Am 24.06.2011 02:59, schrieb Graeme Gill:
A thought that's crossed my mind (even before I heard about the X-Rite colorimeters) is that it would be possible to generically characterise various colorimeter models,
Hello Graeme, well, I'm wondering, how good is the inter-instrument agreement of the sensitivity curves? If it is good, then I would expect the matrices stored in different instruments (of the same model) to differ only by a multiplication by a diagonal matrix (-> representing different channel gains of the individual instruments). But is this really the case? Or do the the inter-instrument differences of the matrices rather have a magnitude as large as for instance the difference between the CRT and the LCD matrix (or the difference between the LCD and a wide-gamut LCD matrix) of one instrument? In the latter case, generic ccmx matrixes for an instrument-type/display-type pair are IMO becoming questionable as well. If a statistically representative quantity (say > 10..20) of user's of a particular instrument type would post the matrices stored in their instruments, then it would be possible to get a feeling for the actual inter-instruments differences, and what to expect consequently. Regards, Gehard