Home › Forums › Help and Support › Gamma tracking OFF with 128 neutral patches but not with 64
- This topic has 9 replies, 2 voices, and was last updated 4 years, 4 months ago by Patrik Gardewall.
-
AuthorPosts
-
2019-12-06 at 13:16 #21470
strange problem with the testcharts here
when i select 64 neutral patches for the testchart gamma after profiling on my KPR-500ALE is spot on
with 128 or even 256 neutral patches gamma is wrong and way too low on the high IREs.
you could almost think that the program calculates the wrong video levels.
i also get inconsistent grayscale and rgb balance with some testcharts.
is the program still after 3 years? broken and bugged out?
these problems wasnt there with my 500M that i profiled in 2017 with the old verisons of dispcal.
questions then is where can i download the old verisons of dispcal from around 2017?
- This topic was modified 4 years, 4 months ago by Patrik Gardewall.
- This topic was modified 4 years, 4 months ago by Patrik Gardewall.
Attachments:
You must be logged in to view attached files.2019-12-06 at 13:25 #21475also as you can see i get worse RGB grayscale with dEs around dE2.0 with the testchart with 128 neutral patches.
The higher dEs for Blue color is normal as it is the limits of the tv
nothing is changed on the tv with the two testcharts
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
2019-12-06 at 13:55 #21481This is display or instrument drift. As this is a Plasma, I strongly recommend white level drift compensation and a small patch size with ~20% ABL background.
2019-12-06 at 14:38 #21484it must be instrument drift then
its an i1D3 but its a couple of years old (7 or something)
i can see from the testchart that peakluminance is different but each time i have checked its stable at 112cd/m2.
(cheked from the calibration window at start before you start the profiling that is)
window size is at 12% so no ABL is triggered here.
what is the default integration time for an i1D3 in dispcal?
can you increase it ?
i know that 1 second is a must for stable readings but i have no clue what it is set at here.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.2019-12-06 at 15:09 #21485i have done some more tests and it seems that you have narrowed down intergration time?/number of measurements per patch or measurement smoothing over older versions of Dispcal
why?
the program only becomes more unstable.
i know this is an fact because profiling goes alot faster now than it did before with the same testcharts.
how do i get back stable readings with Dispcal and I1D3 again?
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
- This reply was modified 4 years, 4 months ago by Patrik Gardewall.
2019-12-06 at 15:11 #21488The instrument or integration time is not the issue. Display drift is the most likely cause. You can counteract that to some extent with white level drift comp and limiting the number of measured patches.
2019-12-06 at 15:12 #21490this is an KRP 500
its an reference display it doesnt drift like new OLEDs does.
so anwer the question instead.
why speed up the program to make it more unstable?
2019-12-06 at 15:15 #21491Loose the piss poor attitude, otherwise I’ll just stop trying to provide help to you. I’m already getting annoyed.
2019-12-06 at 15:28 #21492For the record:
The way to figure out issues caused by changes in your setup is by a process of elimination. You are using an old Plasma, you’re using madTPG. That should be a starting point. I can already guarantee you that changing around DisplayCAL or ArgyllCMS versions won’t solve your issue, so that should save you a lot of time. Good luck.
2019-12-07 at 10:30 #21502its seems that the problem is with the verification not the final 3D LUT.
i have checked grayscale tracking and gamma in other programs and its great.
but the verification measurements are unstable for some reason.
-
AuthorPosts