Hello! I use argyll for photo camera calibration. I found spectral responce curvers for Canon EOS 5d and 5d mark II. (here - http://astrosurf.com/buil/50d/test.htm ) So, now I simulate XYZ and camera responce for some number of paint's of known spectral responce under some light with known visible spectral curve. I use 19 colors: 18 color patches from colochecker24 and 1 "pure white". Spectral responce for colorcheker and standart 4500K light coluld be found in internet easilly. So, it gives 19. I need to get a "gamma 2.2". I suppose camera sensor is linear, and "dcraw -o 0 -g 2.2 0" gives pure "gamma 2.2" raw file. So, I have to make 255*19 charts, for every brightness, and a "0 0 0" for black. It gives 4846 color charts. So, first question: 1. IS IT POSSIBLE TO SET "GAMMA 2.2" IN INPUT FOR COLPROF UTILITY SOMEHOW, NOT BY MAKING A LOT OF "AUXILIARY COLOR CHARTS"? Next, I use this script: @echo off set manufacture=-A "Canon" set model=-M "EOS 5D mark II" set desc=-D "EOS 5D mkII, 4500K, matrix" set common=-qu -v set mode=-aG set input=4500 set output=-O Canon_5Dmk2_4500K_matrix.icm colprof -v %manufacture% %model% %desc% %mode% %output% %input% (with this 4500.ti3 file http://lj.vladimirovich.net/lj11/spectral/4500.zip, so you could download, unzip and try yourself). And it gives me result: No of test patches = 4846 Creating matrix and curves... 100% Matrix = 0.832540 -0.023902 0.143377 0.348156 0.766601 -0.124534 0.026696 -0.169944 0.969710 Gamma = 2.196714 Find white & black points White point XYZ = 0.039651 0.041242 0.034422 Black point XYZ = 0.000000 0.000000 0.000000 Fixup matrix for white point After white point adjust: Matrix = 20.266359 -0.537054 3.421069 8.477817 18.560198 -3.028086 0.636536 -4.059754 23.229008 Done gamma/shaper and matrix creation Profile done Profile check complete, peak err = 4.545609, avg err = 1.047500 So, next question: 2. WHY WHITE POINT IS "0.039651 0.041242 0.034422" AND NOT A19 96.42 100 82.491 100 100 100 1.00000000 1.00000000 1.00000000 FROM A TI3 FILE? Thank you, Yan Sizov