Instability/Artifacts in TRC Generation in ArgyllCMS V3.4.1

Home Forums Help and Support Instability/Artifacts in TRC Generation in ArgyllCMS V3.4.1

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #145239

    Egor S.
    Participant
    • Offline

    Hello everyone,

    I use Palette Master Ultimate (BenQ software) to perform a hardware calibration  for my BenQ SW242Q Monitor with the following targets:
    – Colour gamut: native
    – White point: 6000K
    – Gradation: L*,
    followed by “Profiling Only” in DisplayCAL, skipping software calibration completely, using native white point. The LUT profile results are very good. I get a very smooth and neutral greyscale ramp and the colours look absolutely correct.

    The issue:
    In DisplayCal Profile Info I noticed a significant discrepancy between the TRC tags (rTRC, gTRC, bTRC) for the matrix part of the profile and the Tone Response Curves calculated from the 3D LUT, when the “LUT” Checkbox in DisplayCAL Profile Info is active.
    3D LUT (Correct): The measurement data in the LUT shows a near-perfect linear progression following the hardware-calibrated state (refer to attached screenshot “LUT tone response curve”).
    TRC Tags (Faulty): The generated 1D curves for the matrix part exhibit a massive “kink” at L*  ~30 and channel divergence that is not present in the physical measurements (refer to attached screenshot “TRC tags”).
    The vcgt tag (graphics card V-LUT) is linear.
    I have also attached the monitor ICC profile from DisplayCAL.

    If I generate a matrix profile from the same measuring data as for the LUT profile and apply the matrix profile, the displayed colours get a significant colour shift (due to the wrong TRC ). White is not white and the greyscale is not neutral. In my opinion, this is a bug in the way that ArgyllCMS calculates the TRC tags. This is strange because it didn’t happen with my old PC system, which had an older version of ArgyllCMS and DisplayCAL 3.8.9.3 installed.
    The issue persists regardless of the target gamma. Profiles created for both L* and pure Gamma 2.2 show the identical behaviour, confirming that the error is inherent to the TRC generation logic for this specific ArgyllCMS version and is not related to a specific transfer function. The LUT profile, on the other hand, is always very good.

    Do you agree that this must be a bug in the current version of ArgyllCMS? Should I write a bug report?

    My System:
    Palette Master Ultimate 2.6.0.0 for hardware calibration of the BenQ SW242Q Monitor.
    DisplayCAL 3.9.17
    ArgyllCMS 3.4.1
    Colorimeter: X-Rite i1 Display Pro
    OS: Windows 11 (25H2)
    Graphics card: NVIDIA GeForce RTX 5060 TI with studio driver
    CPU: Intel Core Ultra 5 245K
    Of course, HDR and Windows Auto Color Management (ACM) are deactivated.

    Thanks!

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

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

    #145243

    DaniJ
    Participant
    • Online

    Contact ArgyllCMS via their mailing list.

    #145244

    Kuba Trybowski
    Participant
    • Offline

    DisplayCAL 3.9.17 is pretty buggy.

    Try profiling your monitor using DisplayCAL 3.8.9.3 + Argyll 3.4.1. If the problem remains, contact Graeme Gill, the developer of ArgyllCMS.

    #145245

    Egor S.
    Participant
    • Offline

    Thank you guys for your assessment. I’ll do it the way you suggested.

    Egor

Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.

Log in or Register

Display Calibration and Characterization powered by ArgyllCMS