Black crush on OLED screens (or too dark to be humanly seen?)

[stanlawj]

Hi all, can you check if your OLED screen (laptop or monitor, including QD-OLEDs) is able to show the first 5 levels above black properly at standard 100cd brightness? All my LCDs can (I stare long enough to let my eye accustom to darkness), but not a single OLED screen I have can, even Ipad Pro 13" with Tandem OLED, no matter how bright I increase the screen.

Use this: http://www.lagom.nl/lcd-test/black.php
or this:

Note: some OLEDs are non-uniform, i.e. peripheral of the screen may be brighter or darker than the center causing the sequence of brightness of the squares to vary across the screen.

Because I thought this problem of black crush was been solved for OLED.
It looks like generally any level 5 and below is not possible for OLED? Or is it just really too dark to be discernable by eye?

This has implications for OLED M6 Macbook (Samsung Tandem OLEDs). I was thinking of postponing purchase of Macbook to OLED M6, until I realise my existing OLED Windows laptop screens seem to have also have crushed blacks that deceived me into thinking it has better blacks.

 

[stanlawj]

Ok, I managed to see the squares down to the 3rd black square of Ipad Pro Tandem OLED screen (1st and 2nd still indistinguishable)... but only close-up like 1inch to the screen in absolute pitch black with no reflection on the screen.

I can't distinguish them if there is any slightest surrounding ambient light bouncing off the glossy screen.

 

[Yongkit]

both my woled & qdoled can see every square ok, same for my notebook OLED

 

[stanlawj]

I found an article describing this problem of black crush in detail.
Yes, so the ambient lighting does affect the degree of crushing: the detectable luminance threshold.
Only the 3rd square is visible!

https://tftcentral.co.uk/articles/d...-understanding-and-testing-oled-shadow-detail

Target luminance curve, 2.2 gamma, 200 nits OLED screen
Above is a target luminance curve based on a 2.2 gamma and a 200 nits displays, showing just the first 16 RGB grey scale values from 0 (black) along the horizontal x-axis. The dotted orange line represents the target luminance for each grey shade if the black depth is a true 0.00 nits like it is on an OLED panel.

Target luminance curve, 2.2 gamma, 200 nits OLED screen
As we identified earlier, 0.01 nits is probably about the limit that you could perceive in a dimly lit room and so if you follow the line across at that point on the vertical Y-axis, you can see it’s RGB 3 that would likely be the first that’s visible when the screen is configured to this luminance (the exact target luminance is 0.0114 nits).

So, on an OLED screen like this, if the 2.2 gamma curve was followed accurately and exactly then the best case first shade visible would be RGB 3. Anything lower would be crushed to black! If the luminance was lower, like 120 nits for example, this shifts the target gamma curve downwards and the first visible shade expected would now be RGB 4 (0.0129 nits target). So if gamma is followed accurately, on an OLED screen the loss of some shadow detail is expected and intended!

-----
In other words, it's may not be possible to see the darkest levels depending on the room ambient lighting condition and screen calibration setting (gamma).

What a dilemma!!!

 

[stanlawj]

Conclusion: Does levels 1 to 5 have any significance? Should we completely treat them as black during photo/video editing, i.e. always push any wanted shadow details to be brigher above 10 to avoid OLED black crush issues from ambient lighting overload?

Frankly, I've never read or heard of anyone saying this before.

(Basically, due to gray scale levels are relative to black floor limit of the display, not relative to zero nits brightness)

 

[stanlawj]

both my woled & qdoled can see every square ok, same for my notebook OLED

Haha, based on my investigation and conclusion above, you shouldn't be able to see the first THREE squares if the black floor limit is zero for OLED and WOLED.

 

[Yongkit]

Haha, based on my investigation and conclusion above, you shouldn't be able to see the first THREE squares if the black floor limit is zero for OLED and WOLED.

are you sure you are understand what the article and black crush mean ? opps i forgot to mentioned those are in HDR mode lol.

 

[LiLAsN]

Technically, this test is hard to do in SDR due to the fact that most of you will probably increase the brightness way beyond the 100 nits industry standard for what the maximum brightness level of SDR is.

As a result, you will see some of you being able to see the squares clearly all the way down to the 1st or 2nd box.

But nevertheless, we are mostly using our PC in SDR until HDR is needed.

Use the video to use the 'Black Level' function on your PC monitor or TV to adjust the shadow detail so that the 2nd box should be visible very very slightly.

Or best is to just calibrate your PC monitor to ensure proper tracking of every single brightness point.
I can see a slight box outline down to the 2nd one in that video as intended as mine's calibrated.

So in the proper 100 nits for SDR, I can see a slight box outline down to the 2nd box as intended on my LG C2 OLED TV as mine's calibrated. It should appear very very slightly as the box should still be below 1 nit even.
This is with my camera increasing the exposure to show that it should at least have an elevated black for the box itself as it is above 0 nits but still below 1 nit.

 

[Yongkit]

I try 100 nits calibration before and I do not thinks it suitable for gaming since its my main focus getting OLED so I have it calibrated to my own preference just as screenshot of mine showing clear border while still maintain the black as those side panel as black as possible not raise black to interfere the test pattern.

 

[LiLAsN]

I try 100 nits calibration before and I do not thinks it suitable for gaming since its my main focus getting OLED so I have it calibrated to my own preference just as screenshot of mine showing clear border while still maintain the black as those side panel as black as possible not raise black to interfere the test pattern.

Yes, it's normal for many people to go above 100 nits. So don't worry!
For me, I actually half my brightness from 100 nits as I don't like the bright whites when on websites with white backgrounds. It's too bright for my eyes in my light controlled room.

For games, I prefer to play in HDR because the HDR impact is just wow. Immersive.

But for daily PC use, I prefer to lower my brightness.

 

[Yongkit]

Yes, it's normal for many people to go above 100 nits. So don't worry!
For me, I actually half my brightness from 100 nits as I don't like the bright whites when on websites with white backgrounds. It's too bright for my eyes in my light controlled room.

For games, I prefer to play in HDR because the HDR impact is just wow. Immersive.

But for daily PC use, I prefer to lower my brightness.

on OLED contrast affect those mostly for this logam test, and it is easily confuse between the "brightness" and "contrast" setting in the OLED TV & monitors because it seems to have similar effect but different purpose on picture quality presented.

which is why i am double check if OP understand correctly for the single article he shared.

 

[stanlawj]

on OLED contrast affect those mostly for this logam test, and it is easily confuse between the "brightness" and "contrast" setting in the OLED TV & monitors because it seems to have similar effect but different purpose on picture quality presented.

which is why i am double check if OP understand correctly for the single article he shared.

You mentioned your screen is in HDR mode? That may have changed how the colours appear. My Win PC and laptop screens are calibrated with the i1 Display Pro colorimeter to 100 nits + Gamma 2.2 in SDR. (Ipad Pro can't be calibrated, but obviously Apple is assumed to have done a good job).

But thanks for sharing. I confirm at least the 2nd or 3rd square should be visible (provided in total darkness), at least it shows on the Ipad Pro.

Technically, this test is hard to do in SDR due to the fact that most of you will probably increase the brightness way beyond the 100 nits industry standard for what the maximum brightness level of SDR is.

As a result, you will see some of you being able to see the squares clearly all the way down to the 1st or 2nd box.

But nevertheless, we are mostly using our PC in SDR until HDR is needed.

Use the video to use the 'Black Level' function on your PC monitor or TV to adjust the shadow detail so that the 2nd box should be visible very very slightly.

Or best is to just calibrate your PC monitor to ensure proper tracking of every single brightness point.
I can see a slight box outline down to the 2nd one in that video as intended as mine's calibrated.

So in the proper 100 nits for SDR, I can see a slight box outline down to the 2nd box as intended on my LG C2 OLED TV as mine's calibrated. It should appear very very slightly as the box should still be below 1 nit even.
This is with my camera increasing the exposure to show that it should at least have an elevated black for the box itself as it is above 0 nits but still below 1 nit.

So, one of my laptop OLED screen appears to be wrong and need to be recalibrated. The first 8 black squares are totally black.

The thing is that, any slight reflections off the glossy OLED screen basically makes the first 5 squares invisible at 100 nits brightness. Which means, for normal viewing environment, user have to increase beyond 100 nits or do some magic setting to lift shadows (which makes it non-standard).