There is a "Uniformity Compensation" function and it really works.
Great!!!
Implications here are that Uniformity Compensation (UC) mode requires use of a "calibrated" mode, and you are forced to disable UC if you want to do a custom color temp (anything other than 6500) either via a non-ICC-mode preset, or via custom gain/bias.
Can't have your cake and eat it too ☹
Which leads to a theory about how it works: Panel zone variances must be measured at the factory and storied with the built-in cal data? (If so, it suggests that in principle local measurements could be uploaded into the device in a service mode? And that Dell's Xrite solution can maybe do this? Idk, but fun to think about. (In world of video, is device LUT loading standardized, open or proprietary?)
Seem to remember the specs on Dell's site mentioned a "SDK"? Everything may not be lost, after all. Your theory about factory calibration at a specified CCT makes sense.
To give you a sense of what I'm seeing I've added some screen grabs taken with an iPhone, which tends to exaggerate panel / color aberrations. https://drive.google.com/open?id=1DSpUzIAgxsiG8qi8FKr2uJ5_P4MNmTKu
I can see some differences from left to right ☹ I don't think you'd be able to improve on this difference.
When put in "Adobe RGB" mode, these two displays agree side-by-side as good as my custom cal. No trouble believing they hit < 2dE across board.
Do you mean a 2 DeltaE (whatever) in screen uniformity?
My gestalt at this time is that a designer who hopes to just unbox, set mode and work appears to get what he needs.
I would tend to agree with your gestalt. Many designers would never see the difference between the two screens or complain about the screen uniformity.
I may—or may not—take the time to measure and evaluate the ICC presets.
It will be worth your time.
I'm still climbing steep learning curve for these tools.
It's not for the faint of heart, I agree.
And I'm caught because I'm not sure how much I can trust this DTP94 in my setup.
Florian could best reply as I'm not 100% familiar with DispCal but isn't there some kind of 'correction matrix' to use a DTP94 on a flat panel display?
As I play with DisplayCal the repeatability (or lack of it) keeps me guessing.
It's entirely possible. You have to test this hypothesis but taking repeat measurements on the same screen area, off the same color, like RGB=128. Some variability might come off the display as well.
For example, I just found new instrument modes and compensation data are > available for download in DisplayCal since I last checked a few months ago, which is great! But adds to confusion. Need a more modern puck.
A modern puck would not hurt, or just, go for the gold, get an i1pro.
For now my choice is to run these as described above, enable UC, let DisplayCal set whitepoint and target sRGB. This appears to get me to 95% Display P3 and 98% Adobe, with an sRGB TRC for default and UC enabled. It's a compromise that looks good.
I agree. / Roger