I found something interesting today thats worth opening a new thread. Credit goes to norbert.s in the german hifi-forums forum for instilling the thought in my head and being responsible for me trying it out eventually.
This concerns SDR (bt.709, ...) calibrations only.
Basically the idea is that OLED Light and Contrast shouldn't be lowered "in equal measures" from their starting position in cinema or ISF mode until you reach the desired luminance target of white, but that you could and maybe even should leave OLED Light at a high setting (max, or only slightly lower - will have to test different settings) - and instead set the luminance of white (Y target) pretty much exclusively with contrast.
Why this matters.
If you set the OLED to a Y target of around 100 nits (which is the most usual luminance target for calibrating for a dim environment), the other way around - lets say ending up at OLED Light 31, Contrast 74 - even calibrating a flat PLG 2.2 gamma will result in you loosing perfect black.
On the B6 this comes from mainly the blue corrections that are needed.
To explain -
OLED Light 31, Contrast 74 (= roughly 100 nits)
Will lead you to increase low IRE 2 step blue by 7, then make you lower 5 IRE blue in 20 step by a little to hit the target.
At that point you are loosing perfect black. (Tested on a B6.)
The tricky thing is, that this isn't measurable with an i1d3 (still measures ??? (infinite contrast)) - and your iris adjusts for it - so you can only test it in a dark room on dark patterns or scenes - but once you compare it to the calibration I'l talk about next - it becomes obvious.
OLED Light 100, Contrast 46 (= roughly 100 nits)
Will lead you to increase low IRE 2 step blue by 0 (also no decrease needed), then make you lower 5 IRE blue in 20 step by a little to hit the target.
So the way that doesn't lose you perfect black has you doing less irratic calibration steps as well.
There are other benefits.
First - at OLED Light 100, Contrast 46, APL gets engaged far less. Meaning average picture level (brightness) can be noticeably higher before APL kicks in.
Second - perceptual benefits. Call it color metamerism failure (from my understanding it has to be), call it whatever -
OLED Light 100, Contrast 46 and OLED Light 31, Contrast 74 calibrations for PLG 2.4 look very different.
OLED Light 31, Contrast 74 pushes out stranger colors - perceptively speaking. I will do a dE comparison in the future - (because the other way to explain this is simply dE diffference) but for now - targeting OLED Light 100, Contrast 46 gives you a more "natural" image.
OLED Light 100, Contrast 46 mixes in more white subpixel light - which has much broader spectrum cones - all in all lessening the metamerism failure issue. So that would be one explanation.
Calibrating for "high OLED light" might be worse in two ways though -
- although lowering Contrast doest have a negative impact on brightness clipping or color gradient patterns (at least I have perceived none (using 4:4:4 chroma with 1080p test patterns scaled on the PC side to 4k)), the overall image seems a little less "punchy" (contrasty) - but subjectively I prefer it to the image that results from a OLED Light 31, Contrast 74 calibration.
- Near black levels (first few steps) are more prone to crush on the OLED Light 100, Contrast 46 calibration - on the B6. Apparently (according tho the findings on hifi-forum.de again) higher tier 2016 LG OLEDs have less "indefeatable crush" (regardless of how you correct 5 IRE), than the b6 (lowest end model). So thats something to test as well.
Keep in mind - that I did this comparison with PLG 2.2 (and PLG 2.4), - first tests I made also indicate - that if you use any black compensation , you have to lower blue at 5 IRE even more, potentially increasing the "loosing true black issue" at lower OLED light levels even more. I can't say for sure because again - we cant measure it with the i1d3.
I would highly suggest, that you at least try one of your "best calibration presets" against the same calibation (target wise) with a "near max OLED Light and lower Contrast" calbration - IF you have calibrated anywhere near the 100-120 nits target.
Experiencing "true black" for the first time in a dark room alone is worth it. (Your current calibration might not give you true blacks - even though the i1d3 is telling you it does.)
Any comments appreciated.
In the future I will try slightly lower OLED Light settings (towards the 100 nits target) as well. edit: If you need a frame in between which to test - setting OLED Light to 80-100 is where the ABL doesn't kick in nearly as often.