- This topic has 30 replies, 6 voices, and was last updated 2 hours, 1 minute ago by
.
-
Posts
-
Hi Vincent,
Sorry it took me longer to respond to you than the others, I wanted to take the time to be thorough.
To answer your first question, yes, I use a CCSS to measure/evaluate/profile my reference displays. The Wacom Cintiq Pro 24” that I’m using as reference has an application designed by its parent company, Wacom, to specifically calibrate its panel. I use it to calibrate and profile the monitor to a locked-down sRGB clamp, which yields the colors that most closely match my Apple XDR screen when it comes to sRGB content/test images. The application is called “Wacom Color Manager.” The software manipulates the internal LUT with full control of the Wacom panel, while also pushing out an ICC profile to match. Both the proprietary calibration software and the probe, the Wacom Colorimeter, were in collaboration with Xrite years ago when they made the displays (and when I first bought mine). As a result, Wacom has full control over the hardware (the colorimeter) and the calibration software, along with the CCSS/edr said software uses. I trust this process, but if you have any reason to doubt it please let me know. So you can get an idea of the accuracy of that display, I’ve attached the latest calibration report.
For your second note about SpecPlot app, I decided to give it a go. I’m also attaching an image of the graphs produced, comparing the Calibrite Profiler/Xrite OLED CCSS correction (converted from the proprietary .edr file) to the latest CCSS correction from DisplayCal’s community supplied corrections. The community supplied file seems to line up more closely with the color peaks one would expect from a QD OLED panel. However, again, the Calibrite Profiler correction is the one that yields better results when compared to my reference. Most prominently, the greyscale in the low end is much better. After Calibrite Profiler’s ICC profiling, there is the typical 2.2 gamma human-eye visual crush of RGB values 0-5, ramping up to become gradually more visible crossing over from 4-6-ish. However, when profiling in DisplayCal, the greyscale ramp from the low end is way off. 0-2 are completely black to the human eye, while 3 is a sudden and jarring step up in brightness. Not sure why, but this always seems to happen in DisplayCal. This also used to be an issue in Calibrite Profiler, about a year ago, but that’s no longer the case. Whatever was causing that issue, Calibrite fixed it. When visually comparing greyscale test images on all my displays, they all align very well, assuming I’m using the Calibrite Profiler icc. However, again, the color red is desaturated on the QD OLED.
I will attach the ccss files used. Calibrite Profiler uses a rather old .edr correction file from Xrite for an old Sony OLED monitor from over a decade ago. I’ve converted this into a ccss to get a view of it in spotapp. The other file is a community-created ccss for this OLED model specifically. Calibration with it has been unsuccesful, but perhaps I’m still missing something…
Regarding your earlier comment about “sRGB primaries evaluation missing,” I’m not sure what you mean. Could you please clarify? When I was conducting my test, the brightness of my two displays wasn’t exactly the same. I believe one was closer to 200 nits while the other was 220 nits. However, I’ve since done a visual side-by-side with even closer brightness numbers and the issue remains. So yes, I can confirm have completed a visual test in the sRGB space, comparing primaries. And my eyes see them as two different colors. They appear to be the same hue, at least, but the QD OLED is desaturated in red by comparison.
Also, I’m certain that the RGB range is matched correctly. Gamma shifts never occurred to such an extent that 0-16 was crushed, for example. Again, my current issues don’t really sit with greyscale performance, it’s really just the reds at this point that I keep banging my head against. I can go through the process of “proving” the performance but I’m not interested in tacking on yet another problem to solve. If there IS an issue, it’s something I can deal with later. As for right now, greyscale looks great and matches excellently across all of my displays, so I’m not worrying myself with that at the moment.
Both you and Ben mentioned something about tweaking the ICC, or perhaps creating one from scratch based off of it, and then shifting the red value slightly, experimenting until the color matches visually to (what at this point I’ve considered to be) reference. How would you go about doing this? I attempted this with DisplayCal’s Synthetic ICC Profiler tool but I can’t find a way to dial in the white point, although I might be doing something wrong. I’m very new to creating an ICC from scratch, especially when trying to use a pre-existing ICC’s values as a base from which to start. I’m also attaching the synthetic profile I’ve created, alongside the original ICC from Calibrite Profiler (which still remains the most visually accurate result tested to date).
