(Serious) M1 MacBook Pro Gamma Is Wrong After Calibration

Home Forums Help and Support (Serious) M1 MacBook Pro Gamma Is Wrong After Calibration

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #36115

    ransontham
    Participant
    • Offline

    First time calibrating the miniled MacBook Pro with an i1Display Studio on macOS Monterey (might be important).

    When calibrating to a 2.2 Gamma and, and only and, selecting the XYZ LUT profiling option, the resulting calibration is apparently a 1.0 Gamma.

    I know this cause it is very apparent, and i also ran a verification test. (2 photos 1 file below)

    As you all know, out of the box, DisplayCal doesn’t show the built-in display on the new M1 macs.

    Following this https://www.reddit.com/r/colorists/comments/ndvvxi/solution_got_displaycal_working_on_m1_macs/ thread, i downloaded the Argyll executables v2.3.1. (might be important)

    I put the executables in the folder, and followed ALL instructions. Solution worked.

    More screenshot of my settings below. Please help 🙁

    Attachments:
    You must be logged in to view attached files.

    i1Display Studio on Amazon  
    Disclosure: As an Amazon Associate I earn from qualifying purchases.

    #36123

    ransontham
    Participant
    • Offline

    Here is other photos showing what is going on. They have not been edited.

    The calibration curves seem to not want to apply.

    • This reply was modified 3 weeks, 5 days ago by ransontham.
    • This reply was modified 3 weeks, 5 days ago by ransontham.
    Attachments:
    You must be logged in to view attached files.
    #36132

    Vincent
    Participant
    • Offline
    1. Use latest ArgyllCMS (2.3.1). issues with that to ArgyllCMS maillist
    2. MacOS volor management engine DID nt like complex profiles, so stay with default config.
    3. M1 P3 XDR display should have a truye neutral grey and sRGB-like gamma, so instead running full calinbration, ust use a customized preset and give macOS actual coordinates of your whitepoint so it can correct it by itself… that’s teh point with those screens.
    4. Look at your measured values, It seems that is YOUR user configuration at OS level the one that is messing around. Not profile or calibration. Profile itself (reference values measured after calibration) seems to be as expected. Disable HDR, TRue color, night shift and all that stuff.
    • This reply was modified 3 weeks, 5 days ago by Vincent.
    #36137

    ransontham
    Participant
    • Offline

    2. But XYZ LUT + matrix profiles didn’t use to cause this problem on my 2018 intel MacBook Pro with macOS Monterey. So what changed?

    3. Yes, I did that. After fine tuning the preset, I also ran a calibration according to this tutorial, https://www.youtube.com/watch?v=cLl01EjHU3Q&t=1302s. He didn’t have any problems.

    4. True tone is off, night shift is off. When profiling, the calibration is applied correctly (very apparent, with a 2.2 gamma). But after the profile is generated and applied, the screen turns into one with a 1.0 gamma. Will disable HDR option to see if it works.

    #36138

    ransontham
    Participant
    • Offline

    Update: disabling HDR option makes no difference. Does anyone have this same problem?

    • This reply was modified 3 weeks, 4 days ago by ransontham.
    #36141

    Vincent
    Participant
    • Offline

    4. True tone is off, night shift is off. When profiling, the calibration is applied correctly (very apparent, with a 2.2 gamma). But after the profile is generated and applied, the screen turns into one with a 1.0 gamma.

    Actual calibration response (= display behavior after GPU calibration) is measured & stored in profile. It is in menasurement report & profile. That is working for sure, bu pure expreimental means.
    Hence the culprit is OS color management engine, which is very limited in macOS. Stick to default oversimplified profiles… or blame Apple & demand macOS to properly support all kinds of profiles. An apple P3 display should be modeled by a matrix profile without issue.

    #36142

    Vincent
    Participant
    • Offline

    3. Yes, I did that. After fine tuning the preset, I also ran a calibration according to this tutorial, https://www.youtube.com/watch?v=cLl01EjHU3Q&t=1302s. He didn’t have any problems.

    He is using builtin colorspace simulation on XDR screens. That does not imply using DisplayCAL or i1Profiler for calibration.

    You only need to measure actual white. Display profile will remain the same (because rendering to screen is always white point relative)
    If you give a color managemente ngine actual white, you know primaries coordinates by design (WLED PFS phosphor tuned to exactly P3) and know display TRC (gamma) by fairly good factory calibration… you only need actual white coordinates to make an idealized transformation to whateever colorspace you want. It’s like a simplified version of LUT3D creator, or if you know 3rd party tool “novideo_sRGB” on Windows… that tool is doing the same as macOS colorspace simulation.

    Also HE IS WRONG, they nickname “ArtIsWrong” in Luminous Landscape and other pro printer & color management engine forums, like in many many other videos. i1Profiler lacks of colorimeter correction for mac p3 screens. Closes one will be pfs phosphor 95% P3. He is using “white led” = LED for sRGB only displays… so whoite point will be off by a margin.

    Alse HE IS NOT USING XYZLUT PROFILES (“TABLE” in Xrite software). He is using idealized matrx with 1 TRC curve. It’s in the video… which he skips fast 24:09 so he has not to explain all this stuff.

    What you are experinecing is a set of malfunction & failures due to oversimplification in macOS default color management engine. Yiou have to stick to default idealized matrix 1 curve profiles.

    It’s macOS fault, so
    -do not profile screen, yous correct white point using built in app i mac, using actual white measuremenst from DisplayCAL + macbook retina CCSS (p3 WLED PFS)
    OR
    -calibrate & profile, but use the extremely idealized profiles selected by default (matrix 1 TRC curve).

    #36144

    ransontham
    Participant
    • Offline

    4. True tone is off, night shift is off. When profiling, the calibration is applied correctly (very apparent, with a 2.2 gamma). But after the profile is generated and applied, the screen turns into one with a 1.0 gamma.

    Actual calibration response (= display behavior after GPU calibration) is measured & stored in profile. It is in menasurement report & profile. That is working for sure, bu pure expreimental means.
    Hence the culprit is OS color management engine, which is very limited in macOS. Stick to default oversimplified profiles… or blame Apple & demand macOS to properly support all kinds of profiles. An apple P3 display should be modeled by a matrix profile without issue.

    Agreed.

    #36145

    ransontham
    Participant
    • Offline

    Alse HE IS NOT USING XYZLUT PROFILES (“TABLE” in Xrite software). He is using idealized matrx with 1 TRC curve. It’s in the video… which he skips fast 24:09 so he has not to explain all this stuff.

    It’s macOS fault, so
    -do not profile screen, yous correct white point using built in app i mac, using actual white measuremenst from DisplayCAL + macbook retina CCSS (p3 WLED PFS)
    OR
    -calibrate & profile, but use the extremely idealized profiles selected by default (matrix 1 TRC curve).

    Yea… he did use a 1 TRC curve.

    So no real solution for now… New technology is weird.

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS