Color Management Myths 6-10

From ColorWiki

Revision as of 18:30, 6 June 2007 by Patrick (Talk | contribs)
Jump to: navigation, search
Reserved Article

This page is a
Reserved Article.
For more details see
Reserved ColorWiki Articles

ColorNews

This reserved article originally appeared in CHROMiX ColorNews Issue 11 on February 6, 2004.

Click here to see the original in its original context.
Email
colornews(at)chromix.com to subscribe to the ColorNews newsletter.

by Steve Upton


As there seems to be no end of strange color management advice offered these days, it seemed like a good idea to dispel a few more myths we've run across recently.

Contents

Myth #6: You need to be a color scientist to use color management

Not true! although it helps....

Seriously, color management still isn't quite at the "cookbook" stage - where you can just press buttons and it all works - but it is getting close. We regularly sell and install monitor calibrators, profiles and other color management gear in clients who don't really want to know too much about it - and they're happy with the results. Those who have entered into the world of color management will agree that the water is very deep. There seems to be no limit to how much you can learn and how far you can go in the pursuit of "perfect color".

Rest assured though, you can get into color management and take control of your imaging without breaking either your bank account or your brain!

Myth #7: Setup Photoshop with your monitor profile as the working space.

While this seems to be happening less often as in the past, a recent article in a popular photo magazine article contained this dubious piece of advice.

If you set Photoshop to use your monitor profile as your working space, all new RGB documents will be created using your monitor profile as their gamut definition.

This is a bad idea for a handful of reasons:

  1. All colors will be limited to the gamut of your monitor. As all ColorThink users know, the gamut of your monitor (and sRGB for that matter) tends to be smaller than that of most output systems. Inkjets, presses and photographic output methods typically contain cyans, greens, yellows and even some reds and magentas that monitors cannot display. Photoshop 5 and later versions effectively disconnected your RGB file from your display, allowing non-displayable colors to pass through Photoshop and be printed. This is a good thing, don't undo it.
  2. Each time your re-profile your display you will have a different working space. Color management is enough of a moving target without changing your working space every 2-4 weeks (you are re-profiling every 2-4 weeks right?).
  3. Exchanging files with your monitor profile embedded will cause confusion. The reactions from people who receive files with embedded monitor profiles range from "wha?" to "amateur!". As mentioned in past ColorNews newsletters, when you exchange files with other people you should convert to standard color spaces such as Adobe RGB (1998), sRGB and so forth.

Refer to ColorNews issue #8: [Input Profiles and Working Spaces]

Myth #8: Use perceptual intent for all Photographic-style images

This falls in the "When in doubt, be very very safe" realm. All rendering in Photoshop 4 was relative colorimetric so any out-of-gamut colors were clipped. One of the benefits of Photoshop 5's inclusion of ICC technology was the availability of the perceptual rendering intent. Almost overnight, anyone who grasped ICC profile use was recommending the perceptual intent for photographic images and relative colorimetric for spot colors. In fact, some applications go as far as calling the perceptual intent the "photo" intent.

I think a little rendering intent background would help here...

When an ICC profile is calculated, the color conversions are "hard wired" into its internal tables. This means that the gamut-shrinking capabilities of the perceptual intent are built without any idea of any actual image content. It is a sort of "generic" conversion that is intended to work for any image.

Trouble is, if you have an image that is entirely in-gamut for your printer and you choose the perceptual intent, your image will be desaturated unnecessarily. The can range from a barely noticeable shift of near-neutrals to a significant dulling of saturated colors.

Configure Photoshop's Proof Setup with your printer profile. Then invoke the gamut warning (shift-command-Y Mac or shift-control-Y Win) prior to converting your file. If you see a bunch of out-of-gamut colors in important areas of your image, choose the perceptual intent when you print. If few colors seem affected, chances are that relative colorimetric (with black point compensation ON) is a better choice for your image.

Myth #9: The Saturation rendering intent sucks

"Use perceptual for shrinking your gamut in a friendly way, relative colorimetric for color accuracy and absolute colorimetric for proofing. Don't bother with saturation unless your doing business graphics like graphs"

This has been the line about rendering intents that has been passed around a lot. I have to admit that I had been one of the ones passing around until a few years ago when I started to actually test the saturation intent of profiles we were building.

Most profiling software creates the saturation rendering intent in the same manner as the perceptual intent. So the most likely time you will want to use it is if you are shrinking the gamut of your image and you want to retain details in out-of-gamut colors. In our experiments we've found that the saturation intent will map out-of-gamut colors to brighter and cleaner colors rather that more accurate ones. The color shifting performed by the saturation intent will differ between different profiling packages so your results may vary.

At any rate, if you are not satisfied by the saturation in your profile's perceptual rendering intent, give the saturation intent a try. You may be pleasantly surprised by the results.

Myth #10: Profile Rot or "A good profile gone bad"

We build a lot of profiles here at CHROMiX and with the assumption of ProfileCity's profiling business we expect to build a lot more. A big part of building profiles for people is supporting them. It's something we feel differentiates us from the rest.

One question that seems to crop up regularly in our support team is that of the "freshness" of profiles. People call to ask how long their profile will last or to explain a problem and wonder if their profile has gone bad.

Profiles are hermetically sealed, antibacterial, non-fungal, disease resistant, fade-proof, rust-proof, colorfast little wonders. That is (I guess it needs to be said) they do not undergo any change as they sit on your hard drive and will not experience wear and tear even with years of use.

The devices they represent however, are a totally different matter. If your profile worked well one day and then poorly the next, something in the path to (or from) that device has changed. Most problems seem to arise from a change in consumables. Other common problems include updated software or a maintenance issue - like clogged inkjets with a printer.

When your print profile was first built, a certain CMYK/RGB combination produced a certain Lab color. All the profile knows is that relationship and if anything changes the color produced, then the profile is no longer appropriate for the printer.

There are some cases where all the testing of a profile was with a certain image type so the entire capability of the profile/device was never tested. A similar thing can occur when, for example, a CMYK profile is successfully used for proofing and months later, when the profile is used for separations, a problem shows up. In both these cases the flaw was always present in the profile and incomplete testing lead the user to believe it was a good profile when in fact it was not...

Steve Upton

For more Color Management Myths, see Color_Management_Myths_11-15

Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox