- This topic has 10 replies, 2 voices, and was last updated 2 years, 4 months ago by
.
-
Posts
-
Good day! I am trying to profile my Macbook pro 16 (2019 PFS) diplay to use Davinci resolve SDR rec709 gamma 2.4 100nit So, I have Displaycal software (for profiling and 3D Lut) and I use CS Zro for verification. After hundreds of possible combinations of parameters in DisplayCAL, I picked up such parameters (see screenshots) everything is wonderful with regards to colors, my memorable colors, color coverage, skin tones are all in place. Grayscale is also within normal limits, but I constantly encounter clipping in black and do not understand how to avoid it. Below are screenshots of CS verification. I tried black point compensation, I tried black shift, and so on it did not solve my problem … I tried to measure the black point before profiling and deliberately understand it by 3-5 divisions (I understand that this is nonsense) but this also did not give me results either do you need to increase this value and sacrifice contrast? What actions can I take to improve this performance? It is clipping in black! Once again, I repeat that all primaris colors are in their places. Below screenshots.
I used iProfiler but the results are much worse, it can’t even get into D65…. And yes I forgot, my device is an i1Display pro
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.And yes, if I enable black point compensation in the profiling, it does not get rid of clipping, but it brings chaos in the RGB balance closer to black starting from 0.4 – 0.0 in the CS chart tab CLIP
Vincent, can you please elaborate a little bit more. I understand that I have to use a relative curve? And that clipping is inevitable on this type of display?
And that clipping is inevitable on this type of display?
I’ve not said such thing. Read again.
LUT3D creation: relative TRC.
Also you cannot correct black color without destroying contrast on an additive display, and you cannot track some power law gamma curve down to 0 with a finite contrast display (and very small value 1500-1000:1 BTW).You just have to deal about what will be your black&near black reference on a 2.4 power law once you go close to display non zero output black: this is not black but it will be my reference black.
I’ve wrote it before.
Just keep in mind that when cross verifiying with different apps, validation settings must aim to the same. Like for example in HCFR, otherwise in reduced contrast scenarios ideal reference value may be darker than measured. IDNK about CS Zro valitadion settings equivalent to a relatuve 2.4 gamma with 100% black output offset, maybe it validates that way by default, or try to detect that on the fly (HCFR don’t AFAIK). So beware of that, an error reported by an app may not be an error at all.
It’s very hard for me to understand what you mean and I don’t understand what these abbreviations mean) I understand that it may not be a bug at all, but besides verification by software I use test images on the timeline and I see clipping in the blackest. I also tried using icc profile as a color control but it was even a bit worse in shadows as opposed to LUT. Can you give me recommendations on how to programmatically solve the bug and improve the black? How to set the black point correctly? Should I measure it before profiling and set the values manually instead of leaving it by omission (as measured)?
And that clipping is inevitable on this type of display?
I’ve not said such thing. Read again.
LUT3D creation: relative TRC.
Also you cannot correct black color without destroying contrast on an additive display, and you cannot track some power law gamma curve down to 0 with a finite contrast display (and very small value 1500-1000:1 BTW).You just have to deal about what will be your black&near black reference on a 2.4 power law once you go close to display non zero output black: this is not black but it will be my reference black.
That’s what I meant by saying that I should intentionally set the black point higher then it will set the shadows better (kind of like a hardware calibration where we make profiling easier for programs?)
It’ll try to keep the more contrast you can, it’s never enough on an IPS. Instead of focusing in RGB balance just focus on dE it should be negible after LUT3D unless is display is bad behaved. Suggested configuration for an 1000:1 IPS it’s written in example above.
Then double chck validation config in CS Zro is equivalent to LUT3D configuration in DisplayCAL. Otherwise CS Zero valiodation would be meaningless.
