Help required: Getting primary illuminants and white point from ICC file

[Graeme Gill]

Richard Hughes wrote:
> I'm sort-of stuck here. I'm confused about which cLUT I should be
> using (TRC curves?), as different profiles are doing very different
> things. Could you please elaborate about how to run the primaries
> through the table? Thanks.

Well I can only talk specifically about my set of tools:

Using icclib command line tools it would be something like this:

     icclu -ff -ia -px sRGB.icm
     1 0 0
     1.000000 0.000000 0.000000 [RGB] -> MatrixFwd -> 0.412385 0.212596 0.019306 [XYZ]
     0 1 0
     0.000000 1.000000 0.000000 [RGB] -> MatrixFwd -> 0.357573 0.715199 0.119206 [XYZ]
     0 0 1
     0.000000 0.000000 1.000000 [RGB] -> MatrixFwd -> 0.180500 0.072205 0.950546 [XYZ]

Progromatically, one would do the equivalent, e.g. using icclib
open the profile, create a lookup object to do forward absolute colorimetric
to XYZ space, and then feed colorant only combinations through it.
icclu.c will illustrate the details. I'm sure something similar is
possible using Marti's lcms.

I wouldn't recommend trying to re-create the wheel and do it from
ICC tags as there are a number of details to get right.

Graeme Gill.