Ok, I'm backI think a large part of the problem here is that I'm having trouble getting a consistent white level between my i1d2 and screen.
Post the measured L*a*b* values for white and blue.107.79 9.21 -24.89 56.97 31.62 -101.35
When I do your testchart measurement now, I'm seeing white as: 117.36 -17.01 -16.42If you look back, for first matrix profile I made, I set it to 100 because the screen was measuring IIRC ~115 area or something at the time. When I made the XYZLUT version, the screen was measuring ~100. I remember noticing this discrepancy at the time but didn't really think anything of it.
When running the machine normally, I have the screen brightness set to 'auto' which in practice means it hovers around 50%. When calibrating (or testing calibrations), I always disable the 'auto' option, set the screen brightness to its maximum level and then wait a bit for things to equalize. I do this under the theory that when I need to do something where color accuracy is important, I can get back to a known good state easily.
However.. I'm not sure this is as 'known good' as I thought. When I first max out the screen and do a 'report on uncalibrated display' I get around 160 cd/m^2, which seems to fade slowly over time. In the time it's taken me to run the chart again, re-verify all my settings and write out this email, it's dropped from ~158 to ~146.
I was under the impression that unlike traditional CRTs, LED backlit displays only needed like five or ten minutes to "warm up"... am I wrong? Do I need to wait an hour as with CRTs or is the screen on this machine just being weird?
______________________________________ it has a certain smooth-brained appeal