[argyllcms] Re: [argyllcms]

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Fri, 08 Sep 2006 19:49:20 +1000

Gerhard Fuernkranz wrote:

I guess, when I create a .gam file with iccgamut/tiffgamut for later use with 
the -g/-G options of profile/icclink, then it is important
to create the .gam file with the right -c and -i options, which are consitent with the 
"-c" viewing conditions which I intend to use
later with profile/icclink, is that correct?

Yes, one of the "gotchas" is that there is no direct co-ordination of such details, and little checking either (I think I added some rudimentary checking in the last release of software to icclink, ie. checking for Lab or Jab, but I haven't yet put all the relevant information there to be checked, nor do I thin profile checks.) There is a note to that effect in the tiffgamut documentation.

If I simply create a abscol/relcol L*a*b* .gam file, then I suspect that 
profile/icclink will not convert the supplied .gam file
implicitly to the CAM02 gamut corresponding to the supplied "-c" viewing 
conditions, does it?

No, icclink will issue a warning about the Lab/Jab mismatch, but that is all. It will not notice if the viewing conditions are different, not will (or can) it attempt to correct the situation. There is a subtle difference between creating a gamut surface in a particular colorspace, vs. creating in another, and converting the surface points to a space, and I haven't been down the path of figuring out whether such a thing should be attempted. (An overarching application would solve the problem simply by ensuring that the parameters were the same).

Graeme Gill.

Other related posts: