- This topic has 22 replies, 4 voices, and was last updated 5 years ago by
.
-
Posts
-
Florian – do you see any serious drawbacks to this method if it kills that green in the GUI?
You loose some accuracy. You can check via a measurement report if it’s significant.
So I ended up using WolfPeace’s method and I have results that work for me. Since then, I’ve swapped my UP2516D for an identical UP2516D (Amazon’s nonexistent price match…) and used everything you explained here to calibrate and profile this new panel and make a LUT for use in DaVinci. Great!
One last question. Since the UP2516D has 98% P3-DCI coverage, I’m interested in using the monitor in this mode. This would be exclusively for previewing content graded in DaVinci before DCP output. If I calibrated, profiled, and made an ICC profile for the monitor for P3, and made a DaVinci LUT for P3, that seems to me like it would work no problem. What I’m wondering is, could I also make a DaVinci LUT for Rec709 to use in conjunction with my P3 ICC? P3 has a different white point – would the LUT be able to reconcile this?
What I’m wondering is, could I also make a DaVinci LUT for Rec709 to use in conjunction with my P3 ICC? P3 has a different white point – would the LUT be able to reconcile this?
In your case I’d probably use the calibrated whitepoint (= keep 3D LUT rendering intent at “relative colorimetric”) instead.
Florian, your help 3 years ago was invaluable! I’m back with a new challenge regarding this same monitor:
I’ve built a new computer and migrated from Mac to Windows 10. I’ve used the same principles you explained above to calibrate my monitor and produce a Rec709 LUT for the viewer in Resolve.
The difference now is that Windows 10 can send out a wide gamut image to my wide gamut Dell 2516D. Firefox can not reconcile a wide gamut monitor and streaming video. This results in super saturated videos from youtube or vimeo. Given that I color correct for web almost exclusively, this is annoying when I’m trying to evaluate results once uploaded.
The 2516D has a sRGB mode – my understanding is this is a LUT built into the monitor to constrain the gamut to sRGB. I could potentially switch to that mode and calibrate as I did before in Mac OS. The downside of this is I lose the ability to change the brightness/contrast/color offsets with the OSD controls. My three questions are:
1] Is there a way to apply a LUT, as I have for the Resolve viewer… but instead for the whole display, without additional hardware, in Windows? This would be on top of my ICC profile.
2] Alternatively, is there a way to tell DisplayCAL to make a ICC profile that limits the gamut to Rec709 or sRGB?
3] Or should I just not worry about losing the OSD brightness/contrast/etc controls when I switch the monitor to sRGB mode and re-calibrate with that activated?
To recap, my goal here is to calibrate my monitor, have a calibrated viewer in resolve, and have internet video appear with correct saturation in firefox all on my 2516D
Thanks!!
- Instead of using factory sRGB modoe (which is a LUT-matrix-LUT calibration as you said), try to use Dell’s app for calibration “DUCCS”, currently 1.6.6 as some other Dell ownder posted a day ago.
- You can try to use a DisplayCAL calibration (GPU) of an uncalibrated CAL1/CAL2 to build a lut-matrix-lut an with some C++ work write it to your monitor with Dell’s SDK (which supports UP2516D AFAIK).
It could be easier if future releases of DisplayCAL allowed to write a simplified LUT3D in a generic 3x1024x16bit prelut, 3x3x16bit gamut emulation matrix and 3x1024x16bit postlut.
Hard work, but possible and even easier if DisplayCAL/ArgyllCMS could write that vendor agnostic lut-matrix-lut data that a lot of vendors use for their monitors with HW calibration (Dell, HP… etc) - sRGB factory mode is very likely to have sRGB TRC. It’s not what you expect for video, also factory mode whitepoint could be off.
Unless there is an unsolved issue with DUCCS application and your monitor, use it to calibrate monitor LUTs to Rec709 gamma 2.2 or whatever you use (use DUCCS “custom xy” preset, set manually calibration target). Then test it with DisplayCAL. If it is too off in white or gamma, use DisplayCAL again in than CAL1/CAL2 mode to solve these issues.
P.S: AFAIK ICCv2 profiles from i1Profiler o i1Profiler variants were not supported very well by Firefox color management. Firefox may behave as if it had no color management (images or HTML colors).
It does not matter for you if your CAL1/CAL2 is calibrated to Rec709/g2.2 but if you wish to use native gamut and use Firefox you may want to “profile without calibrate” in DisplayCAL after you calibrate in DUCCS. Maybe now this issue is solved in current versions, but you should verify it.Ok this is definitely starting to slip outside my skill level here. Anything involving C++ and the Dell SDK is definitely beyond my capabilities.
My understanding of DUCCS is that it creates a LUT-matrix-LUT that it loads into CAL1/CAL2 and then makes an ICC profile for the operating system. Does this ICC profile have a calibration or is it just a profile of what the LUT is doing?
Could I use DUCCS to make a Rec709 LUT-matrix-LUT and then use DisplayCAL to make and load the ICC profile? Would it be wrong to not just profile, but also calibrate with DisplayCAL at this step? In theory the calibration would have a very minimal effect so long as DUCCS has done a good job.
One point of clarification – the monitor has multiple gamut modes, not just sRGB. There’s Rec709, Adobe 1998, etc. So I could use the factory Rec709 mode but the lack of control over whitepoint and brightness with the OSD would remain.
My understanding of DUCCS is that it creates a LUT-matrix-LUT that it loads into CAL1/CAL2 and then makes an ICC profile for the operating system. Does this ICC profile have a calibration or is it just a profile of what the LUT is doing?
Just a profile with a linear 1D-LUT calibration for graphics card (to ensure that no GPU calibration is applied)
Could I use DUCCS to make a Rec709 LUT-matrix-LUT and then use DisplayCAL to make and load the ICC profile? Would it be wrong to not just profile, but also calibrate with DisplayCAL at this step? In theory the calibration would have a very minimal effect so long as DUCCS has done a good job.
You can:
-use DUCCS, create HW calibration and use the generated ICC (but Firefox may not recognize it… but since it’s Rec709 it does not matter too much)
-use DUCCS, create HW calibration and create a profile without calibration in DisplayCAL (set all calibration tab to as measured)
-use DUCCS, create HW calibration and create a profile WITH calibration in DisplayCAL if DUCCS miss to get desired white, or if it could not solve some neutral grey issues in your screen.Which one do you need? Validate DUCCS result with DisplayCAL and you’ll know.
