2023-01-08 at 22:12 #38471
Hello, I am attempting to create a 3D LUT for use with DWMLUT_GUI to color correct my display (LG 27GN950-B). After the research I’ve done, it seems like the optimal target is 6500k sRGB, as I really only use my PC for gaming, and most games dont use any of the larger color spaces. The way I understand the process for creating a 3D lut, I should be selecting sRGB in the “source colorspace” box, and gamma 2.2 in the tone curve box, however sRGB doesn’t appear in the list. The profile does exist in the arygyll cms folder I downloaded (version 2.3.1), but displaycal just doesn’t see it. Is this a bug or am I misunderstanding the process? I have already tried reinstalling displaycal, and also redownloading the argyllcms folder.
Using a colorchecker display pro if that is relevant2023-01-08 at 23:30 #38474
May have figured this out, only difference between sRGB and REC.709 is gamma 2.2 vs 2.4 respectively, so setting REC.709 with 2.2 gamma curve ≈ sRGB2023-01-09 at 16:14 #38477
May have figured this out, only difference between sRGB and REC.709 is gamma 2.2 vs 2.4 respectively, so setting REC.709 with 2.2 gamma curve ≈ sRGB
Depends on what you call sRGB, the official version uses a specific TRC. The approximate one that follows a pure 2.2 gamma curve is what your settings will get, and probably the most used.
2023-01-09 at 20:14 #38481
With an IPS display it’s very likely that you want 2.2 for games, not sRGB TRC. Same with 2.4 for movies.
TRC shape (whatever it is) is going to be lifted/deformed to preserve tonal separation near black (alternative to that would be clipping and 99% you do not want that), hence ypu’ll end with too much bright dark greys that you expect, although in validation (black shifted) it goes all OK.
Use 2.2.2023-01-09 at 21:35 #38485
Good to know, would explain the strange deviations I was seeing in low brightness.
So target colorspace of rec.709 with gamma 2.2 is optimal for just games? Or is there another set of parameters that would lend itself better to my needs?
This display does seem to do good coverage of DCI.P3 and adobe RGB, but as far as I can tell those are rarely used in anything other than professional editing/mastering software.2023-01-09 at 22:43 #38490
no, use that: srgb/rec709 primaries and 2.2 gamma black output 100% (relative)
Edir: Same if your display had some kind of HW calibration: srgb primaries, 2.2g
2023-01-10 at 1:58 #38492
- This reply was modified 2 weeks, 6 days ago by Vincent.
srgb/rec709 primaries and 2.2 gamma black output 100% (relative)
Hi Vincent, btw how exactly does it shift output with 100% output offset?
Do we get luminocity ^ 2.2 + shift ?
And for 0% output offset we get ( luminocity + shift ) ^ 2.2 ?2023-01-12 at 16:25 #38516
Try to see it as a 2D plot, Y = gamma value per input, X input.
A true 2.2 will remain flat at 2.2=Y… and it will need infinite contrast to do that.
On a finite contrats display calibration too will aim for Y=2.2 for most of the X plot, but at or near 0 it cannot be 2.2 because it has some A nits brightness.
So you can keep 2.2 while you can, then clip (not advised)
keep 2.2 while you can, and when you cannot do it “bend” that 2D graph to lower and lower gamma as you approach to 0 input.
The “bending” can be more or less agressive, depending on your intent. DisplayCAL will aim to preserve gery neutrality & brightness separation and record (if you choose s) actual TRC in ICC profile because for color managed apps it is the sensible approach => this way you can rebuild whetever TRC you want to track at different settings because grey is neutral and dark grey steps are as visually difefernt in brightness as you can for that TRC reference.
It is no longer constant 2.2, see your own TRCs in measurement report or profile information.2023-01-12 at 18:33 #38527
Would it be fair to sum up the issue with sRGB TRC on non infinite contrast displays is the compound compression near black?