- This topic has 26 replies, 3 voices, and was last updated 8 years, 6 months ago by
.
-
Posts
-
Hi
i went through a couple of testpatterns for colors and i noticed that the Blue Color has a Magenta shift to it in the higher saturations/amplitudes with the 3D LUT.
i checked all of my 3D LUTs from some time ago and all of them has this magenta shift.
Its not there if i disable the 3D LUT so its not the Monitor that is wrong.The Monitor 500M has some slight high dE for just blue color (dE 2.0) that cant be fixed with the standard cms settings.
i just wonder what can cause this magenta shift?
does the 3D LUT compensate that slight error wrong or is it the testpatterns that doesnt have enough blue color patches?or can it be the i1 Display Pro meter that reads wrong?
if i had to use a setting for color patches i would set it so it doesnt change or measure blue color at all
is that possible to make?here is a screenshot i made in PS that shows the problem
it pretty much looks like this on the tv with a visible line were the magenta shift starts.
http://privat.bahnhof.se/wb192876/magentaerror.jpgoriginal testpattern
http://privat.bahnhof.se/wb192876/TZTF60.pngEdit: its only blue color that has this problem not the other colors
ti3 file if that can be of any help
http://privat.bahnhof.se/wb192876/madVR%202015-09-13%200.3127x%200.329y%20S%20XYZLUT+MTX.ti3- This topic was modified on 2015-10-10 14:31:33 by patrik10.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Hi,
The Monitor 500M has some slight high dE for just blue color (dE 2.0) that cant be fixed with the standard cms settings.
If the monitor gamut is smaller than Rec. 709, then out-of-gamut colors cannot be displayed and will thus be clipped by the 3D LUT.
[ Is that a plasma? It seems you didn’t use a colorimeter correction ]icm
Please attach the display profile as well, thanks.
yes blue color is slight lower than the rec 709 target but not much
i dont have this magenta tint without the 3d lut so it must change it not only clip those levels.if i use the plasma colorimeter correction with the i1 Display Pro i get a green cast to the picture so its not accurate.
where can i find the display profile?
here is the result without the 3D LUT
(the blue color result here is the same with the wider gamut i use for the 3D LUT)
http://privat.bahnhof.se/wb192876/AdvancedColorManagementReport_KRP500M.pdf- This reply was modified on 2015-10-10 15:09:04 by patrik10.
is it this file you want?
http://privat.bahnhof.se/wb192876/adjust.log.2015-09-13i dont have this magenta tint without the 3d lut so it must change it not only clip those levels.
There’s no way to change the out-of-gamut points. Only points within the gamut can be changed by the 3D LUT.
if i use the plasma colorimeter correction with the i1 Display Pro i get a green cast to the picture so its not accurate.
That seems unusual.
where can i find the display profile?
The other file with .icm extension in the same folder.
here is the result without the 3D LUT (the blue color result here is the same with the wider gamut i use for the 3D LUT)
That report doesn’t tell what delta E formula is being used (1976? 2000?)
- This reply was modified on 2015-10-10 15:54:29 by fhoech.
the plasma color correction file for the i1 Display pro is a no go.
if anyone is using it with their plasma tv they will have the wrong whitepoint with a greenish looking picture.CIE94 is used
ok it must be this one then
http://privat.bahnhof.se/wb192876/madVR%202015-09-13%200.3127x%200.329y%20S%20XYZLUT+MTX.icmone question
there is supposed to be one Correction TXT file.
its around 16mbi have searched but i cant find it
do you know where its saved?Edit: solved
i had ot create a 3d lut for eecolor- This reply was modified on 2015-10-10 18:28:19 by patrik10.
Thanks, the profile looks fine with respect to the measurement data. I re-created the 3D LUT with the same settings you used. Looking at the TZTF60.png test image you provided through the 3D LUT using madVR and MPC-HC (see attached screenshot), I don’t actually see any banding or abrupt color shifts, which would indicate the problem in your case lies elsewhere.
there is supposed to be one Correction TXT file. its around 16mb
16mb would be way too large (typo?). 16kb is more like it. The correction files are normally located at C:UsersYourUserNameAppDataRoamingArgyllCMS. You can also hover the correction dropdown with the mouse to get a tooltip with the location of the selected correction.
I looked at it more closley using Teds calibration disc and the pattern for Blue Amplitudes and Saturation.
Its around 45% to 50% Amplitude i have these magenta bands instead of blue with the 3d lut.
Without the 3d lut its completely even steps.Its almost feels like a bug with madvrs pattern generator as its just at those % i get a magenta tint.
Its the same with the 9200 test chart i have.
I was using a 3200 patches chart here.How can you check that all patches are where they should be with the pattern generator ?
Its almost looks like dispcal thinks these amplitudes are for magenta color and adjusts it to that instead of blue.
What i meant was are you sure that the patches are synced and lined up properly or can it be one or two that is off ?But ok if you cant see anything wrong using this 3d lut i guess something must have knocked the signal off.
I have seen problems before that suddenly pops up with the 3d lut and that was after i reseted the tv to get back the low blacklevel.
Then the near black grayscale had suddenly turned blue.I guess it must be failures or shifts somewhere in the signal here.
I will do another run with 3200 patches.But ok if you cant see anything wrong using this 3d lut i guess something must have knocked the signal off.
I see a very slight band going vertically through the origin point of the gradient (not nearly as strong as the banding in your screenshot), but that’s about it. Still, only you can ultimately tell how it looks on the particular display in front of you. What should definitely be the case though is that madVR output on your system with the 3D LUT from the profile you posted should match the screenshot I attached (when viewed in a not color managed application like MS Paint).
What exactly does that plasma EDR correction do, Florian? Does it work under assumption that all plasmas are born the same? Or is it something else?
It’s hard to know for sure because I think X-Rite has not published the Plasma TV model(s) that the EDR was created from. But it’s possible that the EDR is a better match for Plasma TVs with more CRT-like phosphors, while the Kuros seems to have phosphors that allows them to reach a gamut somewhat closer to DCI P3.
i can see it now that the screenshot i made was overdone
its more close to what you have on your picture with just a slight magenta band.i was able to get the magenta tint on the camera from the tv and this is how it looks like
with 3D LUT
http://privat.bahnhof.se/wb192876/3d%20lut.jpgwithout 3D LUT
http://privat.bahnhof.se/wb192876/no%203d%20lut.jpgi was reading the Madvr ArgyllCMS thread on avs and there is one user “Masharak”on the last page that got this pink band at 50% on his grayscale.
also using an I1 Display Proi have not checked the grayscale yet but can it be a bug with madvrs testpattern generator that causes these pink/magenta tints around 50% ?
can you see from my 3D LUT why it corrects 50% amplitude for Blue color to magenta?
- This reply was modified on 2015-10-12 03:17:43 by patrik10.
Colorspace 1 is Pioneers native colorspace and its around 130% of srgb and 90-92% of Adobe RGB.
These Kuros are the only Plasmas that i know about that gets this wide gamuts.
Panasonic claimed that they could go wider than srgb with their ZT and “new” phosphors but it turned out to be just a marketing gimmick.i would really like to do a 3D LUT from this native colorspace but i have tried several settings and it allways reduces the 3d lut to rec 709.
but thats OT here 🙂i have some Sony Mastered in 4K movies that uses xvcolor
but to get the best colors form those i cant use the 3D LUT as it is now.
i just use Pure Mode with color space 1 for those
