Guys, perhaps you can help me out.
I keep having some issues with gamma calibration of my tv using HCFR with a Chromecast as pattern generator. Despite my efforts somehow the results differ when I use the automatic measurement of greyscale & gamma (F2 'run greyscale') vs a manual routine (by clicking the 5, 10 etc cells and pressing "measure" every time) with the exact same settings. Below the details.
Chromecast 3 connected via HDMI2
i1 Display Pro (latest version) with display type WRGB OLED (LG B7 2017)
TV: LG C8 OLED
TV service menu: TPC (ASBL) off
TV settings: ISF Expert dark mode, BT.1886 gamma, things like energy saving, eye comfort, motion eye care all off
20-step greyscale, BT.1886 gamma
10% window pattern, latency 300ms
delta E CIE2000 with absolute Y w/gamma
When I perform a manual
20-point greyscale measurement (which takes me about 100 seconds) the result looks like this:
If after that I immediately perform the same measurement with the exact same settings in automatic
mode (60 seconds) the result is different, with a distinct stepping pattern. What you see is that output is going down in 4 steps and then shoots up again in 1 step (f.e. IRE 65 -90). As you can see this is a repetitive pattern.
Here is the result in one picture, with the solid yellow line being the auto mode and the dotted yellow line the manual mode:
Measurement data with the xyY values are available in the attachments for reference.
This process is very repeatable: every time I perform a manual measurement the result is different from the automatic mode with the latter showing that stepping pattern.
I enabled the beep with every measurement so I can clearly see that the pattern is already present on the tv some time before the measurement starts. So no issues there.
I also did a measurement using my pc connected via HDMI to rule out any problem with the chromecast, and there too the same stepping pattern is visible in auto mode.
The ONLY difference I can think of is that in manual mode the chromecast returns to its opening screen shortly between each measurement, while in auto mode the windows are shown without interruption. I tend to think the manual mode is the most correct one, since in this mode the picture is reset in between every measurement and the pixels are heating up less in the process of measuring increasing brightness. I noticed that due to pixel heating the brightness is going up the longer you leave a pattern on the screen, so especially in the higher IRE range it is better to perform measurements quickly and then return to a black or dark grey screen.
It seems as if somehow something does not reset itself correctly in auto mode, although I went through the trouble of disabling every related setting in the normal and service menu of the tv.
I even tried disabling any possible output limiter by pressing the SETTINGS or VOLUME button of the tv at every single measurement step of the auto mode, to prevent any brightness limiter kicking in. But as could be expected in a total measurement time of less than 60 seconds, this did not make a difference.
Does anyone know what is going on here? Of course I would like to rely on auto mode but I can't since the results do not seem correct. You could argue that the result in auto mode is still valid since the differences are small and near gamma 2.4, but I think this stepping should not occur.