Re: LAB as working/interchange space
Re: LAB as working/interchange space
- Subject: Re: LAB as working/interchange space
- From: Peter MacLeod <email@hidden>
- Date: Fri, 25 May 2001 16:17:26 -0700
email@hidden writes:
Just a question. Why is it assumed CIELAB is the best color appearance
model?
It's not. A color appearance model is a model that predicts what
a color would look like under changes in viewing conditions,
such as changes in illuminant, surround luminance, etc.
CIELAB is considered a color appearance model under this definition,
but not a very good one--in particular, it's chromatic adaptation
model is inaccurate, and it doesn't have control over any appearance
parameters but chromatic adaptation.
CIELAB was developed as a perceptually uniform color space,
in which one could compute euclidean distances between colors
in the space and get a meaningful measure of how far apart
the colors were perceptually. There are more recent color
difference formulae that correct some inaccuracies in
CIELAB, such as CMC, CIE94, etc. CIELAB is a fairly old
standard (1976).
There are more comprehensive color appearance models that
have been proposed, including at least one CIE standard
called CIECAM97s.
--Peter