When to use ‘Absolute Colorimetric’?
The concept of ‘Absolute Colorimetric’ is relatively easy. It counts paper color. Usually we use (255, 255, 255) as white, and we expect (245, 245, 245) prints very light gray. But in the color management world, most papers are darker than (245, 245, 245). In a sense, (245, 245, 245) should print no inks, i.e. clipping around white. This is what ‘Absolute Colorimetric’ does. In most cases, this is not what we expect. So we use ‘Relative Colorimetric’ with BPC everyday, and feel curious about the raison d’être of ‘Absolute Colorimetric.’
You might never encounter it, but there are some cases which require ‘Absolute Colorimetric.’ Let’s see three of them.
Printing a replica of painting
Most traditional paintings don’t have any blank margin. Its replica shouldn’t have it too. In this case, ‘Relative Colorimetric’ is irrelevant because there is no area of zero inks. Use ‘Absolute Colorimetric.’ The printing image should be very colorimetric, of course.
Printing the same colors on multiple materials
The idea which absolutely makes you complain: Printing the same colors on multiple materials.
Most materials have illuminant dependency. The color matching over multiple materials are always metameric. Inks degrade unevenly on different materials. The color matching is not only metameric but also ephemeral.
But someday you might encounter a client who immunes to your persuasion. ‘Absolute Colorimetric’ is for the day.
(255, 255, 255) is out-of-gamut, so you need to adjust levels to avoid clipping around white. Black requires the adjustment too (in ‘Relative Colorimetric’ BPC does the job). And treat saturated colors which are out-of-gamut. Different materials have different gamuts, so you need to jam colors in the intersection of all gamuts.
When you don’t like paper color cast
‘Relative Colorimetric’ ignores paper color. So paper color cast occurs unless the paper is true gray. If you don’t like the color cast, you should use ‘Absolute Colorimetric’ and treat out-of-gamut colors manually.