I want to demonstrate my Photoshop students the inner working of a
Matrix/TRC monitor profile.
I created a profile for my monitor using its default 2.2/5166K calibration.
I then extracted the XYZ White Point and Primaries from the profile.
I then set the Linear Tristimulus matrix from the primaries the usual way :
Red
Green
Blue
86.526
29.523
20.755
39.162
95.009
6.892
1.406
13.347
111.285
Then I calculate the Inverse Matrix (easy to do in Excel):
Inverse Matrix
0.013325
-0.00382
-0.00225
-0.00553
0.012205
0.000275
0.000495
-0.00142
0.008981
Then I run some XYZ color through the inverted matrix.
For demonstration sake, I use the Yellow patch on the ColorChecker chart:
60.476
61.307
8.302
Which gets me the Linear RGB values.
Then I correct for 2.2 gamma and get these gamma-corrected RGB values :
194.75
171.19
40.76
Naturally, it does not agree with Photoshop RGB values :
227
198
56
There are not CAT involved of any kind in this v2.1 profile so the numbers I
get from the profile agree with the 'absolute' measurements coming from my
instrument.
What could I possibly be missing?
/ Roger