2020-06-29 at 16:56 #25283
I have some questions regarding the measured gamma in the verification graph.
Should it be flat across the graph or having some deviation on x: 5 or x:10 is fine?
I mean I’m calibrating to Rec.1886 2.4 gamma (absolute), and the low end is a bit bright. Is there a way to improve gamma calibration? Maybe a longer patch sequence?
Using i1Display Pro Plus and Dell U2715H
2020-06-29 at 19:30 #25290
- This topic was modified 5 days ago by pyron.
Do not use Rec1886, it not meant for that kind of low contrast displays (less that 3000:1 or maybe higher). Use 2.4 instead, otherwise Rec1886 will work as intened by specificiation (google ITU Rec1886) and result in a lower gamma than 2.4
Also near black it will go to lower gamma values, no matter what you do, because of limited display contrast and keeping a minimum of brightness separation between greys. Your display cannot go darker there.
2020-06-29 at 20:51 #25292
- This reply was modified 4 days, 22 hours ago by Vincent.
Thanks for the reply. I though it had to do more with the black level rather than contrast ratio. Something around 0.03 cd/m2 in order to use absolute. Also the contrast ratio diminishes when you set the white level to 100 nits.
I will try to calibrate to 2.4 relative tonight and longer patches and see how it goes.2020-06-29 at 21:06 #25293
Yes, black level. But black level with typical grading setup is about 0.10 on a typical IPS. With other technologies with higer contrast black level is low at the same white level. Since typical white level is fixed fo these tasks, let’s say 100cd/m2 @D65 … it is actually a matter of contrast, not to black level. Static contrast ratio is a “window”, you fix white level => you get a black level, not the opposite.2020-07-02 at 15:29 #25330
Actually the gamma image was from the HDTV, here is the before/after of my PC display. As you can see increasing the patches to ~822 yields worse results. At 15 black gamma still doesn’t reach 2.4 on the new calibration.
Is there any tip to improve gamma? Would a Rec.1886 compliant display show a better measurements?
These are my settings:
2020-07-02 at 15:43 #25332
- This reply was modified 2 days, 2 hours ago by pyron.
On limited contrast displays gamma has to bend and go lower. Otherwise you’ll clip and you do not want it. This is the best you can get on a 1000:1 IPS.
2020-07-02 at 15:44 #25334
- This reply was modified 2 days, 2 hours ago by Vincent.
I understand, thanks for the help!