Originally Posted by Slower
After opening some older calibration files in 3.4.2, I found some bugs not present in 3.4.1:
Selecting BT 1886, in near-black graph the target luminance goes to zero instead to actual black level for 0% white stimulus. Also noted by Dominic Chan in the above post, I now see.
Luminance and near-black 0% issue has been confirmed and fixed.
Saturation shift graphs look completely different: in 3.4.1 starting with 0% shift (as a result of near-perfect grayscale) to max +/- 5% for fully saturated primaries/secondaries. On the other hand, 3.4.2 starts with +/- 3% (even for near-perfect grayscale!!) with fully saturated green and magenta going to 20-50%. The resulting dE graph of course is also different compared to the one in 3.4.1.
The satshift graph issue has been confirmed and will be fixed in next update.
Sorry to bother you with this, but perhaps you missed my post #7657
about scaling ISO 12233?
Originally Posted by Dominic Chan
Here's a picture. I use a pattern size of 50%. The pattern (including background) seems to have the same aspect ratio, but only occupies the upper left portion of the screen.
I can't reproduce this on any resolution or display combo that I can access so it's not a generic problem. Will have to wait for more reports to see if some consistent reason can be discovered.
Originally Posted by carneb
I've attached my chc file. My Greyscale 50% is around 92 nits (the target 94) and the white window for primaries/secondaries is 90 nits.
Ok, so for your colorchecker measurements, some of the error you are seeing is because your display is not hitting it's PQ targets (98.735 nits
). PQ is an absolute scale and the targets are being calculated to that absolute scale, not your measured white.