata searches in browsers such as Bridge and digital asset management software such as
Microsoft Expression Media/iView, Extensis Portfolio or Canto Cumulus. For example, if a
design firm commissions a series of portraits, it is often critical to include the portrait subject’s
name in the Description and/or Headline fields. Other fields, such as Origin, can be equally
important to a designer assembling a printed piece. The photographer should ascertain these
metadata needs, or make the designer aware of how this custom metadata can improve the
designer’s workflow.
Web designers often have the same organizational and search needs as publications designers.
But they have an additional responsibility to prevent metadata stripping. Unfortunately, in the
initial release of Adobe Photoshop CS3, Save For Web (called Save For Web & Devices in CS3)
strips most metadata. Although it saves the file description and the creator's copyright notice by
default, the feature changes the copyright status to Unknown. Beginning with the first CS3
Page 35 update (Photoshop 10.0.1) it has become easier to
discover how to include XMP metadata when using
Save For Web. In the updated version, the Save for Web
and Devices command includes the option to "Include
XMP" in the main Save for Web dialog, where users
will more likely find it. (In the first CS3 release, the
option was buried in a sub-sub-menu.) When this option
is enabled, all XMP data – including the creator
copyright and contact info – is included in the optimized
file, and the copyright status is preserved.
Regrettably, while easier to find than before, the option
is not enabled by default. Users must discover it by
clicking on a tiny triangle to open a fly-out menu (see
illustration), and set it. Once set, however, it stays set.
For most users, Photoshop CS3 will automatically
request the upgrade from the Adobe web site and install
it. If not, you can download the updater:
For Macintosh:
http://www.adobe.com/support/downloads/detail.jsp?ftp
ID=3774
For Windows: http://www.adobe.com/support/downloads/detail.jsp?ftpID=3775
Please note "include XMP" means just that. The new file will not include Legacy IPTC-format
metadata. Although many programs that read metadata recognize XMP, there are exceptions,
including Apple Preview. This means web designers may still want to use "Save As" instead of
"Save For Web & Devices," particularly if their clients are photographers, since some magazine
editors review photos in the Preview.
10. Fine Art Reproduction – What
Museums Need
Creating and preparing digital reproductions of artworks requires careful attention to details
noted above, as well as some specific concerns. Unlike typical photo reproduction, the colors
within digital files of artworks should not be subject to interpretation or individual preference.
The reproduction must produce the same perceived color (within physical limitations) as the
original artwork. To help achieve this goal, it is particularly important to implement the section
of the UPDIG guidelines that concerns preserving embedded profiles and calibrating your
monitor.
An ICC color-managed workflow is essential for proper color handling. This workflow is based
on colorimetric matching. In fine-art reproduction, appearance matching must also be considered
as a further refinement to the ICC workflow. Practically, this means the image must be
considered as a whole and within the context in which it will be viewed. Some appearance
matching issues to consider include:
Page 36 1. When comparing a soft proof to original art, match the white point in both viewing
conditions. If viewing the artwork in a booth with 5000K lighting, then the monitor
displaying the screen image should also be calibrated to a 5000K white point. Otherwise,
the mismatch will invalidate any evaluations of color, because your eyes cannot
chromatically adapt to both white points simultaneously. (see footnote)
NOTE: For more detail, please refer to ISO/DIS 12646:2008 (Graphic technology —
displays for colour proofing – characteristics and viewing conditions) and ISO 3664:2000
(Viewing conditions – Graphic technology and photography).
2. When comparing a soft proof to a hard-copy proof, a white border of at least 1 inch must
surround the soft proof.
3. Do not compare individual colors in isolation, since they may need to be different
colorimetric colors in order to produce the same color appearance within the context of
the whole image.
4. View the reproduction at an appropriate viewing distance. If it is to be printed on an
8½x11-inch page and the original art is much larger, if viewed together they must not be
side-by-side, but rather each should be assessed at their natural viewing distances.
For art reproduction, raw and DNG formats are not recommended as submission formats, since
the images will need different rendering, depending on the final output, in order to maintain
consistent color appearance. The only way to control this is by providing a rendered image
tagged with an ICC profile.
Imaging Workflow
Image Capture. The only way to control digital captures for art reproduction is with
professional equipment. Consumer-grade cameras and scanners introduce image processing
intended to create preferred color for typical photography. This type of processing is counterproductive when the goal is appearance matching.
For two-dimensional art, first perform flat fielding to compensate for lens falloff and uneven
lighting. For 3-dimensional art, flat fielding is necessary only while capturing the calibration
target. Next, calibrate the camera's color. There are several solutions available for this.
Depending on the art being photographed, some solutions may perform better than others, so try
multiple solutions to find which work(s) best for different art media.
The most common reason for a camera calibration to fail is metamerism. Because cameras see
color differently than people, different artists' materials (oil paint, watercolor, silver-gelatin
photographs, chalk, etc.) can appear to have the same color yet photograph differently. The more
widely the spectral nature of the materials photographed differ from the materials in the
calibration target (such as a GretagMacbeth Digital ColorChecker SG), the more likely there will
be color errors. Making a custom target from the same materials you are photographing can
alleviate this problem, although doing so properly is difficult.
Scanning. Scanning transparencies is more straightforward than digital capture. Use one of the
excellent, available software solutions to characterize (profile) your scanner. It is also very
important to consider whether a transparency accurately represents the art. Transparencies have
Page 37 narrower limits in color fidelity when compared to digital captures and generally require more
color-editing effort.
Printing. Characterize your printer by creating an ICC profile, and apply that profile in your
printing workflow. Affordable consultants can help with this.
File Standards. Images extracted from the internet and video are unacceptable for art
reproduction. Most images from these sources contain insufficient resolution or color gamut to
produce quality reproductions, lack proper color management information and could have
copyright problems.
Applying these best practices should produce digital files that meet the general specifications
outlined above, although modifying existing files to meet these standards can be difficult. Such
practices as resampling, color-space conversions (changing the color gamut), color mode
conversions (between RGB and CMYK) and file format conversion (e.g., low-quality JPEG to
TIFF) may cause more harm than good. Unless you have the experience to understand the
consequences of each adjustment, leave an image as it is. The best workflow is one that
conforms to these standards from the initial image capture.
File delivery
Outbound. When supplying images, it is also important to conform to the guidelines in this
document. Ideally, provide a guide print along with a ReadMe file to instruct printers in correct
color handling. As online delivery systems for press-ready files continue to proliferate, it will not
always be possible or cost-effective to deliver guide prints. For this reason, printers who have
adopted an ICC workflow should be favored.
Inbound. To receive images from outside sources, ask the provider for image files that conform
to the guidelines in this document. Sample request language follows:
File Request Template
Please provide digital files as follows:
Resolution: 400 pixels per inch
Size:
12 inches long dimension
Bit-Depth:
8-bit
Color Space: Adobe RGB (1998)
File Format: TIFF or high-quality JPEG
ICC Profile: Embedded
Page 38 NOTES:
!
!
!
!
!
When viewed under standard prepress conditions – 5000K illumination of the artwork
and 5000K monitor white point – the digital file's appearance should match the original
work.
You may sharpen the image to compensate for loss of detail during capture, but do not
apply output sharpening, which will vary based on a particular press and reproduction
size.
The artworks should be photographed unframed and unmatted.
Please provide information to link digital file names to the works of art.
The production of these files should follow best practices in color management: Profile
the camera and assign that profile to the original capture. Then convert that file to Adobe
RGB (1998) color space and embed that profile.
Files submitted as above allow us to extract details and make moderate-size reproductions
without up-sampling.
A more detailed discussion of the topics introduced here, including templates and instructions for
printers, can be found at www.ImageMuse.org.
Appendix
ICC Color Management
ICC International Color Consortium ICC White Papers
X-rite color measurement and color management, offering hardware, software and services for measuring,
formulating and matching color including Color Checkers
Chromix Color Services, Color Think Software
SWOP Specifications Web Offset Publications
SWOP-certified systems This page lists the SWOP certified proofing systems
GRACoL General Requirements for Applications in Commercial Offset Lithography
FOGRA Graphic Technology Research Association
IFRA Global research and service organization for the news publishing industry.
ECI European Color Initiative
ISO International Organization for Standardization
Dry Creek Photo Digital Imaging Resources for Photographers by Photographers
Colorremedies Color Management Essential Training
Metadata
EXIF.org - EXIF and related resources
Orphan Works Orphan works are basically works whose copyright owners cannot be located.
EXIF The specification uses the existing JPEG, TIFF Rev. 6.0, and RIFF WAV file formats, with the
addition of specific metadata tags. It is not supported in JPEG 2000, PNG, or GIF.
IPTC IIM the first multi-media news exchange format
IPTC Core This page is about IPTC metadata to be used within Adobe's Extensible Metadata Platform
framework.
XMP (Extensible Metadata Platform) a standard for metadata, created by Adobe Systems Inc.
PLUS standards An international non-profit initiative on a mission to simplify and facilitate the
Page 39 communication and management of image rights.
Controlled Vocabulary web site Recommendations for Limitations on Image Filenaming
http://www.adobe.com/products/xmp/custompanel.html The custom File Info panel for XMP metadata
allows you to define, create, and manage custom metadata properties using standard Adobe applications
Software
Bibble Professional Workflow and raw Conversion software for Windows, Mac OS X, and Linux
Capture NX Nikon’s software for simpler, more intuitive full-scale digital image processing and editing
Capture One Raw workflow software from Phase One
Aperture Import, manage, and enhance your photos in one simple, integrated workflow
Microsoft Expression Media 2 is a powerful and customizable image management software application
for importing, annotating, organizing, archiving, searching and distributing your ever-growing collection
of digital files.
HindSight Software for Professional Photographers including MetaMachine
PhotoMechanic Photo Mechanic is designed to work with groups of photos together in order to manage
them.
ImageIngester double backup; naming; moving to folders; applying bulk metadata and Camera
Raw/Lightroom settings; GPS tagging; converting to DNG; verifying; correlating multi-camera shoots.
BreezeBrowser Used to download, browse, select, organize and post process their images, create web
galleries and control cameras from a Windows PC.
Cumulus Digital Asset Management software and hardware
Portfolio Digital Asset Management software and hardware
DXO Optics Pro is a multi-award winning program running on Windows or Macintosh to automatically
improve image quality.
Silkypix Developer Studio adjust the color, tone, and the white balance of the image can truly help you
express your feelings through the photography.
DAM Useful Digital Asset Management and Workflow Software
Adobe Software
ACR (Adobe Camera Raw) Digital camera raw file support.
AcrCalibrator Tom Fors’ calibration script for ACR. Visual Vacations ACR calibration tutorial
Lightroom Lightroom® software is the professional photographer's essential toolbox, providing one easy
application for managing, adjusting, and presenting large volumes of digital photographs so you can
spend less time in front of the computer and more time behind the lens.
Monitor Calibration
PDF manual how to calibrate and profile both CRT and LCD monitors
Gretag-MacBeth Eye-One Colorimeter
Color Vision Spyder 3 Elite Colorimeter
ColorEyes Display Color profiling software and hardware
Profile Verification Kit by Pixl Ltd.
Black Point Check
Grayscale Test
W cathode ray tube
W liquid crystal display
W white point
W color temperature
W color look up table
Page 40 W gamma correction
W luminance
File Formats
DNG file format (Digital Negative) The public, archival format for digital camera raw data
Luminous Landscape site A good analysis of the raw Flaw
OpenRAW Organization hoping to persuade camera makers to provide open documentation of their
proprietary raw formats.
OpenRAW survey The 2006 raw Survey – A Report on the Experiences, Requirements, Belifs, and
Preferences of Photographers and Imaging Professionals regarding raw Imaging Technology
File Delivery
YouSendIt FTP file delivery service
SmugMug Online gallery and image sharing service
Web Color
http://www.northlight-images.co.uk/article_pages/web-browser-color-management.html* web browser
color management
http://regex.info/blog/photo-tech/color-spaces-page2/* web browser color management
Books
The DAM Book: Digital Asset Management for Photographers, By Peter Krogh
Practical Color Management: Eddie Tapp Real World Camera RAW: Broce Fraser and Jeff Schewe Real World Color Management: Bruce Fraser, Chris Murphy, Fred Bunting O’Reilly O'Reilly Media spreads the knowledge of innovators through its books, online services, magazine, and conferences Peach Pit Essential resources for the creative community Blogs and Forums
The Dam Forum
John Nack on Adobe
The Lightroom Journal
Photoshop News
Consultants
Peter Krogh- Digital Asset Management
Andrew Rodney- Color Management
Neil Barstow‐ Color Management
Chris Murphy‐ Color Management
Page 41