Nice work with the i1d3. Thanks for the new version. :) So, you say the expected error which comes from the imperfections of the filters is always dE2000<5 with your test sample? I guess it means we can assume <2.5 as the highest expected error on the particular display types when they are measured with the appropriate "general correction"s (LCD mode for LCD, etc...) This makes me think I shouldn't even try to generate corrections with my ColorMunki. It isn't a perfect sensor either, I guess it could also show a similarly sized error if I could compare it to a recently validated labor quality spectrometer. Doesn't it? (I didn't buy an i1d3 yet to test their differences for myself. I waited for ArgyllCMS support.) By the way, do you think the OEM version of the sensor (which you can buy from SpectraCal and other third-party retailers) should also work with this version or would it require additional work on the ArgyllCMS driver? Didn't you think about making a tool which can create custom corrections for spectros like CM and i1Pro? This way I could at least compare my CM with a freshly re-certified i1Pro (or may be even something "bigger" - I guess the university where I learn has a lab spectro somewhere and it would only take a bottle of wine or box of cigarettes to play with it for some minutes. :D).