- This topic has 325 replies, 56 voices, and was last updated 2 months ago by
.
-
Posts
-
Thanks for the best calibration application.
This is the display characteristics before and after calibration of INNOCN 27M2U (HDR mode) using a DWM LUT and applying a calibration 3D LUT (cube). (EOTF, RGB balance, CIE uv saturation mapping)
It cannot be used for grading before calibration, but after calibration, it can be used sufficiently for grading HDR footage.
* When watching video or playing games, I applying a LUT with luminance map, because the high luminance side saturates quickly.The calibration processing is as follows.
1…Created CCMX colorimeter correction matrix from DispCal. I used i1Pro2 for reference and i1D3 for colorimeter.
2…I used DispCal’s visual white point editor to set a white point that perceptually matched the D65 target.
3… While correcting i1D3 with CCMX, measured 100 point grayscale with HCFR (the white point is the same as the target of DispCal), and make 1D correction LUT(primary) from the measured values.
4…Create a 4000 point measurement patch for DispCal, measure 4000 points with the 1D correction applied via DWM LUT, and create a 3D LUT for saturation mapping correction.
5…Combined the DispCal 3D LUT and HCFR 1D primary.
(I used DaVinci Resolve for combinate the LUTs)
6…Measure 100 points again on HCFR with applied the combination LUT, and perform 1D correction(secondary) from the measured values.
7…Combined the 1D correction(secondary) and the 3D LUT(combined HCFR 1D(primary) and DispCal 3D LUT) by DaVinci Resolve.
8…apply the combined LUT(HCFR 1D secondary-DispCal 3D-HCFR 1D primary)
(I used DaVinci Resolve for combinate the LUTs)1D correction from HCFR measurement values uses my original Excel sheet to create a 1D cube for DaVinci Resolve.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Hello neldrip, thank a bunch for sharing your experiences. I’m a novice looking to learn from experience and tricks used by others
I used DispCal’s visual white point editor to set a white point that perceptually matched the D65 target.
So what did you perceptually match the monitor to? Another monitor? Did you do the tweaking via OSD on the monitor?
While correcting i1D3 with CCMX, measured 100 point grayscale with HCFR (the white point is the same as the target of DispCal)
Haven’t yet had a chance to play with HCFR. Does it have some automation to do that? Or was it all manual and very time-consuming? Am I right to understand that DisplayCal doesn’t have ability to create 1D LUT-s? Is there an advantage to doing a 1D LUT and then building a LUT on top of that? Did you find it necessary to make any tweaks from monitor’s OSD at this stage?
Thanks again for sharing!
Am I right to understand that DisplayCal doesn’t have ability to create 1D LUT-s?
A 3DLut can easily contain a 1DLut, just use a matrix profile as a base.
So what did you perceptually match the monitor to? Another monitor? Did you do the tweaking via OSD on the monitor?
Calibrate another non-wide-gamut standard monitor to D65 and use it as a target for perceptual white point matching.
Please refer to the following Lightillusion article.
https://www.lightillusion.com/perceptual_match_guide.htmlHaven’t yet had a chance to play with HCFR. Does it have some automation to do that? Or was it all manual and very time-consuming? Am I right to understand that DisplayCal doesn’t have ability to create 1D LUT-s? Is there an advantage to doing a 1D LUT and then building a LUT on top of that? Did you find it necessary to make any tweaks from monitor’s OSD at this stage?
It’s not automated.
Surely, dispCal can also create a 1D-only LUT, but HCFR provides better EOTF reproducibility and a better white point matching than DispCal.When measuring for 1D correction from a calibrated state (with luminance map), HCFR outputs a LUT with almost no change like this.
By the DispCal, It seems strange in the low luminance region and the high luminance side and i dislike it.As for the number of measurement points, the standard setting of DispCal has only about 30 measurement points for grayscale, I think there is a problem with accuracy.
If you create a measurement patch, you can measure 250 points, but even if you use that, you can’t get an accurate output. (See attached SS)Calibrate another non-wide-gamut standard monitor to D65 and use it as a target for perceptual white point matching. Please refer to the following Lightillusion article. https://www.lightillusion.com/perceptual_match_guide.html
Metameric Failure is a very difficult topic and I don’t know of a good reliable solution.
But I’ve seen differences even when comparing two small gamut displays adjusted to D65.
One a LED LCD with about 89% Rec.709 coverage.
And another a CCFL LCD with about 70%-80% Rec.709 coverage. And both looked visually different.
Even when I used a 2nm Jeti 1501 probe.
So which display to use?
There might be something else rather then gamut causing metameric failure. But I don’t understand what.
It’s not automated.
Surely, dispCal can also create a 1D-only LUT, but HCFR provides better EOTF reproducibility and a better white point matching than DispCal.When measuring for 1D correction from a calibrated state (with luminance map), HCFR outputs a LUT with almost no change like this.
By the DispCal, It seems strange in the low luminance region and the high luminance side and i dislike it.Because of your choice of paramsin ref TRC + black output
(See attached SS)
You are correcting WP on the 2nd one (DisplayCAL?). Don’t do that = if you use a visual whitepoint match, you cannot perform LUT3D calculations against a Rec709 D65 or equivalent colorspace. You’ll have to go to relatieve colorimetric way or absolute vs equivalent colorspace with numerical coordinates of your “visually matched white”
Calibrate another non-wide-gamut standard monitor to D65 and use it as a target for perceptual white point matching. Please refer to the following Lightillusion article. https://www.lightillusion.com/perceptual_match_guide.html
Metameric Failure is a very difficult topic and I don’t know of a good reliable solution.
But I’ve seen differences even when comparing two small gamut displays adjusted to D65.
One a LED LCD with about 89% Rec.709 coverage.
And another a CCFL LCD with about 70%-80% Rec.709 coverage. And both looked visually different.
Even when I used a 2nm Jeti 1501 probe.
So which display to use?
There might be something else rather then gamut causing metameric failure. But I don’t understand what.
If you are dealing with your own metameruic failure vs std observer… the answer is easy: the one that looks whiter (=less color pink-green tint) although people can have several candidates for white in b* axis if in “his observer” all displays have a*=0.
It’s very important to think about it.
CCFL, WHITE LED, in both cases, I think it’s important to create CCMX in CIE1931-2, and set to the D65 white point.
I set my White LED non-wide gamut monitor to D65 and used that as a reference.I think you need for high-resolution spectroradiometer is particularly pronounced for recent wide-gamut displays with sharp-angle spectra like a quantom-dot .
Of course, it is important to use a high-resolution spectroradiometer, but the more important is a measurement and make correction LUT should be performed with an appropriate workflow.
Use HCFR and careful 1D correction(Primary), and make the 3D LUT that corrects the saturation map on DispCal with 1D primary correction, after combining these 2LUTs, perform 1D correction again(secondary) and combining secondary 1D – DispCal 3D LUT – 1D primary, complete make calibration LUT for HDR display.
*In my environment, DispCal’s 3D LUT caused EOTF and white balance to deviate from the target, so I needed 1D recalibration.I have used INNOCN 27M2U, ASUS ProArt PA32UCX, and INNOCN 27M2V calibrated in this workflow, and I am very satisfied with the fact that these HDR displays of all looks identical, except for the difference in maximum brightness.
Calibrate another non-wide-gamut standard monitor to D65 and use it as a target for perceptual white point matching. Please refer to the following Lightillusion article. https://www.lightillusion.com/perceptual_match_guide.html
Metameric Failure is a very difficult topic and I don’t know of a good reliable solution.
But I’ve seen differences even when comparing two small gamut displays adjusted to D65.
One a LED LCD with about 89% Rec.709 coverage.
And another a CCFL LCD with about 70%-80% Rec.709 coverage. And both looked visually different.
Even when I used a 2nm Jeti 1501 probe.
So which display to use?
There might be something else rather then gamut causing metameric failure. But I don’t understand what.
What does the spectral graph look like for those?
How big of a difference are you seeing?
Also, how come they are both so much less than rec709?
I’d probably use the one that has the higher gamut coverage for the reference. Unless one looks more “white” to you, then go with that one.
2 different humans don’t necessarily even see the color the exact same, so there is no 100% true reference anyways.
Because of your choice of paramsin ref TRC + black output
You are correcting WP on the 2nd one (DisplayCAL?). Don’t do that = if you use a visual whitepoint match, you cannot perform LUT3D calculations against a Rec709 D65 or equivalent colorspace. You’ll have to go to relatieve colorimetric way or absolute vs equivalent colorspace with numerical coordinates of your “visually matched white”
Hello Vincent,
Thanks for your suggestion, yes, I used these rendering intent(surely include relative colorimetric), but not works well.
it seems strange in black side EOTF.*I use madTPG HDR mode for HDR pattern generator,
target on Calibration tub, White and Black level is “As measured”, 3D LUT tub, target peak luminance is 800cd/m2, Mastering display black level 0.0000cd/m2
I do not use Black point compensation, no offset.I think you need for high-resolution spectroradiometer is particularly pronounced for recent wide-gamut displays with sharp-angle spectra like a quantom-dot .
Q-LED can be measured at 3nm with ArgyllCMS + Xrite spectros easily, no narrow graph in SPD.
The narrow peaks that get blured at 10nm or 3nm resolution and need high end equipment to be captured with accuracy are the Widegamut-LED with PFS/KFS phosphors. Maybe you were talking about this 2nd group.Because of your choice of paramsin ref TRC + black output
You are correcting WP on the 2nd one (DisplayCAL?). Don’t do that = if you use a visual whitepoint match, you cannot perform LUT3D calculations against a Rec709 D65 or equivalent colorspace. You’ll have to go to relatieve colorimetric way or absolute vs equivalent colorspace with numerical coordinates of your “visually matched white”
Hello Vincent,
Thanks for your suggestion, yes, I used these rendering intent(surely include relative colorimetric), but not works well.Then aim for equivalent colorspace with transformed white as source profile for making the LUT3D.
Example: Rec709 with modified wp to be some xy coordinate. Use DisplayCAL synth profile editor.Another option is using perceptual (relative whitepoint) but since there is no single rule of doing perceptual, you’ll have to create display profile with your own cooked recipe: some % desaturation, some colorspace as reference… It should be done on “gear” icon in DisplayCAL in profile type, or wrinting your own perceptual comfiguration in colprof in argyll comand line. I think that DisplayCAL does not cover the whole set of perceptual configuration for colprof:
https://argyllcms.com/doc/colprof.htmlSome user complained about “colorimetric intent” (which is one way to get relative whitepoint = no changed white, the other is perceptual) and although i do not remember the detailt he got what he aimed for by using an equivalent Rec 709 colorspace with a whitepoint numerically matched to the “visual white” he wanted.
From the verification of Argyll CMS, I think that a high-resolution spectroradiometer is necessary (for high-precision measurement) for Q-dots or laser light sources that have sharp-angle spectra.
If it is a normal non-wide color gamut type, I think that X-rite equipment is enough.
https://www.argyllcms.com/doc/i1proHiRes.htmlThen aim for equivalent colorspace with transformed white as source profile for making the LUT3D.
yes, I set target white point on calibration tub, and used relative intent, but not works well.
comfirmed the exported LUT on lightspace, black side seemes strange on eotf and white balance.
so I correct it use HCFR grayscale measurement.
From the verification of Argyll CMS, I think that a high-resolution spectroradiometer is necessary (for high-precision measurement) for Q-dots or laser light sources that have sharp-angle spectra.
If it is a normal non-wide color gamut type, I think that X-rite equipment is enough.
https://www.argyllcms.com/doc/i1proHiRes.htmlPlot a QLED, like bundled Samsung TV, or some widegamut PG models from Asus. There are not sharp figures in a QLED spectral power distribution.
The sharp ones are in WLED PFS/KFS, or a laser proyector… or some older CCFL.
