- This topic has 9 replies, 3 voices, and was last updated 9 months, 3 weeks ago by
.
-
Posts
-
Hi there!
If my reference monitor is not connected via DeckLink, but it is a second desktop monitor and I use it in Davinci Resolve like reference viewing via the “Video Clean Feed” option. I did not find an explanation in the Wiki.
How do I properly calibrate it with ICC-profile compilation for MacOS (it should be a single curve+matrix for MacOS) ? How do I get a 3DLUT for this monitor? What steps should I take correctly in this case? Thank you!
I think this topic is discussed on other threads:
-reuse system wide ICC grey calibration (2nd display)
-make a detailed XYZLUT profile for 2nd display, set no calibration params so you can reuse current dekstop grey calibration. DO not install it system wide
-use that XYZLUT profile to make a Resolve LUT3D, DO NOT use “embed VCGT tables” since they will be applied system wide through OS grey calibration on dsplay 2.
-use that LUT3D for Resolve in the corect slot.
-reuse system wide ICC grey calibration (2nd display)
I’m using a monitor with a built-in 3DLUT, so do I need to calibrate it to the Native color space wide gamut?
-make a detailed XYZLUT profile for 2nd display, set no calibration params so you can reuse current dekstop grey calibration. DO not install it system wide
if I understand correctly, this means to create an XYZLUT profile with a large testchart. I am also using a monitor calibration profile not through the Davinci Resolve viewer? Just direct monitor calibration as it is? The calibration tab settings should all be “as measured”?
-use that XYZLUT profile to make a Resolve LUT3D, DO NOT use “embed VCGT tables” since they will be applied system wide through OS grey calibration on dsplay 2.
-use that LUT3D for Resolve in the corect slot.
This 3DLUT that I will use in the Davinci Resolve settings for the viewer or/and reference monitor?
As I’ve read on other forums, using the “Video clean feed” function gives a clean flat signal to the reference monitor, ignoring the system ICC profiles as I used DeckLink for it. https://forum.blackmagicdesign.com/viewtopic.php?f=21&t=108481#wrapper
At this time works only the monitor built-in 3DLUT and the 3DLUT used in Davinci Resolve for Viewer. In this case, for other using time I can use for the second monitor the correct ICC file, created based on the measurements for the 3DLUT compilation. This profile will display the correct colors when the second monitor’s desktop is not used as a reference monitor in Davinci Resolve. Am I right?
Same as above but since there is HW cal, no VCGT is involved…. unless it is not reliable/bad quality HW calibration / bad quality monitor like the ones from Benq SW line. In thta case it is very probable that you still need full grey calibration and VCGT as explained above. But let’s say that you have an Eizo CG:
-Use vendor software for HW cal. Use native colorspace & your desired white & brightness target. You’ll end with an ICC for color management, usually macos friendly due OS limitations. That ICC will be display profile for all apps.
-DisplayCAL, all calibration tab settings to native. Make XYZLUT profle with lots of patches. Instead of “calibrate & profile” should be just “profile only” os soemthing like that. YOu can use thehn same linear VCGT GPU calibration (no calibration) fro that DIsplaYCAL profile.
Do not install that new ICC as systemwide profile, just remember where it is stored. Then use that ICC as destination for LUT3D maker in DisplayCAL.OR
-You can use monitor vendor software and calibrate to your actual resolve target.. let’s say as an example D65, 100cdm2, 2.4 gamma, Rec709 primaries.
Validate results with DisplayCAL using simulation profile and “use simulation profile as display profile” and a huge set of patches to see how it behaves in target colorspace “volume”. If error is low no need to use software LUT3D in Resolve.You need to use a fast & good colorimeter for that task, at least for 1st option since mumber of patches can be very high, spectrophotometers are slow and WLED PFS type of backlights are not friendly for Xrite 10nm spectros using vendor software. Also you’ll need proper correction for colorimeter. Several threads talking about this.
Nothing like what I wrote about earlier. The ICC profile of the operating system affects the output through the Video Clean Feed in Davinco Resolve.
I put a crazy profile in MacOS and turned on Video clean feed, but the MacOS profile works the same way, although I have disabled the function of using it in Davinci. How to be in this situation then?
Thank you Vincent for your reply!
Yes, you are right that BenQ is not a very good monitor, but that is what it is.
In this case using the Video Clean Feed can I do the follow steps:
– Calibrate in PME to get native color space through the built-in monitor 3DLUT
– calibrate in DisplayCAL to get a regular ICC-profile for macOS in single curve+matrix profile type
– measure that with lots of patches to compile Davinci Resolve 3D LUT without VCGT to refine the installed ICC profile.– Calibrate in PME to get native color space through the built-in monitor 3DLUT
It is not using it, even if it actually has one. At best (eizo level) it will have a 17side cube, but PME does not measure ~5K patches at calibration stage (not profilling, calibration, more likely to 10×4 on grey or close to that). Grey is usually corrected through a 1Dlut will all entries (typical 1024 x3, like in a GPU.
So the main problem with those low quality monitors, specially on LG and Benq with so low quality cal software is:
-grey may not be grey. Report & RGB+grey balance, take a look at combined a*b* range or significative a* deviations in measured grey
-white may not be white because vendor software fro Benq & LG doe snot use the proper corrections for i1DisplayPro & that partiocular backlight (and spyders have none)
Check this fist. If some of this is wrong you’ll have to decide “WHERE” make the correction with the additional handicap of a all macOS limitations (=> default settings for DisplayCAL on macos).
-If grey is bad instead of using PME profile and if your mac has an AMD card, reaclibrate grey using DIsplayCAL & default profile type to overcome macos limitations. This way you’ll have c olorless gray & no banding, like if if was pure HW cal. That ICC profile will have VCGT calibration.
-If white is bad and you had an AMD card on Windows it can be corrected in VCGT without banding to some reasonable distance in DIsplayCAL calibrtaion. Not sure about this on macos, AFAIK default settings suggest do not correct white in GPU, hence you may want to do it in Resolve LUT3D.
Also if white is not very bad and you want to keep THAT whitepoint, use relative colorimetric when making LUT3Ds fro that display.Another option if grey and white is bad is just to do not use PME at all. Move to User/custom OSD preset, full GPU calibration in grey, tweak white point with OSD gains controls. Then proceed as in 1st example: macoS system wide matrix type profile with VCGT, XYZLUT profile using same VCGT, when making LUT3D do not embed VCGT since it is applied system wide by ICC.
Since your system has so many potential flaws is better to identify them first => PME then check results with displaycal. Then decide what needs to be corrected and where. Above is just a hint where to do it minimizing side effects. Also remember to use the proper correction for colorimeter when using DisplayCAL and if using spectrophotometer use 3nm high res mode.
Calibrite Display Pro HL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Today PME updated and I again updated the monitor’s 3DLUT in hopes of a better result. After that I could not immediately check the color as the button was not active. I did the measurement again and created an ICC profile for macOS. After that I measured the monitor again with Gamma 2.2. with the correction from the i1Profiler – LCD GB-r-LED IPS (DEll U2413). I hope DisplayCAL doesn’t take the ICC profile which I installed on the system, because I selected “Use simulation profile as display profile”. I left the device link profile empty and inactive.
This is the result.Grey has typical oscilation of these Dells, Benqs.. etc HW cal, they need more calibration patches in vendor software… but, that’s what you have. It can be corrected in GPU displayCAL calibration if your mac has an AMD GPU. It should have bigh bitdepth LUT and dithered outputs so correcting grey “system wide” with a DIsplayCAL slow cal with mac default settings for profile (to avoid macos issues with complex profiles) should work without issue. No visual differences with the proper HW cal it should have.
I meant, grey range got a pass, but it could be better. It is possible to spot visually a* errors in grey in a not color managed gradient with such range.With high errors in red it looks like you have HW cal to native but you are trying to verify a match to AdobeRGB… IDNK why you are doing that.
