Graeme, > The sequel library that the eye-one sdk uses, has capabilities to set > calibration matrices, but I don't think these are exposed by the > eye-one sdk itself. I'll have to ask the question to X-Rite. > It would make far more sense > to add a correction in the driver to the XYZ values. Yes. No question. I was trying to think of the case when using other profilers. > Before jumping to conclusions though, how would you characterize > the errors between the DTP94 and the CS-100 ? What I've done in the past with the Canadian National Research Center was to simultaneously measure my Eizo CG21 with my CS-100 and their CS-1000. Dr. Réjean Baribeau then used Matlab to calculate a correction matrix to apply to the CS-100 measurements, yielding very low residuals. > If you scaled the > XYZ values to make the white points agree, how much in > error would other colors be ? I don't know and that's probably not the route to go, perhaps. Suppose I use dispcal to calibrate my Eizo CG221. Presently, this would only go as far as calibrating the monitor through the host video LUT -- not the monitor LUT. So, not a good solutino for this part of the work :( But suppose I use some other profiler, like ColorEyes Display, who have code to calibrate the Eizo in hardware. Then, conceivably, I could use targread to read the patches, with my CS-100, in some kind of manual mode, and then profile would build me the final 3DLUT monitor profile from my corrected CS-100 measurements! Bingo. Would that work? Roger Breton | Laval, Canada | graxx@xxxxxxxxxxxx http://pages.infinit.net/graxx