hdr-compute-peak triggers on HDR10+ content

[Tjoma96]

mpv Information

mpv v0.41.0-42-g468e2a847 Copyright © 2000-2025 mpv/MPlayer/mplayer2 projects
 built on Dec 29 2025 11:06:08
libplacebo version: v7.358.0 (v7.351.0-106-gbc90ef9-dirty)
FFmpeg version: N-122276-g9ab2a437a
FFmpeg library versions:
   libavcodec      62.23.102
   libavdevice     62.2.100
   libavfilter     11.10.101
   libavformat     62.8.101
   libavutil       60.20.100
   libswresample   6.2.100
   libswscale      9.3.100

Other Information

- Windows version: Microsoft Windows [Version 10.0.26200.7462]
- GPU model, driver and version: NVIDIA RTX 3060, driver ver. 591.44
- Source of mpv: https://github.com/zhongfly/mpv-winbuild/releases
- Latest known working version: N/A
- Issue started after the following happened: Playing HDR10+ video

Reproduction Steps

Playing a video with HDR10+ metadata in it, peak detection gets enabled. It doesn't happen with Dolby Vision content.

Expected Behavior

Tonemapping using HDR10+ metadata

Actual Behavior

Tonemapping using MPV peak detection

Log File

output.txt

Sample Files

https://mega.nz/file/nehDka6Z#C5_OPbSZkONdOp1jRmc09C9-viDc3zMj8ZHruHcWKyA

I carefully read all instruction and confirm that I did the following:

• I tested with the latest mpv version to validate that the issue is not already fixed.
• I provided all required information including system and mpv version.
• I produced the log file with the exact same set of files, parameters, and conditions used in "Reproduction Steps", with the addition of --log-file=output.txt.
• I produced the log file while the behaviors described in "Actual Behavior" were actively observed.
• I attached the full, untruncated log file.
• I attached the backtrace in the case of a crash.

[kasper93]

@haasn decided that HDR10+ metadata is not worth using haasn/libplacebo@defc776

[Tjoma96]

@haasn decided that HDR10+ metadata is not worth using haasn/libplacebo@defc776

That doesn't make sense, why is this the case for HDR10+ but not for DV?

You can see here that MPV tonemapping is overtracking the Max and undertracking the Avg compared to HDR10+ metadata, how is it worse?

[llyyr]

HDR10+ metadata was garbage in most cases back when that commit was made, maybe reality is different now and we need to reconsider using it

[Tjoma96]

HDR10+ metadata was garbage in most cases back when that commit was made, maybe reality is different now and we need to reconsider using it

I see, is there an option I can use in the meantime to enable computation of HDR10+ metadata?

[llyyr]

--tone-mapping=st2094-40 should do it, if I understand that commit right.

[kasper93]

--hdr-compute-peak=no

[Tjoma96]

--hdr-compute-peak=no

That did the trick, thanks. Will just have to manually disable hdr-compute-peak for HDR10+ content then until it gets reintegrated into auto. Used the same scene as earlier and it is detecting ~500 nits less peak using HDR10+ metadata than it did with peak detection which is a lot more accurate.