September 2022 - FFmpeg-trac - lists.ffmpeg.org

Re: [FFmpeg-trac] #2676(avformat:closed): FFprobe doesn't detect cover art for .tta file having id3v2 metadata
by FFmpeg 04 Sep '22

04 Sep '22

#2676: FFprobe doesn't detect cover art for .tta file having id3v2 metadata ------------------------------------+------------------------------------ Reporter: cyril | Owner: (none) Type: defect | Status: closed Priority: normal | Component: avformat Version: git-master | Resolution: fixed Keywords: tta id3 | Blocked By: Blocking: | Reproduced by developer: 1 Analyzed by developer: 0 | ------------------------------------+------------------------------------ Comment (by Balling): There is still a problem if you rename the file to .mp3. -- Ticket URL: <https://trac.ffmpeg.org/ticket/2676#comment:5> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 0

#9903(avformat:new): hls demux emits eof packets on some tls/ssl errors
by FFmpeg 04 Sep '22

04 Sep '22

#9903: hls demux emits eof packets on some tls/ssl errors ----------------------------------+-------------------------------------- Reporter: sylware | Type: defect Status: new | Priority: normal Component: avformat | Version: git-master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | ----------------------------------+-------------------------------------- My IAP (french "free") hls peering with twitch.tv is accute trash (was fine with SFR IAP). While reading twitch.tv streams, the hls demuxer will generates tons of tls/ssl errors, but with some streams, not all, some tls/ssl errors will generate empty/EOF packets in the middle of valid packets. Then when I send valid packets after those spurious(?) empty/EOF packets I get the error: non null packet after EOF packet. Since it is a network quality issue, as a workaround, I do "clean" those empty/EOF packets from the packets queues(video and audio) before trying to read from the hls demuxer again after an error, and playback seems more than fine, namely I do not need to reset/flush the decoders. This is nasty, because it would seems that if I reset/flush the decoders I will get probably choppy playback. It feels like normal and fine playback is expected to work around in some specific way those network errors. -- Ticket URL: <https://trac.ffmpeg.org/ticket/9903> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#4773(undetermined:new): Multidimensional Quantization needs aeons
by FFmpeg 03 Sep '22

03 Sep '22

#4773: Multidimensional Quantization needs aeons -------------------------------------+------------------------------------- Reporter: xtemp09 | Type: defect Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: flac | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- When multi_dim_quant is on, ffmpeg requires too many days to convert. Could you add multithreading and/or openCL support? -- Ticket URL: <https://trac.ffmpeg.org/ticket/4773> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 4

#9366(undetermined:new): [libtls] twitch broke libressl libtls backend
by FFmpeg 03 Sep '22

03 Sep '22

#9366: [libtls] twitch broke libressl libtls backend -------------------------------------+------------------------------------- Reporter: sylware | Type: defect Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- twitch did something yesterday and last week git ffmpeg compiled with a libressl 3.3.3 stop being able to read their hls streams. if I compile this ffmpeg with libressl 3.3.3 but enabling the tls_openssl backend instead of the tls_libtls backend, then I can read their hls streams again. ffmpeg -i "$(youtube-dl -g https://www.twitch.tv/monstercat)" -c copy /tmp/test.ts -- Ticket URL: <https://trac.ffmpeg.org/ticket/9366> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 1

#9687(avfilter:new): vf_yadif_videotoolbox.metal is always recompiled when running make (install/test)
by FFmpeg 01 Sep '22

01 Sep '22

#9687: vf_yadif_videotoolbox.metal is always recompiled when running make (install/test) -------------------------------------+------------------------------------- Reporter: Hanspeter | Type: defect Niederstrasser | Status: new | Priority: normal Component: avfilter | Version: git- | master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: At the end of `make`, the generated files belonging to vf_yadif_videotoolbox seem to be removed, triggering their rebuild along with libavfilter.dylib when `make install` or `make test` are run. How to reproduce: {{{ $ ./configure <options> $ make <lots of compilation> mkdir -p libavfilter/x86/ gcc -I. -Isrc/ -D_ISOC99_SOURCE -std=c99 -Wall -O3 -MMD -MF ffbuild/bin2c_host.d -MT ffbuild/bin2c_host.o -c -o ffbuild/bin2c_host.o src/ffbuild/bin2c.c xcrun -sdk macosx metal /sw/build.build/libavcodec59-5.0-shlibs-5.0-1/ffmpeg-5.0/libavfilter/metal/vf_yadif_videotoolbox.metal -o libavfilter/metal/vf_yadif_videotoolbox.metal.air <more compiling until the end> strip -x -o ffprobe ffprobe_g strip -x -o ffmpeg ffmpeg_g rm libavfilter/metal/vf_yadif_videotoolbox.metal.air libavfilter/metal/vf_yadif_videotoolbox.metallib libavfilter/metal/vf_yadif_videotoolbox.metallib.c $ make install (or make test) xcrun -sdk macosx metal /sw/build.build/libavcodec59-5.0-shlibs-5.0-1/ffmpeg-5.0/libavfilter/metal/vf_yadif_videotoolbox.metal -o libavfilter/metal/vf_yadif_videotoolbox.metal.air xcrun -sdk macosx metallib --split-module-without-linking libavfilter/metal/vf_yadif_videotoolbox.metal.air -o libavfilter/metal/vf_yadif_videotoolbox.metallib ffbuild/bin2c libavfilter/metal/vf_yadif_videotoolbox.metallib libavfilter/metal/vf_yadif_videotoolbox.metallib.c vf_yadif_videotoolbox_metallib gcc -I. -Isrc/ -I/sw/include -D_ISOC99_SOURCE -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -I/sw/build.build/libavcodec59-5.0-shlibs-5.0-1/ffmpeg-5.0/compat/dispatch_semaphore -DPIC -DZLIB_CONST -DHAVE_AV_CONFIG_H -DBUILDING_avfilter -fno-stack-check -I/sw/lib/libspeex1/include -I/sw/include -I/opt/X11/include -std=c11 -Werror=partial-availability -fomit-frame-pointer -fPIC -I/opt/X11/include -I/opt/X11/include -I/opt/X11/include -pthread -I/sw/include -I/sw/include -I/sw/include/libxml2 -I/sw/include -I/sw/include/freetype2 -I/sw/include -I/sw/include/bs2b -I/sw/include -I/sw/include -I/sw/include/freetype2 -I/sw/include/fribidi -I/sw/include -I/sw/include -I/sw/include -I/sw/include/openjpeg-2.4 -I/sw/include -I/sw/include/opus -I/sw/include/opus -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/include -I/sw/lib/libcdio-2.1/include -I/opt/X11/include -I/opt/X11/include -I/opt/X11/include -I/opt/X11/include -g -Wdeclaration-after-statement -Wall -Wdisabled-optimization -Wpointer-arith -Wredundant-decls -Wwrite- strings -Wtype-limits -Wundef -Wmissing-prototypes -Wstrict-prototypes -Wempty-body -Wno-parentheses -Wno-switch -Wno-format-zero-length -Wno- pointer-sign -Wno-unused-const-variable -Wno-char-subscripts -O3 -fno- math-errno -fno-signed-zeros -mstack-alignment=16 -Qunused-arguments -Werror=implicit-function-declaration -Werror=missing-prototypes -Werror =return-type -I/sw/include/SDL2 -D_THREAD_SAFE -MMD -MF libavfilter/metal/vf_yadif_videotoolbox.metallib.d -MT libavfilter/metal/vf_yadif_videotoolbox.metallib.o -c -o libavfilter/metal/vf_yadif_videotoolbox.metallib.o libavfilter/metal/vf_yadif_videotoolbox.metallib.c sed 's/MAJOR/8/' /sw/build.build/libavcodec59-5.0-shlibs-5.0-1/ffmpeg-5.0/libavfilter/libavfilter.v | tr " " "\n" | sed -n /global:/,/local:/p | grep ";" | tr ";" "\n" | sed -E "s/(.+)/_\1/g" | sed -E "s/(.+[^*])$/\1*/" > libavfilter/libavfilter.ver <libavfilter.dylib gets rebuilt as well> <which triggers rebuild of libavdevice.dylib and executables as well> <finish installing> install -m 644 libavutil/libavutil.pc "/tmp/prefix4/usr/local/ffmpeg- clang/lib/pkgconfig" rm libavfilter/metal/vf_yadif_videotoolbox.metal.air libavfilter/metal/vf_yadif_videotoolbox.metallib }}} Originally found with v5.0 (which is not yet available in the dropdown for bug reports btw) on macOS 10.14.6 but also happens with git-head. -- Ticket URL: <https://trac.ffmpeg.org/ticket/9687> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 1

#3920(ffprobe:new): ffprobe can't parse dash init.mp4
by FFmpeg 01 Sep '22

01 Sep '22

#3920: ffprobe can't parse dash init.mp4 ---------------------------------+-------------------------------------- Reporter: sinkers | Type: defect Status: new | Priority: normal Component: ffprobe | Version: git-master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | ---------------------------------+-------------------------------------- ffprobe seems unable to parse the init.mp4's generated for dash files whereas mediainfo can. Example:ffprobe http://progressive.netshow.ninemsn.com.au/media2/2738061917001/2014/09/3764… ffprobe version N-65977-gd9e2ace Copyright (c) 2007-2014 the FFmpeg developers built on Aug 30 2014 17:48:03 with Apple LLVM version 5.1 (clang-503.0.40) (based on LLVM 3.4svn) configuration: --enable-libx264 --enable-libfdk_aac --enable-nonfree --enable-gpl --enable-ffplay --enable-libfreetype --enable-libx265 libavutil 54. 7.100 / 54. 7.100 libavcodec 56. 1.100 / 56. 1.100 libavformat 56. 3.100 / 56. 3.100 libavdevice 56. 0.100 / 56. 0.100 libavfilter 5. 0.103 / 5. 0.103 libswscale 3. 0.100 / 3. 0.100 libswresample 1. 1.100 / 1. 1.100 libpostproc 53. 0.100 / 53. 0.100 [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7fa82c005200] Could not find codec parameters for stream 0 (Video: h264 (avc1 / 0x31637661), 512x288): unspecified pixel format Consider increasing the value for the 'analyzeduration' and 'probesize' options http://progressive.netshow.ninemsn.com.au/media2/2738061917001/2014/09/3764…: End of file Mediainfo output mediainfo http://progressive.netshow.ninemsn.com.au/media2/2738061917001/2014/09/3764… General Complete name : http://progressive.netshow.ninemsn.com.au/media2/2738061917001/2014/09/3764… Format : MPEG-4 Format profile : Base Media Codec ID : isom File size : 713 Bytes Duration : 5mn 45s Overall bit rate : 17 bps Encoded date : UTC 1904-01-01 00:00:00 Tagged date : UTC 1904-01-01 00:00:00 Video ID : 1 Format : AVC Format/Info : Advanced Video Codec Format profile : Main(a)L2.1 Format settings, CABAC : Yes Format settings, ReFrames : 3 frames Codec ID : avc1 Codec ID/Info : Advanced Video Coding Duration : 5mn 45s Width : 512 pixels Height : 288 pixels Display aspect ratio : 16:9 Frame rate : 23.976 fps Color space : YUV Chroma subsampling : 4:2:0 Bit depth : 8 bits Scan type : Progressive Encoded date : UTC 1904-01-01 00:00:00 Tagged date : UTC 1904-01-01 00:00:00 -- Ticket URL: <https://trac.ffmpeg.org/ticket/3920> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 11

#9862(avfilter:new): Can't render subtitles on a transparent video
by FFmpeg 01 Sep '22

01 Sep '22

#9862: Can't render subtitles on a transparent video -----------------------------------+--------------------------------------- Reporter: fduck | Type: defect Status: new | Priority: normal Component: avfilter | Version: unspecified Keywords: subtitles | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -----------------------------------+--------------------------------------- Rendering subtitles on a transparent video track (with the alpha option set to 1) results in subtitles that don't disappear when they should. Later subtitles are drawn right over earlier. {{{ ffmpeg -y\ -ss 0 -t 15 -i 'video.mov'\ -filter_complex " nullsrc=size=1080x1080 [subtitles]; [subtitles] subtitles=subtitles.srt:alpha=1 [subtitles]; [0][subtitles] overlay=x=0:y=0:shortest=1 [out] "\ -map [out] \ '../test.mp4' }}} ffmpeg version 5.1 built with Apple clang version 13.1.6 (clang-1316.0.21.2.5) -- Ticket URL: <https://trac.ffmpeg.org/ticket/9862> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 4

#9871(avformat:new): AccessViolationException setting seekable = 1
by FFmpeg 01 Sep '22

01 Sep '22

#9871: AccessViolationException setting seekable = 1 -------------------------------------+------------------------------------- Reporter: sean | Type: defect tearney | Status: new | Priority: normal Component: avformat | Version: git- | master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: How to reproduce: {{{ var outputFormat = av_guess_format("mp4", @"D:\test.mp4", null); var formatContext = avformat_alloc_context(); formatContext->oformat = outputFormat; var videoStream = avformat_new_stream(formatContext, null); var videoCodec = avcodec_find_encoder(AVCodecID.AV_CODEC_ID_H264); var videoCodecContext = avcodec_alloc_context3(videoCodec); avcodec_open2(videoCodecContext, videoCodec, null); avio_open(&formatContext->pb, @"D:\test.mp4", AVIO_FLAG_WRITE); formatContext->pb->seekable = 1; avformat_write_header(formatContext, null); avio_close(formatContext->pb); }}} setting formatContext->pb->seekable = 1 is throwing an access violation exception when calling avio_close. comment out that line and avio_close works fine. -- Ticket URL: <https://trac.ffmpeg.org/ticket/9871> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 4

#9900(avcodec:new): Native amrnb decoder bug
by FFmpeg 01 Sep '22

01 Sep '22

#9900: Native amrnb decoder bug -------------------------------------+------------------------------------- Reporter: | Type: defect Aleksoid1978 | Status: new | Priority: normal Component: avcodec | Version: | unspecified Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: How to reproduce: {{{ ffmpeg.exe -i amrnb_Fast_decode_issue.3gp amrnb_Fast_decode_issue.mp4 ffplay amrnb_Fast_decode_issue.mp4 - fast playback. ffmpeg.exe -c:a libopencore_amrnb -i amrnb_Fast_decode_issue.3gp amrnb_Fast_decode_issue.mp4 ffplay amrnb_Fast_decode_issue.mp4 - normal playback. ffmpeg version 2022-08-22-git-f23e3ce858-full_build-www.gyan.dev Copyright (c) 2000-2022 the FFmpeg developers built with gcc 12.1.0 (Rev2, Built by MSYS2 project) }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/9900> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 2

#9899(undetermined:new): silenceremove bug
by FFmpeg 01 Sep '22

01 Sep '22

#9899: silenceremove bug -------------------------------------+------------------------------------- Reporter: Jozef | Type: defect Chutka | Status: new | Priority: normal Component: | Version: undetermined | unspecified Keywords: | Blocked By: silenceremove | Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- using ffmpeg version 2022-07-31-git-1368b5a725-full_build-www.gyan.dev `silenceremove` seems to remove whole audio content in some cases. Having supershort `adelay` before `silenceremove` serves as a fix. {{{ ffmpeg -i input -af silenceremove=start_periods=1:stop_periods=1:detection=peak out-empty.m4a -y ffmpeg -i input -af adelay=delays=1:all=1,silenceremove=start_periods=1:stop_periods=1:detection=peak out-ok.m4a -y }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/9899> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 5