May 2025 - FFmpeg-trac - lists.ffmpeg.org

#11572(undetermined:new): HEVC interlaced reported as progressive; wrong DAR
by FFmpeg 02 May '25

02 May '25

#11572: HEVC interlaced reported as progressive; wrong DAR -------------------------------------+------------------------------------- Reporter: Peter | Type: defect Krefting | Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: hevc | Blocked By: interlaced | Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: A HEVC broadcast video stream encoded by the Makito encoder configured to use HEVC 1920×1080i video is identified as 1920×540p. How to reproduce: The following captures were provided to us by the broadcaster, and I have been given permission to share them for the purpose of this bug report. They were created by the Makito encoder software: * [https://www.dropbox.com/scl/fi/77pmthkf3uv3zf3a4yjfk/hevc- interlace-1.ts?rlkey=ray5xnepc2lwuzzomccujiaqf&dl=1 Sample 1] * [https://www.dropbox.com/scl/fi/oqag1b7f8t8lqettjolst/hevc- interlace-2.ts?rlkey=pp1wah7ep32pbx355tcbszb7v&dl=1 Sample 2] Running ffprobe on the samples detects them as 1920×540p with a 32:9 DAR, and playing them with ffplay do display them in 32:9. {{{ $ ffprobe /tmp/hevc-interlace-2.ts ffprobe version N-119410-g33d0d1c672 Copyright (c) 2007-2025 the FFmpeg developers built with gcc 12 (Debian 12.2.0-14) configuration: libavutil 60. 2.100 / 60. 2.100 libavcodec 62. 2.100 / 62. 2.100 libavformat 62. 0.102 / 62. 0.102 libavdevice 62. 0.100 / 62. 0.100 libavfilter 11. 0.100 / 11. 0.100 libswscale 9. 0.100 / 9. 0.100 libswresample 6. 0.100 / 6. 0.100 [hevc @ 0x55a556e149c0] PPS id out of range: 0 [hevc @ 0x55a556e149c0] Skipping invalid undecodable NALU: 39 [hevc @ 0x55a556e149c0] PPS id out of range: 0 [hevc @ 0x55a556e149c0] Skipping invalid undecodable NALU: 0 [hevc @ 0x55a556e149c0] PPS id out of range: 0 [hevc @ 0x55a556e149c0] Skipping invalid undecodable NALU: 39 [hevc @ 0x55a556e149c0] PPS id out of range: 0 [hevc @ 0x55a556e149c0] Skipping invalid undecodable NALU: 1 [hevc @ 0x55a556e149c0] PPS id out of range: 0 (...messages repeats...) [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 0, dts = 6953319369). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 1, dts = 6953274618). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 2, dts = 6953274618). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 3, dts = 6953274614). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 4, dts = 6953274614). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 5, dts = 6953274614). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 6, dts = 6953274622). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 7, dts = 6953274618). [mpegts @ 0x55a556e0ea00] PES packet size mismatch [mpegts @ 0x55a556e0ea00] Packet corrupt (stream = 8, dts = 6953274614). Input #0, mpegts, from '/tmp/hevc-interlace-2.ts': Duration: 00:01:01.53, start: 77197.593489, bitrate: 9267 kb/s Program 1 Metadata: service_name : (redacted; run on sample to see) service_provider: (redacted; run on sample to see) Stream #0:0[0x79]: Video: hevc (Main 10) ([36][0][0][0] / 0x0024), yuv420p10le(tv, bt709, progressive), 1920x540 [SAR 1:1 DAR 32:9], 59.94 fps, 59.94 tbr, 90k tbn, Start-Time 77197.709s Stream #0:1[0xdd](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.594s Stream #0:2[0xde](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.594s Stream #0:3[0xdf](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.593s Stream #0:4[0xe0](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.593s Stream #0:5[0xe1](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.593s Stream #0:6[0xe2](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.594s Stream #0:7[0xe3](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.594s Stream #0:8[0xe4](und): Audio: aac (LC) ([15][0][0][0] / 0x000F), 48000 Hz, stereo, fltp, 129 kb/s, Start-Time 77197.593s }}} When feeding frames directly to the codec, it does look like the AV_FRAME_FLAG_INTERLACED bit is set correctly in the returned frame. According to the provider of the sample, the Makito encoder (the version number was not indicated) is configured as follows: * Codec Algorithm: HEVC/H.265 * Resolution: 1920×1080i * GOP Size: 60 * Encoding Profile: Main 10 * Resizing: Scale * Closed Captioning: on * Chroma Subsampling: 4:2:0 10-bit * Frame Rate: Automatic * Transcode Source: System * Rate Control: CBR * Framing: IBBBBP * Counting Mode: UTC Conversion * Bitrate: 6000 * Aspect Ratio: Automatic * Scene Change Detection: off * Partial Image Skip: on * Intra Refresh: off -- Ticket URL: <https://trac.ffmpeg.org/ticket/11572> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 1

#11569(documentation:new): A correction to the docs
by FFmpeg 02 May '25

02 May '25

#11569: A correction to the docs -------------------------------------+------------------------------------- Reporter: idest | Type: defect Status: new | Priority: normal Component: | Version: documentation | unspecified Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- In the documentation to the ffmpeg tool, at the end of the -aspect option description there's a phrase: >If used together with -vcodec copy, it will affect the aspect ratio stored at container level, but not the aspect ratio stored in encoded frames, if it exists. Here "stored at ''container'' level" should probably be replaced with "stored at ''video stream metadata'' level". ffmpeg N-118724-g73f4668ef8-20250310. -- Ticket URL: <https://trac.ffmpeg.org/ticket/11569> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 9

#11563(undetermined:new): Implement better jpeg compression now that patents on those two algorithms expired
by FFmpeg 01 May '25

01 May '25

#11563: Implement better jpeg compression now that patents on those two algorithms expired -------------------------------------+------------------------------------- Reporter: Balling | Type: | enhancement Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: mjpeg | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: bigger bug is here "Image specific quantization tables" which can source code that should compile just fine, https://pages.cs.wisc.edu/~ratnakar/rdopt.html and there is even a better algorithm that doubles average size savings: Joint Optimization of Run-Length Coding, Huffman Coding, and Quantization Table With Complete Baseline JPEG Decoder Compatibility. Unfortunately, it is patent encumbered probably until 2024 (2025 now, btw): https://www.google.com/patents/US7570827 There are also images comparison https://cloud.githubusercontent.com/assets/72159/8814942/99255002-300a- 11e5-8fcd-4f810b0aa78b.jpg vs https://cloud.githubusercontent.com/assets/72159/8814943/992ee824-300a- 11e5-99fe-0f56e3229c84.jpg -- Ticket URL: <https://trac.ffmpeg.org/ticket/11563> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 5

#11566(ffmpeg:new): Arabic subtitles rendering artifact BUG
by FFmpeg 01 May '25

01 May '25

#11566: Arabic subtitles rendering artifact BUG -----------------------------------+-------------------------------------- Reporter: ellat | Type: defect Status: new | Priority: normal Component: ffmpeg | Version: git-master Keywords: subtitles | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -----------------------------------+-------------------------------------- Summary of the bug: How to reproduce: Create the file a.srt: {{{ 1 00:00:00,000 --> 00:00:15,000 الأسد }}} Use a random input video and run this command to burn subtitles: `%./ffmpeg -i input_video.mp4 -vf subtitles=a.srt:force_style='FontName=Arial' out.mp4` A box outline will appear in the middle of the arabic word. {{{ ffmpeg version N-119396-g48c0dba23b-20250430 Copyright (c) 2000-2025 the FFmpeg developers built with gcc 14.2.0 (crosstool-NG 1.27.0.18_7458341) }}} According to: https://bugs.kde.org/show_bug.cgi?id=502725 this issue has existed for 2 years. -- Ticket URL: <https://trac.ffmpeg.org/ticket/11566> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 2