Getting the right calibration process: BenQ PD2700U with Macbook Pro M1 MAX

[EvanX]

I’m diving into calibration again and would appreciate any help. I’m esentially doing color grading with a reference monitor via UltraStudio Monitor 3G, but trying to learn the basics of calibration so i’m starting out by just calibrating my macbook and the monitor directly connected to it without the ultrastudio. Here’s my setup:

1. Macbook Pro M1 Max
2. BenQ PD2700U
3. UltraStudio Monitor 3G
4. Calibrite ColorChecker Display Pro

Target: Rec.709 Gamma 2.4 at 100-120 nits (I calibrated to 110 just as a middle ground)

I’ve done a calibration with settings i’ve attached below from previous documentations and reading upon several threads. I hope the questions i ask below resonate with others starting out and this can serve as a guide for other people as well. Things I would like to inquire are:

1. Although the calibration verification results show good delta E across the board, I view my monitor as losing contrast from before it was calibrated (Specifically from the apps icon, Resolve’s gray background is more lifted as well as Photoshop etc). Is this just to be expected from a calibration and my eyes have gotten too used to the contrast before? Are my calibration settings correct to begin with?
2. Are my verification settings correct? I provided the description in the name of the attachments. From what I grasped from this thread https://hub.displaycal.net/forums/topic/do-i-have-this-right/ , I used the Simulation Profile + Unchecking Use as Display Profile to provide me with a calibration of my monitor using the calibration against a rec.709 2.4 standard.
3. The gamma graph on the report shows a curved line with the blacks performing fairly rough and then coming to a more stable line at 2.2-2.3. I calibrated for 2.4 so I’m not sure why I’m not getting more consistent results. A year ago, I did a calibration and it gave me a straighter line on the gamma graph and much closer to 2.4. Is this a sign of monitor performance degradation? (Here’s my last year measurement report from another thread https://hub-assets.displaycal.net/wp-content/uploads/users/evanx/2023/07/15/Measurement-Report-3.8.9.3-%E2%80%94-Resolve-%E2%80%94-2023-07-15-21-48-13.html)
4. In the calibration tab under tone curve, I used custom > 2.4 > absolute. Through reading the forums, some have also suggested to just use As Measured. This gave me slightly worse results from a delta E of 1.07 to 1.7. Which is the correct way?
5. When should i use the ambient light settings properly? I sometimes grade in a brighter environment for ‘less serious’ work.
6. I would like to inquire more regarding situations to use the different profiling types: single curve + matrix, XYZ + matrix etc. If that’s too in depth to go in to, I would like to inquire for the case of calibrating my display via a normal signal chain & calibrating my reference monitor via ultrastudio through resolve.

In the future, I’d like to move forward with my calibration and recalibrate my reference monitor through the ultrastudio signal chain. If anyone can give me an affirmation/correction, but I presume this is the correct workflow: Calibrating through the “Monitor Calibration” menu in Resolve, calibrating and verifying as usual, then feeding the LUT created to the viewing monitor in project settings. (Can anyone give me a description of the LUT creation settings?)

Appreciate all the help i can get. Hope this helps others who are confused as well. Thanks!

Attachments:

You must be logged in to view attached files.

Calibrite Display Pro HL on Amazon   Calibrite Display SL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Vincent]

1. Unlikely, I bet you mislabel “loosing contrast” as actuakl color management based on actual TRC value for yor display vs nominal value in EDID ICC profile.
   If in doubt just test a no calibration scenario agains PD2700U buikt in profile created by macOS when you plugged it.
2. No. MacOS is color managed. Since you are aiming for an OS profile (all purpose profile) it is usless to aim to 2.4 or 1.8. OS color management will undo all.
   That includes Resolve ICC color management. The other settings are ok. It does not means that usong 2.4 is wrong… just useless, on a color managed enviroment 2.2 will do the same.
   If you wish a LUT3D for resolve yiu can use XYZLUT profile and depening if you are using a deckling or a regular GPU from your mac you should embed or do not embed VCGT in LUT3D.
3. It bends near black because your display has no infinite contrast. Bending of the graph cannot be avoided because at RGB 0 0 0 your display will not give you 0 nit.
4. Useless for an all “purpose-OS display profile”, explained in “2”
5. Never. It does not work the way you thing. Even if you can compensate gamma to “fake”/”alter” desired TRC to match ambient light, then on profiling stage it will measure actual TRC and whatever you did will be undone bu color management. Explained in “2”
6. No for an all purpose display profile in macOS. macOS color management engine is very limited, stick to single curve matrix with fake infinite contrast (BPC).
   For a ICC profile that is going to be used to create a LUT3D you can use other profile settings since this profile is not going to be assigned in OS settings, but beware with VCGT as explained before. Also such profile on a regular GPU (non decklink) must share VCGT calibration with your single curve matrix profile in OS.

This is the actual measurement report that matters
https://hub-assets.displaycal.net/wp-content/uploads/users/evanx/2023/07/15/Measurement-Report-3.8.9.3-%E2%80%94-Resolve-%E2%80%94-2023-07-15-21-48-13.html
The othr ones seem to be misconfigured. Also do not aimg for rec1886 on a 1000:1 display… it does not behave as you think, just washes dark greys. Read ITUR rec1886 definition (ist a variable TRC depending on contrast). If you wish to simulate color managed Rec709 2.4, just simulate that, not rec1886.

[EvanX]

Thanks a lot for the clearing of misguided concepts, Vincent. Will post again soon with an updated calibration according to your advice.

[EvanX]

Dear Vincent, I’ve tried to run another set of calibration mimicking the settings on the measurement report you said was well configured. Could you help to confirm whether my process is correct please? The only thing I changed was setting the target tone curve to custom > 2.4 > relative (Yes i’ve understood you said this is useless, but I tried to minimize factors I didn’t know and just followed the previous measurement report to spare any confusion). Below I attach the measurement reports. Still a bit unsure which settings provide just the verification report that tells me that my calibration was done properly, so i attached all three variants (labels stated in the attachment name):
1. No simulation profile checked
2. Simulation profile checked, but not using “Use Simulation Profile As Display Profile”
3. Simulation profile checked + using “Use Simulation Profile As Display Profile”

1. Unlikely, I bet you mislabel “loosing contrast” as actuakl color management based on actual TRC value for yor display vs nominal value in EDID ICC profile.
   If in doubt just test a no calibration scenario agains PD2700U buikt in profile created by macOS when you plugged it.

To jump back on to your #1 response, how would I do a test of no calibration scenario against my PD2700U’s built in profile when plugged in? Below I will attach a measurement report where i:

1. Changed the display color profile setting of my monitor to the profile created by macOS (Default name is “BenQ PD2700U”) -> attachment file name is the one I labeled with “Default OS Profile” in the name.
2. Did a measurement report where I presume the one needed is the one with no simulation profile checked because I just want to check the ‘default’ state of my monitor?
3. Although, the “Settings” on the top part of the app wouldn’t allow me to change the settings to anything else in order to do the measurement report. So the settings were still on my calibration settings. I don’t really know if this affected the result. If yes, would love any more tips.

Note: For all of these verification tests, I had the tone curve set to Custom > 2.4 > Relative, Black output offset to 100% (Referencing to the answer by Florian on this thread regarding tone curve https://hub.displaycal.net/forums/topic/do-i-have-this-right/ )

Attachments:

You must be logged in to view attached files.

[Vincent]

Dear Vincent, I’ve tried to run another set of calibration mimicking the settings on the measurement report you said was well configured. Could you help to confirm whether my process is correct please? The only thing I changed was setting the target tone curve to custom > 2.4 > relative (Yes i’ve understood you said this is useless, but I tried to minimize factors I didn’t know and just followed the previous measurement report to spare any confusion). Below I attach the measurement reports. Still a bit unsure which settings provide just the verification report that tells me that my calibration was done properly, so i attached all three variants (labels stated in the attachment name):
1. No simulation profile checked
2. Simulation profile checked, but not using “Use Simulation Profile As Display Profile”
3. Simulation profile checked + using “Use Simulation Profile As Display Profile”

You are in a macOS color managed enviroment, hence:
-your desired white (assume D65)
-your desired brightnes
-gamma close to native, likely to be 2.2

Then you test if resulting profile + embeded VCGT calibration matches display behavior: no simulation profile.

Optionally you can test how that monitor renders an image encoded in certain colorpsace in that color managed system: simulation profile (rec709 gamma 2.4) but NO use simulation profile as display profile.

“use simulation profile as display profile” is used for testing HW cal or LUT3D (with embeded VCGT calibration)  or factory presets on a display, and since you are on macOS without a decklink it does not apply, unless you want to test factory modes for your various displays.

1. Unlikely, I bet you mislabel “loosing contrast” as actuakl color management based on actual TRC value for yor display vs nominal value in EDID ICC profile.
   If in doubt just test a no calibration scenario agains PD2700U buikt in profile created by macOS when you plugged it.

To jump back on to your #1 response, how would I do a test of no calibration scenario against my PD2700U’s built in profile when plugged in? Below I will attach a measurement report where i:

1. Changed the display color profile setting of my monitor to the profile created by macOS (Default name is “BenQ PD2700U”) -> attachment file name is the one I labeled with “Default OS Profile” in the name.

No, it is using DisplayCAL profile. Check file, it is shown.

1. Did a measurement report where I presume the one needed is the one with no simulation profile checked because I just want to check the ‘default’ state of my monitor?
2. Although, the “Settings” on the top part of the app wouldn’t allow me to change the settings to anything else in order to do the measurement report. So the settings were still on my calibration settings. I don’t really know if this affected the result. If yes, would love any more tips.

Note: For all of these verification tests, I had the tone curve set to Custom > 2.4 > Relative, Black output offset to 100% (Referencing to the answer by Florian on this thread regarding tone curve https://hub.displaycal.net/forums/topic/do-i-have-this-right/ )

Since you are on macOS thus your options are limited, try to simulate driver/default ICC and use it as Display profile to test factory behavior. Keep in mind that your changes in OSD settings wil be present in output, like gamam settinsg in monitor OSD or whitepoint coordinates.

[Author]

Posts