- This topic has 29 replies, 2 voices, and was last updated 3 years, 3 months ago by
.
-
Posts
-
Redo 2.2. => should be closer to native. Just in case dark patches are too bright even for color managed apps to undo.
Not sure why my last post didn’t go through. Sorry if this is a double. Re-did validation with 2.2 gamma – better contrast. I still don’t see anything that would explain green tones other than it being a poor correction for my monitor… The rgb balance is a bit off but not SO much…that wouldn’t explain it right?
Anything in that profile validation or the 3d projection that looks off to you?
Also to avoid some GPU issues try to “profile only” factory AdobeRGB preset (choose calibration tab all to native). It will measure display and make a taylor made profile (simple or complex depending on youir choice), no GPU calibration at all.
If issues dissapear it is GPU issue.This is missing
Also to avoid some GPU issues try to “profile only” factory AdobeRGB preset (choose calibration tab all to native). It will measure display and make a taylor made profile (simple or complex depending on youir choice), no GPU calibration at all.
If issues dissapear it is GPU issue.This is missing
I mean, no GPU calibration at all, simple (Mtx 1 curve) or complex (xyzlut) profile on factory calibration preset. Then check with each one if the visual artifacts errors you report are there.
If AdobeRGB profile in OS settings works poperly then 1st option should give close results. If not compare 2 ICCs, primaries location because it may be working properly and your issues are seeing things wrong with AdoberGB factory preset and AdobeRGB icc as display profile => device is working properly.I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues. Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim. RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim. RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I previously tried white matching. It doesn’t change the green tinted blues. Even with the white matched, I see substantial differences between calibrated “custom” setting and the adobe setting profiled – as well as my other calibrated monitor.
Everything leads me to believe there is a correction issue because with no correction on, the adobe RGB white shows as a perfect 6500k white; with the correction on its way off.
I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim.RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I previously tried white matching. It doesn’t change the green tinted blues. Even with the white matched, I see substantial differences between calibrated “custom” setting and the adobe setting profiled – as well as my other calibrated monitor.
Then profile only some native gamut OSD preset with some 6500K correlated color temperature preset and see if it dissapears => GPU issues when loading calibration.
Also compare measured blue & cyan in factory AdobeRGB calibration and in that native gamut preset. No huge differences are expected => no issuesEverything leads me to believe there is a correction issue because with no correction on, the adobe RGB white shows as a perfect 6500k white; with the correction on its way off.
Because AdobeRGB preset is actually way off in white.
Also “6500K” (alone) is meaningless, it has no information about white, there is a missing coordinate, 6500K CCT could be green or pink beause there are many 6500K CCT.
No colorimeter correction does not mean no correction. It actually means “correct readings from my display as if that display spectral power distribution is like my colorimeter filters spectral sensivity” (AND IT IS NOT, 100% SURE).
Also :
(poor color uniformity may make things look different if one preset has uniformity compensation and the other does not)
Metameric observer failure (your visual system vs std observer) is another elephant in the room, but since narrow spectral spikes happen in red channel for WLED PFS and you have no issues with RMY… blue led is almost the same for all these LED backlight displays. But who knows, you may need to rent an i1Pro2 and measure @3nm if for that particular unit blu spike is placed +-10nm from typical position.
I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim.RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I previously tried white matching. It doesn’t change the green tinted blues. Even with the white matched, I see substantial differences between calibrated “custom” setting and the adobe setting profiled – as well as my other calibrated monitor.
Then profile only some native gamut OSD preset with some 6500K correlated color temperature preset and see if it dissapears => GPU issues when loading calibration.
Also compare measured blue & cyan in factory AdobeRGB calibration and in that native gamut preset. No huge differences are expected => no issuesEverything leads me to believe there is a correction issue because with no correction on, the adobe RGB white shows as a perfect 6500k white; with the correction on its way off.
Because AdobeRGB preset is actually way off in white.
Also “6500K” (alone) is meaningless, it has no information about white, there is a missing coordinate, 6500K CCT could be green or pink beause there are many 6500K CCT.
No colorimeter correction does not mean no correction. It actually means “correct readings from my display as if that display spectral power distribution is like my colorimeter filters spectral sensivity” (AND IT IS NOT, 100% SURE).
Also :
(poor color uniformity may make things look different if one preset has uniformity compensation and the other does not)
Metameric observer failure (your visual system vs std observer) is another elephant in the room, but since narrow spectral spikes happen in red channel for WLED PFS and you have no issues with RMY… blue led is almost the same for all these LED backlight displays. But who knows, you may need to rent an i1Pro2 and measure @3nm if for that particular unit blu spike is placed +-10nm from typical position.
Is it possible to manually adjust the correction to provide a slightly wider spike (or just moved) in the red (or green) channels?
I have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim.RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I previously tried white matching. It doesn’t change the green tinted blues. Even with the white matched, I see substantial differences between calibrated “custom” setting and the adobe setting profiled – as well as my other calibrated monitor.
Then profile only some native gamut OSD preset with some 6500K correlated color temperature preset and see if it dissapears => GPU issues when loading calibration.
Also compare measured blue & cyan in factory AdobeRGB calibration and in that native gamut preset. No huge differences are expected => no issuesEverything leads me to believe there is a correction issue because with no correction on, the adobe RGB white shows as a perfect 6500k white; with the correction on its way off.
Because AdobeRGB preset is actually way off in white.
Also “6500K” (alone) is meaningless, it has no information about white, there is a missing coordinate, 6500K CCT could be green or pink beause there are many 6500K CCT.
No colorimeter correction does not mean no correction. It actually means “correct readings from my display as if that display spectral power distribution is like my colorimeter filters spectral sensivity” (AND IT IS NOT, 100% SURE).
Also :
(poor color uniformity may make things look different if one preset has uniformity compensation and the other does not)
Metameric observer failure (your visual system vs std observer) is another elephant in the room, but since narrow spectral spikes happen in red channel for WLED PFS and you have no issues with RMY… blue led is almost the same for all these LED backlight displays. But who knows, you may need to rent an i1Pro2 and measure @3nm if for that particular unit blu spike is placed +-10nm from typical position.
Is it possible to manually adjust the correction to provide a slightly wider spike (or just moved) in the red (or green) channels?
Does “profile only” in native gamut solve the issue?
You are trying to tweak something that a priori does not need further refinement just to do not check some of the more probable issues.
Profile only in matrix & XYZLUT, native gamut, some color temperature factory preset 6500K CCT, or white visual match. This way you check:
-potential cyan non ideal beavior that a factory ARGB preset may have solved if it is a LUT3D (instead of typical Lut-matrix-lut which cost wise may use instead of a LUT3D). profile only XYZLUT goes away vs profile matrix remain
-potential GPU LUT issues, hence profile onlyI have another monitor hooked up to the same GPU – without any issues. Doesnt that rule out a GPU issue?
Not with the same calibration and the same display repsonse.
In any event, I ran the profile only on the adobe RGB setting and I don’t see any issues.
If a custom “profile only” with DisplayCAL on AdobeRGB preset shows thing “Ok” in color managed apps then it is not a “profile” (display description) issue. Thta was the required test.
Honestly, the ICC profiles don’t make a huge difference when I’m on the adobe RGB setting. The biggest difference is when I switch from adobe RGB to my custom setting. Its a huge difference – even when both settings are calibrated. The “custom” turns yellowish green compared to adobe RGB setting. That doesn’t make much sense to me.
That is white point, nothing related to your “blue colors” previous claim.RGB gain to your desired white, visual whitepoint editor if it does not match your expectations… and it’s done.
(also poor color uniformity may make things look different if one preset has uniformity compensation and the other no)
There was no issue after all… 2 pages fr nothing…
I previously tried white matching. It doesn’t change the green tinted blues. Even with the white matched, I see substantial differences between calibrated “custom” setting and the adobe setting profiled – as well as my other calibrated monitor.
Then profile only some native gamut OSD preset with some 6500K correlated color temperature preset and see if it dissapears => GPU issues when loading calibration.
Also compare measured blue & cyan in factory AdobeRGB calibration and in that native gamut preset. No huge differences are expected => no issuesEverything leads me to believe there is a correction issue because with no correction on, the adobe RGB white shows as a perfect 6500k white; with the correction on its way off.
Because AdobeRGB preset is actually way off in white.
Also “6500K” (alone) is meaningless, it has no information about white, there is a missing coordinate, 6500K CCT could be green or pink beause there are many 6500K CCT.
No colorimeter correction does not mean no correction. It actually means “correct readings from my display as if that display spectral power distribution is like my colorimeter filters spectral sensivity” (AND IT IS NOT, 100% SURE).
Also :
(poor color uniformity may make things look different if one preset has uniformity compensation and the other does not)
Metameric observer failure (your visual system vs std observer) is another elephant in the room, but since narrow spectral spikes happen in red channel for WLED PFS and you have no issues with RMY… blue led is almost the same for all these LED backlight displays. But who knows, you may need to rent an i1Pro2 and measure @3nm if for that particular unit blu spike is placed +-10nm from typical position.
Is it possible to manually adjust the correction to provide a slightly wider spike (or just moved) in the red (or green) channels?
Does “profile only” in native gamut solve the issue?
You are trying to tweak something that a priori does not need further refinement just to do not check some of the more probable issues.
Profile only in matrix & XYZLUT, native gamut, some color temperature factory preset 6500K CCT, or white visual match. This way you check:
-potential cyan non ideal beavior that a factory ARGB preset may have solved if it is a LUT3D (instead of typical Lut-matrix-lut which cost wise may use instead of a LUT3D). profile only XYZLUT goes away vs profile matrix remain
-potential GPU LUT issues, hence profile onlyI’ll give that a try. The native gamma response isn’t perfect so its not ideal to entirely profile but it may shed some light on the issue.
It asks if I want to embed current calibration curves or use linear ones. Does that make a difference?
Yes
Linear = no GPU calibration, profile only
Current = whatever GPU calibration you have loaded now because of some profile set as active/current/default for that display.
… but if you actually “profiled only” AdobeRGB factory preset it should have asked you the same. So… if you did not, you must test profilye only AdobeRGB factory preset before this.
Yes
Linear = no GPU calibration, profile only
Current = whatever GPU calibration you have loaded now because of some profile set as active/current/default for that display.
… but if you actually “profiled only” AdobeRGB factory preset it should have asked you the same. So… if you did not, you must test profilye only AdobeRGB factory preset before this.
So, a combination of profiling only with XYZ lut and reducing the green in the white point have made the issue less substantial. Seems like thats as good as I’ll get it.
Changing the white point (no longer being balanced in the middle bars) doesn’t affect the color accuracy of the profile? I’m about 2.0 de off of what the white point says is on the kelvin scale.
Profile accuracy won’t be affected… it will store that non whiteness. Image editors render images to screen with relative white point… so they do not care as long as profile is accurate (PCS transformed).
The main issue is that if you made a visual whitepoint match to what you see as white because some observer metameric failure, other people may not see that screen as white as you.
