Absolute vs Relative Colorimetry
Absolute vs Relative Colorimetry
- Subject: Absolute vs Relative Colorimetry
- From: Roger Breton <email@hidden>
- Date: Thu, 05 Apr 2001 23:00:15 -0400
In Relative Colorimetry, I understand colors get mapped from the Source
space to the Destination space using common L* a* b* values. That would be
going from Input Device coordinates to Lab -- through the Input Profile A2B1
tag, on the one hand as a first transform, then going from Lab to Output
Device coordinates -- this time through the Output Profile B2A1 tag. Right?
Colors falling outside the Destination gamut getting clipped to the outer
boundery of the Destination gamut, preserving hue angle at the expense of
Lightness and Chroma. (Please correct me if I am wrong!)
The Source white point somehow gets "dropped", "ignored" or "forgotten" in
this conversion (it's not "simulated" onto the Destination like in absolute
colorimetric conversion).
But how do individual in-gamut colors get "formulated" in the Destination
space? Don't they have to take into account the color of the Destination
Media (Destination White point), somehow, as that's what they were measured
on in the first place, at the time of measuring the profiling target?
Finally, if I were to compare the spectral curve of, say, patch F8 from a
real Kodak IT8.7/2 versus patch F8 as imaged through a combination of a
scanner profile (created from that very IT8) and a printer profile, what
should I find, in theory? Would the two different spectral curves match (not
in theory, you say, because of metamerism)? **But** how close would you say
the resulting Lab value shluld read? Plus or minus a few Delta Es, if my
system is well profiled and behaved and I have used good quality profiling
software?
Thank's in advance for bearing with me and for your kind and patient help.
Roger Breton
Quibec Institute of Graphic Communications