[argyllcms] Re: Specifying an algorithm type override for display profiles

  • From: Gerhard Fuernkranz <nospam456@xxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Wed, 06 Feb 2013 15:03:55 +0100

Am 06.02.2013 12:37, schrieb Graeme Gill:

Maybe something could be done before profiling by evaluating how additive the 
sample set is

Hi Graeme,

in case of tri-chromatic readings (with colorimeter) the only way coming into 
my mind is attempting to fit a matrix/TRC model to the data.

With spectral readings, the additivity of the _emitted spectra_ can be verified easily, 
but the display electronics could unfortunately still introduce some non-linear 
interdependencies between the RGB channels (maybe even intentionally, for "color 
correction" purposes, which certainly would not make profiling easier), so that the 
whole monitor may still not appear as additive, even if it's panel is additive.

One quandary is that in the ICC way of doing things, you can only have the 
(baked in) gamut mapping with cLUT type profiles. Use of matrix profiles tends 
to imply gamut per component clipping, which is not very elegant.

Some limited perceptual mapping (e.g. adjust gamma in order to account for 
viewing conditions, black point mapping, desaturate colors via the matrix if 
the display can't reproduce satured colors) could be basically done with a 
matrix profile as well, but this would be a non-standard profile then, since 
matix profiles are colorimetric by definiton.

Best Regards,

Other related posts: