- This topic has 3 replies, 2 voices, and was last updated 6 years, 8 months ago by
.
-
Posts
-
- Offline
its damn impossible to get a 3D LUT with proper result on a 9G KURO!
the 3D LUTs dEs is WORSE than what i get stock from the 500M.
Colorluminance is all over the place
displcal reports over 125% of srgb gamut coverage but the final 3d lut only covers 96%?
also its impossible to use fullscreen with madtpg becase it looses signal in the begining of the profiling.
and if you dont know how a kuro looks when it doesnt get a signal i can report it
blacklevel rises above LCD level for a few seconds before panel drive kicks in
the problem here is dispcal continues to take reading when that happends.
also as i have wrote about before
with the newest version of dispcal it throws in random white readings after 2-3 patches
those white patches are NOT in the testchart
where the hell do they come from and why is it only happend with the newest version?
i think its time to get to the bottom with all these problems.
is the problem MadTPG or Dispcal?
files
http://privat.bahnhof.se/wb192876/madVR%202017-08-22%2018-52%20S%20XYZLUT+MTX.rar
testchart that i use
We already discussed this.
- Offline
its not that
i now use an older version of madvr from april this year.
and i have also checked the window area where the meter is and there is no mouse pointer blocking anything
it seems maybe that the testchart i was using was broken? (how now that can be possible?)
i took another one that had 9200 patches and reduced them to 3000 and did another run
this time color coverage was as is should around 99% of srgb
also color luminance was better for all primary and secondary colors except blue
still blue colors luminance is up +15% and thats higher than what i have stock
grayscale is the same
Dispcal cant correct grayscale from the 3d lut even with 256 grayscale step measurements
i still get dE of around 1.0 to 2.0 at IRE 10 to 30
thats worse than the stock dE from the 500M
why is it that so fine measurements of the grayscale cant be fixed to reference?
it seems maybe that the testchart i was using was broken? (how now that can be possible?)
No, that’s not possible.
why is it that so fine measurements of the grayscale cant be fixed to reference?
It requires a stable display. If the display drifts or ABL kicks in you have no chance. White level drift compensation will help some, as will reducing the number of measurements, and making sure the test patch area is as small as possible with either 20% APL or black background.
