Ben Goren wrote:
I intend to play around with this this weekend -- assuming I can get the continuous ink flow system for my printer to actually work.... I'm trying to figure out if you meant: xicclu -f if -i r -p x or xiiclu -f ib -i r -p x in order to generate XYZ from RGB -- or possibly something else entirely for -f.
OK, let's say we have a lut based RGB->PCS profile that we've created called device.icm. To create the candidate test point RGB values I'd use: xicclu -fif -ir device.icm and feed in suitable Lab values: 90.000000 0.000000 0.000000 [Lab] -> Lut -> 0.912927 0.927358 0.937031 [RGB] 80.000000 0.000000 0.000000 [Lab] -> Lut -> 0.819480 0.853146 0.855490 [RGB] . . 20.000000 0.000000 0.000000 [Lab] -> Lut -> 0.287507 0.345025 0.367405 [RGB] 10.000000 0.000000 0.000000 [Lab] -> Lut -> 0.143174 0.177814 0.191570 [RGB] So the RGB values would form the new grey axis test points. If I want to add real XYZ expected target values for these points (to be consistent with the rest of the chart recognition stuff), then I would feed the RGB value back into the profile in the forward direction. xicclu -ff -ia -px device.icm 0.912927 0.927358 0.937031 [RGB] -> Lut -> 0.633962 0.660017 0.585528 [XYZ] 0.819480 0.853146 0.855490 [RGB] -> Lut -> 0.470933 0.490288 0.434955 [XYZ] Since (by convention) the CGATS XYZ values are scaled to 100, you would need to multiply the XYZ by 100, e.g. add 63.3962 66.0017 58.5528, 47.0933 49.0288 43.4955 etc. as the expected XYZ in the .ti1 file. Graeme Gill.