CHROMiX

ColorNews Issue#29 ColorThink Pro tip - Pat Herold

Reading the interesting article of “fine-tune” profiling you can read that if you have an image that has some problem colors in it you can sample those colors and give the profile software more detail how to treat them. It also says you have to convert the image to the printer profile you are using before making any samples. This is what I don’t understand:

Why do you have to convert the image to the profile? As this profile does not do a great job with this colors, by converting the image to the printer profile this will not be the correct colors to sample right!! … If you have color problems should you not use the colors before the conversion so the profiling software have the “true” colors it has problems with?? When the problem colors are converted they are already treated into the printer profile. It does not sound logic to me or am I missing something here? :open_mouth:

Regards
Soren

Your question sounds logical to me, in that it seems silly to convert to a profile one wants to change, especially since the colours that are the worst/going to be chaged are the ones we are going to be adding to the target & using to rebuild the profile.

Anyone know what the deal here is?

It certainly is a logical question.

However, keep in mind that what you’re making through this process is a profiling target that will be printed on your printer. Therefore, all the colors on the target that get printed out will necessarily be in the color space of the printer anyway. We’re trying to sample similar colors to the ones that you’re having trouble with; it does not have to be perfect, we just need something that is nearby, to give the profiling software more information about that color area so it can do a more complete job. Even though the existing profile is not perfect, it will be good enough for this purpose (unless you have a really bad printer profile, and in that case you should probably start all over!)

If you are really concerned about this, then take it as an iterative process. Convert the image to the profile, sample the offending colors, make a new target, create a new profile, test it by printing with it. If it’s still bad, convert the image to the new profile, and sample the colors again, etc…