This is the core article of the section. Select items from the menu to clarify subjects otherwise not fully covered here.

Review of CM Concepts and Definitions

By now we know what CM is - its definition and objectives and why we might want to learn more. Following is a summary and review of useful definitions.

Color Management (CM):  The collection of processes and technologies whereby we minimize changes in the intended appearance of an image as it moves from one presentation environment to another. Alternatively, if we desire preparing images for a variety of viewing environments, CM affords predictability of results.

Color Space (CS): The formal, mathematical description of the range of colors or color gamut an image could represent or the color gamut a device such as a printer could display. Color spaces may be device independent and standardized, such as sRGB or they may be device specific such as when they are part of a Device Profile. There are numerous standard color spaces and countless device dependent color spaces. Even though a device dependent color space may not be standard, it is represented in a standardized format which makes it possible for the Color Management Module to locate it within the Profile Connection Space. CM makes it easy to move an image among Color Spaces with controlled changes to appearance. The most widely used standard color spaces are sRGB and aRGB (Adobe). A color space is a subset of the Profile Connection Space (see below).

Device Profile: Combination of a device native color space plus calibration data used to standardize and correct characteristics such as Gamma (essentially contrast), Color Temperature (color cast) and deviations from expected tonal progression as we move from darker to brighter parts of an image.

ICM/ICC Extension: As far as I can determine, these file extensions are interchangeable and are used to identify both device profiles and color spaces. There is an unfortunate ambiguity rampant in discussion of these files. A file of this type could contain just a color space or ... a color space plus device calibration data. Almost universally, both types of file are referred to as "profiles". Properly speaking, only the second type is a profile while the first is a color space - period.

Device Calibration: The process of building a Device Profile (see above). Calibration can be achieved by eye, using test charts but a much more satisfactory result is obtained using a colorimeter and supporting software. A calibrated monitor is the heart of a color managed environment. A device profile contains a description of the native color space plus data used to correct to some standard the device's response to color information.

Profile Connection Space (PCS): The superset of all color spaces, comprising all visible colors, derived from the 1931 CIE model based on perceptual research.

Color Aware: A term used to describe applications that recognize image/device color spaces and are able tp use Color Management to preserve intended appearance as images move from one color space to the next. Most simple display software (Internet explorer, Windows FAX and Image Viewer, etc.) is not color aware and ignores embedded Color Spaces. Images displayed by these applications are, therefore rendered as if they existed in the default CS of the environment. This is usually sRGB or the current CS of the monitor if it is using a calibration profile.

Color Management Module (CMM): Standard software used by color aware systems and applications to map a source color space into a receiving or target color space. This is the translation process that minimizes changes in image appearance when moving an image from its native color space into one that is different. The CMM first locates the source color space within the PCS and then determines where in the target color space that subset of the PCS lies. Colors from the subset outside the target space (out of gamut colors, if any) are modified according to rendering intent so that they fit. Unless conversion is requested, translation of an image into a new color space is temporary or non-destructive.

sRGB/aRGB: The most popular and second most popular color spaces. These are standard, device independent spaces. sRGB is an implied standard for Internet image exchange and assumed by non-color aware display software and devices. aRGB is a superior/larger color space and is gaining favor, especially for printing.

Embedded Color Space: When a camera or scanner image is produced the device implants the Color Space specification used. This becomes part of the image file. It informs Color Management software of the image Color Space without your having to remember or document it. This embedded space can easily be removed or changed, depending on the conversions (if any) you later perform. Following conversion from camera RAW an image always has an embedded CS.

Color Space Mismatch: When the embedded color space of an image does not match the color space of a target device or the Working Color Space of software the image will not be correctly displayed unless it is color managed into the new space. This is what Color Management is all about. It is now necessary to go through the PCS using the CMM as described elsewhere, remapping the image CS into the target CS. Dealing with a mismatch is usually automated but may involve having to make a choice, with conversion (as opposed to temporary translation) being an option. This will depend on your software settings.

Color Working Space (CWS): This applies to image editing software. The monitor will usually be set up to work within its native CS using a calibrated Device Profile. Working in this space will result in many color space mismatches, requiring manual responses (in the case of some software although you can easily automate this in Photoshop). It is usually best to have the software configured to use the CS of your images. This will invoke color management to automatically map the embedded image CS into the monitor CS. The monitor calibration corrections will still be effective (a Color Space is not, after all, a profile).

The rest of this article deals with specific applications of Color Management technology and methods. Certain topics will be referenced and elaborated upon by means of links which are also available from the menu.

Monitor Calibration

CLICK HERE for complete discussion - A properly calibrated monitor is they key to preparing images in a way that ensures they will look the way you intend. Of course, this will also depend somewhat upon how well your viewers'  monitors are calibrated but these days we can usually assume reasonable presets are in effect. Furthermore, Internet software is increasingly Color Aware. When it comes to printing, you are in full control. A calibrated monitor can provide an excellent approximation of how your work will look in print. Whether or not you decide to get involved with color management, calibration will help you produce better images. Note that not all monitors are the same. You can find out more HERE.

Cameras and Color Spaces

The better cameras provide you a choice of shooting color space which is invariably sRGB/aRGB. This applies only to the optional production of JPEG and TIFF images. For best results, we shoot RAW/NEF which does not acquire a color space until it is "developed" by software such as ACR, Silkypix and many others. It is usual for post-RAW images to have an embedded color space, supplied by the software according to user option.

A useful strategy is to set your camera to sRGB and choose the option producing both RAW and JPG. You then have camera processed JPGs almost ready for the Internet (you may want to resize them) and the RAWs which you can develop as aRGBs for printing or other highest quality presentation.

TIFF is a higher quality format than JPEG but when produced in the camera is always 8 bits / channel. This seems like a waste of the format. Better to shoot RAW and then have the option of converting to 8 or 16 bit TIFF in software.

There is to me a problem with the observation so often made that RAW files do not come with a color space. Clearly, the gamut of a color sensor must have limitations and these must define a color space that is a subset of the Profile Connection Space. The camera must use this when converting RAW internally to JPEG or TIFF (???). It isn't clear how or even if the software converter on your computer gets access to the camera profile. When I can answer this question I will document findings here.  For now, let's assume that since cameras offer aRGB as their best option, that is the assumed color space of their RAW files. As for calibration data needed to complete the camera device profile, I have no idea where or if this exists.

Image Preparation

If you are mainly scrap-booking or sharing a few shots on the Internet you can just print or post you JPEGs directly and they might be OK. At The White Creek Gallery we aspire to fine photography (at least when the mood strikes us) so we are going to use an image processing package or packages to make our images look their best.

RAW/NEF Development:

For the best results choose an output color space aRGB at 16 bits per channel. That means a TIFF output file format because JPEG supports only 8 bits/channel. 8 bits/channel TIFF looks just as good to me as 16 bit TIFF but will not withstand as much processing before posterization threatens, especially with aRGB color space which compresses colors more than sRGB. This is somewhat a judgment call. If you do not run into posterization problems with the sort of processing you do, 8 bits/channel ought to be fine.

sRGB is more congenial to 8 bits/channel because it does not compress colors as much as aRGB. 8 bit sRGB TIFFs can endure a lot of processing without posterization so this may be a better choice if you heavily massage your images.

If your intention is Internet display via browser you will have to produce JPEGs at some point. These are easily created from 8 or 16 bit TIFFs. OR ... you can use the JPEGs created by the camera.

Never discard your RAW files. They are the ultimate source from which you can create any other format image files. You burn your bridges by making and keeping only JPEGs. If you really are determined to get rid of your RAWs, maintain your image archive in TIFF format.

Color Working Space

For image processing your monitor should be set to its Custom preset mode with the calibration profile loaded. This maximizes the monitor's ability to render properly any image you load (especially if it is a wide gamut monitor). Your imaging package such as Photoshop should usually be set to use a Color Working Space that is the same as that of the image you will be working with. Here we get into a potentially swampy area dealt with in more detail HERE.

Printing

In the color managed environment the CMM is invoked to map your image CS into the CS of printer+paper and ink. The printer manufacturer typically supplies the numerous profiles comprising the color spaces and calibration data for each case. These are professionally developed profiles and more than adequate for most work although you can purchase the very expensive equipment and go through the tedious effort of printer calibration. These printer profiles are automatically stored in the same place as monitor and other profiles when you install the printer software. There are also sources of 3rd party profiles that might be useful if you need to print using other than the printer manufacturer's papers. Naturally, the printer maker is not going to provide profiles for a competitor's media.

There are three main ways to color manage and print your images:

1. Turn off Color Management (for printing) in your imaging software and configure your printer drivers to do the job. This is almost never recommended owing to the minimalist environment environment offered. The drivers will read the image CS specification and use the CMM to map this into the printer CS.
2. Turn off Color Management in the printer drivers (this is an option in the printer control panel) and enable CM for printing in the imaging software. This is the preferred route to take, offering better flexibility.
3. The very best way (my opinion) ... use a dedicated printing application. My current choice is QImage. Being dedicated to printing, this product has many more useful printing features than an imaging package, such as advanced interpolation and re-sizing algorithms and an all-inclusive approach to color management. QImage is inexpensive and easy to use. There are other similar products so consider this option.

Internet Images

It is totally unnecessary and a waste of both storage and bandwidth to post high density images to the Internet unless you have some unusual intention in mind, an example being huge files you want someone to download and then print on a poster. For convenient monitor viewing we are looking at images 8 X 10 maximum.

Consider the following:

• Even the best, hi-def monitors are low resolution devices. They have a native resolution of approximately 100 dpi and can resolve about 1/3 to 1/4 the detail of a glossy printed page. In photographic terms, monitor resolution is poor to fair. If you post an image at greater than 100 dpi it will be reduced to that resolution by means of the browser expanding it. It may then be too big to see all of it without scrolling. The viewer could use one of the browser re-scaling options to zoom out. Any extra detail is then lost because of interpolation. There simply is no point in posting high density images for ordinary viewing.
• High density images can take a long time to download. This will irritate the people trying to view them and could discourage visits to your web site if that's where you are posting.
• Internet images have to be JPEGs. If you convert to JPEG from a very large file there will be contouring.

The steps - take them in the order shown:

1. Images should be in the sRGB Color Space. Your image processing software can convert if necessary. Some day it may not matter what color space you use. We will all have wide-gamut monitors and 100% color aware viewing environments. Wait 10 years.
2. Finish any editing touch-up that still needs to be done.
3. Re-size to actual physical dimensions with optimum re-sampling. This ensures the image has the right size on a monitor running in native-mode resolution.  Good sizes (in inches) are 8 X 10, 5 X 7, 4 X 6. Choose bi-cubic re-sampling, which minimizes aliasing effects (straight lines going jagged). Use 100 dots/inch. This is a close match to most monitors. Longest side 10 inches will nicely fit most screens but viewers running their units in non-native (low resolution) mode may be obliged to do some scrolling or zooming out. For variety try other dimensions that complement your composition. Why not 1 X 8 for a snake?
4. Optionally apply borders or frame. This is a matter of taste. Very easy to do in Paintshop but have so far found nothing in Photoshop. The best borders are black and/or white because they are color-neutral.
5. Convert to JPEG. Again, your imaging software knows how. Your image may already be JPEG so just check the size and re-convert choosing more compression if necessary. Choose a higher quality setting. In Paintshop, select 20% compression. In Photoshop, select high Medium or low High.
6. Check the imaqe size. I find after all this it seldom exceeds 300k which is very manageable for most people. Most files end up a good deal smaller - in the 100K to 200K range.
7. Be careful with backgrounds. You will find most of the better photo sites use black or dark gray backgrounds for images. Such a background is color neutral and avoids the illusion of color-cast in images. Again, this is somewhat a matter of taste. There may be cases where a colored background complements an image. Although white is color-neutral it can be distracting as a background because it floods the eye with too much light.

These final images have the following characteristics:

• They are as sharp as can be represented for their physical size on a monitor.
• There are no significant compression artifacts
• They are compact, download quickly and display almost instantaneously.
• Being sRGB they will look great on any monitor.
• They will not withstand any enlargement. Make them bigger and they will begin looking soft. They are optimum just as they are. If you want people to be able to download higher density versions for making big prints, provide a separate file they can retrieve.
• If viewed on a monitor not running in native mode resolution, appearance may be degraded. This is because the virtual pixel size is increased in non-native resolution modes. The image is, therefore, expanded as pointed outearlier. The degradation is really only visible in more extreme cases, such as viewing at 1024 X 768 on a 1920 X 1080 monitor. The only way you can partially anticipate this problem is by making your images smaller and increasing the dot density. For example, settle on a largest size of 4 X 6 using 120 dpi. In native resolution mode this would expand to 100 dpi (full resolution), 7.2 inches on the long side whereas on a monitor at 1.5 X pixel expansion that would work out to 9 inches and 80 dpi which is still reasonable for detail. Really, though, this is not a critical issue. I have tested my 100 dpi images in non-native mode at considerable expansion and they are still acceptable for the most part. If your objective is image quality, why not cater for the native mode environment?

These images look every bit as good as a 2 megabyte image scaled down to fit the monitor by viewing software. Far too many people post or send as attachments humungous images in the belief these have more detail. They don't, unless you view them at much larger than screen size, which no one does.