[argyllcms] Re: Spectral illuminant sample

Marco N. wrote:
I have tried the replacement in the .sp file, the module spec2cie makes the conversion, but regarding the found values ...
I have tried two conversions, a D50 and a custom illuminant (10000K). I have opened the ti3 resulting files and look at the first sample (first patch of the colorchecker) after BEGIN DATA 


1. the RGB value are the same in the two files (in what colorspace?)


The colorspace of the device of course. The whole point of the data
file is as a basis to create the device <-> CIE model.
2. the spectral values are the same (they are the spectral data samples of the CC) 

Yes.
3. The Lab value is very different, the D50 value are ok, the 10000K value are wrong (first patch D50: 39.773 12.878 13.695 10000K: 38.689 7.1901 -7.0360 this last is a purple-gray?) 

Sounds reasonable. You've switched to a very blue colored illuminant
compared to D50, so naturally the CIE values have shifted towards blue.
Note that throughout all of the code, XYZ and Lab values are
effectively interchangeable, the Lab is always a D50 white point
conversion of the XYZ, as is the ICC convention.

Graeme Gill.

Other related posts: