July 2019 - FFmpeg-trac - lists.ffmpeg.org

#8015(avformat:new): hls muxer crash
by FFmpeg 15 Jul '19

15 Jul '19

#8015: hls muxer crash ----------------------------------+-------------------------------------- Reporter: mypopy | Type: defect Status: new | Priority: normal Component: avformat | Version: git-master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | ----------------------------------+-------------------------------------- Used the command as follow: ./ffmpeg -f lavfi -re -i testsrc=s=720x480:r=25:d=100 -vcodec libx264 -r 25 -preset medium -keyint_min 25 -g 25 -f hls -hls_flags +program_date_time -hls_segment_type fmp4 -hls_fmp4_init_filename now.mp4 -use_localtime 1 -hls_list_size 0 -hls_time 6 -hls_segment_filename %Y-%m-%d_%H-%M-%S.m4s now-data.m3u8 will crash with the message like: [hls @ 0x559ea8797000] Opening '2019-07-13_15-27-41.m4s' for writinged=0.775x [hls @ 0x559ea8797000] Opening 'now-data.m3u8.tmp' for writing [hls @ 0x559ea8797000] Opening '2019-07-13_15-27-49.m4s' for writinged=0.871x '''av_interleaved_write_frame(): Invalid argument''' [hls @ 0x559ea8797000] Opening '2019-07-13_15-27-49.m4s' for writinged=0.877x Segmentation fault (core dumped) -- Ticket URL: <https://trac.ffmpeg.org/ticket/8015> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 6

#7117(undetermined:new): DVB subtitles render incorrectly
by FFmpeg 14 Jul '19

14 Jul '19

#7117: DVB subtitles render incorrectly -------------------------------------+------------------------------------- Reporter: leoenc | Type: defect Status: new | Priority: normal Component: | Version: undetermined | unspecified Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: Conversion of sup subtitles (pgssub) to DVB subtitles and muxing to mpeg- ts results in bad looking subtitles. When using the overlay filter instead, they render correctly. I am using the latest nightly build 20180331-be502ec and testing with VLC Player 3.0.1 (Windows). I also tried playing the dvb output with ffplay but subtitles aren't showing but instead the screen is just flashing from black to video. I tried with or without canvas_size 1920x1080 and/or s:s 1920x1080, makes no difference. How to reproduce: I've attached the input video and subtitles needed to reproduce + both outputs I got on my end. Output with DVB subtitles which looks wrong: {{{ ffmpeg -y -i input.mp4 -fix_sub_duration -i subtitles.sup -vcodec libx264 -crf 24 -f mpegts -scodec dvbsub output_dvb.ts ffmpeg version N-90553-gbe502ec6cd Copyright (c) 2000-2018 the FFmpeg developers built with gcc 7.3.0 (GCC) configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable- bzlib --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable- libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable- libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable- nvenc --enable-nvdec --enable-dxva2 --enable-avisynth libavutil 56. 12.100 / 56. 12.100 libavcodec 58. 16.100 / 58. 16.100 libavformat 58. 10.100 / 58. 10.100 libavdevice 58. 2.100 / 58. 2.100 libavfilter 7. 13.100 / 7. 13.100 libswscale 5. 0.102 / 5. 0.102 libswresample 3. 0.101 / 3. 0.101 libpostproc 55. 0.100 / 55. 0.100 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'input.mp4': Metadata: major_brand : isom minor_version : 1 compatible_brands: isomavc1 creation_time : 2018-04-01T16:37:29.000000Z Duration: 00:00:30.12, start: 0.000000, bitrate: 20 kb/s Stream #0:0(und): Video: h264 (High) (avc1 / 0x31637661), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], 18 kb/s, 25 fps, 25 tbr, 100 tbn, 50 tbc (default) Metadata: creation_time : 2018-04-01T16:37:29.000000Z handler_name : 264(a)GPAC0.7.0-rev0-gbd5c9af-master Input #1, sup, from 'subtitles.sup': Duration: N/A, start: 3.440000, bitrate: N/A Stream #1:0: Subtitle: hdmv_pgs_subtitle, 1920x1080 Stream mapping: Stream #0:0 -> #0:0 (h264 (native) -> h264 (libx264)) Stream #1:0 -> #0:1 (hdmv_pgs_subtitle (pgssub) -> dvb_subtitle (dvbsub)) Press [q] to stop, [?] for help [sup @ 00000271594a0140] Invalid timestamps stream=0, pts=0, dts=303570, size=4974 [libx264 @ 00000271594833c0] using SAR=1/1 [libx264 @ 00000271594833c0] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2 [libx264 @ 00000271594833c0] profile High, level 4.0 Output #0, mpegts, to 'output_dvb.ts': Metadata: major_brand : isom minor_version : 1 compatible_brands: isomavc1 encoder : Lavf58.10.100 Stream #0:0(und): Video: h264 (libx264), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], q=-1--1, 25 fps, 90k tbn, 25 tbc (default) Metadata: creation_time : 2018-04-01T16:37:29.000000Z handler_name : 264(a)GPAC0.7.0-rev0-gbd5c9af-master encoder : Lavc58.16.100 libx264 Side data: cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: -1 Stream #0:1: Subtitle: dvb_subtitle (dvbsub), 1920x1080 Metadata: encoder : Lavc58.16.100 dvbsub [sup @ 00000271594a0140] Invalid timestamps stream=0, pts=0, dts=767997, size=6238 [sup @ 00000271594a0140] Invalid timestamps stream=0, pts=0, dts=1135096, size=9926 frame= 150 fps=0.0 q=29.0 size= 27kB time=00:00:06.68 bitrate= 33.1kbits/s sp[sup @ 00000271594a0140] Invalid timestamps stream=0, pts=0, dts=1635183, size=17502 frame= 310 fps=309 q=29.0 size= 80kB time=00:00:11.80 bitrate= 55.6kbits/s sp[sup @ 00000271594a0140] Invalid timestamps stream=0, pts=0, dts=2218393, size=18698 frame= 478 fps=317 q=29.0 size= 133kB time=00:00:18.60 bitrate= 58.7kbits/s spframe= 639 fps=318 q=29.0 size= 187kB time=00:00:24.20 bitrate= 63.1kbits/s spframe= 753 fps=310 q=-1.0 Lsize= 221kB time=00:00:30.00 bitrate= 60.3kbits/s speed=12.3x video:54kB audio:0kB subtitle:65kB other streams:0kB global headers:0kB muxing overhead: 85.350807% [libx264 @ 00000271594833c0] frame I:4 Avg QP: 8.50 size: 642 [libx264 @ 00000271594833c0] frame P:190 Avg QP:10.10 size: 77 [libx264 @ 00000271594833c0] frame B:559 Avg QP:13.67 size: 69 [libx264 @ 00000271594833c0] consecutive B-frames: 0.9% 0.3% 0.0% 98.8% [libx264 @ 00000271594833c0] mb I I16..4: 100.0% 0.0% 0.0% [libx264 @ 00000271594833c0] mb P I16..4: 0.0% 0.0% 0.0% P16..4: 0.0% 0.0% 0.0% 0.0% 0.0% skip:100.0% [libx264 @ 00000271594833c0] mb B I16..4: 0.0% 0.0% 0.0% B16..8: 0.0% 0.0% 0.0% direct: 0.0% skip:100.0% [libx264 @ 00000271594833c0] 8x8 transform intra:0.0% [libx264 @ 00000271594833c0] coded y,uvDC,uvAC intra: 0.0% 0.0% 0.0% inter: 0.0% 0.0% 0.0% [libx264 @ 00000271594833c0] i16 v,h,dc,p: 99% 0% 1% 0% [libx264 @ 00000271594833c0] i8c dc,h,v,p: 100% 0% 0% 0% [libx264 @ 00000271594833c0] Weighted P-Frames: Y:0.0% UV:0.0% [libx264 @ 00000271594833c0] kb/s:14.75 }}} Output with overlay subtitles which looks correct: {{{ ffmpeg -y -i input.mp4 -i subtitles.sup -filter_complex "[1:s]scale=width=1920:height=1080[sub];[0:v][sub]overlay" -vcodec libx264 -crf 24 -f mpegts output_overlay.ts ffmpeg version N-90553-gbe502ec6cd Copyright (c) 2000-2018 the FFmpeg developers built with gcc 7.3.0 (GCC) configuration: --enable-gpl --enable-version3 --enable-sdl2 --enable- bzlib --enable-fontconfig --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libfreetype --enable-libmp3lame --enable- libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-libshine --enable-libsnappy --enable-libsoxr --enable-libtheora --enable-libtwolame --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libzimg --enable-lzma --enable-zlib --enable-gmp --enable- libvidstab --enable-libvorbis --enable-libvo-amrwbenc --enable-libmysofa --enable-libspeex --enable-libxvid --enable-libaom --enable-libmfx --enable-amf --enable-ffnvcodec --enable-cuvid --enable-d3d11va --enable- nvenc --enable-nvdec --enable-dxva2 --enable-avisynth libavutil 56. 12.100 / 56. 12.100 libavcodec 58. 16.100 / 58. 16.100 libavformat 58. 10.100 / 58. 10.100 libavdevice 58. 2.100 / 58. 2.100 libavfilter 7. 13.100 / 7. 13.100 libswscale 5. 0.102 / 5. 0.102 libswresample 3. 0.101 / 3. 0.101 libpostproc 55. 0.100 / 55. 0.100 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'input.mp4': Metadata: major_brand : isom minor_version : 1 compatible_brands: isomavc1 creation_time : 2018-04-01T16:37:29.000000Z Duration: 00:00:30.12, start: 0.000000, bitrate: 20 kb/s Stream #0:0(und): Video: h264 (High) (avc1 / 0x31637661), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], 18 kb/s, 25 fps, 25 tbr, 100 tbn, 50 tbc (default) Metadata: creation_time : 2018-04-01T16:37:29.000000Z handler_name : 264(a)GPAC0.7.0-rev0-gbd5c9af-master Input #1, sup, from 'subtitles.sup': Duration: N/A, start: 3.440000, bitrate: N/A Stream #1:0: Subtitle: hdmv_pgs_subtitle, 1920x1080 Stream mapping: Stream #0:0 (h264) -> overlay:main Stream #1:0 (pgssub) -> scale overlay -> Stream #0:0 (libx264) Press [q] to stop, [?] for help [sup @ 00000292aeae0600] Invalid timestamps stream=0, pts=0, dts=303570, size=4974 [sup @ 00000292aeae0600] Invalid timestamps stream=0, pts=0, dts=767997, size=6238 [sup @ 00000292aeae0600] sub2video: using 720x576 canvas [libx264 @ 00000292ae6e2700] using SAR=1/1 [libx264 @ 00000292ae6e2700] using cpu capabilities: MMX2 SSE2Fast SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2 [libx264 @ 00000292ae6e2700] profile High, level 4.0 Output #0, mpegts, to 'output_overlay.ts': Metadata: major_brand : isom minor_version : 1 compatible_brands: isomavc1 encoder : Lavf58.10.100 Stream #0:0: Video: h264 (libx264), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], q=-1--1, 25 fps, 90k tbn, 25 tbc (default) Metadata: encoder : Lavc58.16.100 libx264 Side data: cpb: bitrate max/min/avg: 0/0/0 buffer size: 0 vbv_delay: -1 frame= 57 fps=0.0 q=29.0 size= 7kB time=-00:00:00.07 bitrate=N/A speed=N/A frame= 123 fps=123 q=29.0 size= 20kB time=00:00:02.56 bitrate= 62.9kbits/s sp[sup @ 00000292aeae0600] Invalid timestamps stream=0, pts=0, dts=1135096, size=9926 frame= 189 fps=126 q=29.0 size= 40kB time=00:00:05.20 bitrate= 63.6kbits/s sp[sup @ 00000292aeae0600] Invalid timestamps stream=0, pts=0, dts=1635183, size=17502 frame= 255 fps=127 q=29.0 size= 53kB time=00:00:07.84 bitrate= 55.2kbits/s spframe= 322 fps=128 q=29.0 size= 78kB time=00:00:10.52 bitrate= 60.5kbits/s sp[sup @ 00000292aeae0600] Invalid timestamps stream=0, pts=0, dts=2218393, size=18698 frame= 388 fps=129 q=29.0 size= 91kB time=00:00:13.16 bitrate= 56.5kbits/s spframe= 454 fps=129 q=29.0 size= 125kB time=00:00:15.80 bitrate= 64.6kbits/s spframe= 518 fps=129 q=29.0 size= 137kB time=00:00:18.36 bitrate= 61.2kbits/s spframe= 583 fps=129 q=29.0 size= 149kB time=00:00:20.96 bitrate= 58.4kbits/s spframe= 649 fps=129 q=29.0 size= 184kB time=00:00:23.60 bitrate= 64.0kbits/s spframe= 716 fps=129 q=29.0 size= 198kB time=00:00:26.28 bitrate= 61.6kbits/s spframe= 754 fps=127 q=-1.0 Lsize= 216kB time=00:00:30.04 bitrate= 58.8kbits/s speed=5.07x video:116kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 85.782951% [libx264 @ 00000292ae6e2700] frame I:5 Avg QP: 8.08 size: 12890 [libx264 @ 00000292ae6e2700] frame P:190 Avg QP: 9.85 size: 81 [libx264 @ 00000292ae6e2700] frame B:559 Avg QP:13.61 size: 70 [libx264 @ 00000292ae6e2700] consecutive B-frames: 1.1% 0.3% 0.0% 98.7% [libx264 @ 00000292ae6e2700] mb I I16..4: 95.4% 3.1% 1.5% [libx264 @ 00000292ae6e2700] mb P I16..4: 0.0% 0.0% 0.0% P16..4: 0.0% 0.0% 0.0% 0.0% 0.0% skip:100.0% [libx264 @ 00000292ae6e2700] mb B I16..4: 0.0% 0.0% 0.0% B16..8: 0.0% 0.0% 0.0% direct: 0.0% skip:100.0% L0:42.3% L1:57.7% BI: 0.0% [libx264 @ 00000292ae6e2700] 8x8 transform intra:3.0% inter:6.6% [libx264 @ 00000292ae6e2700] coded y,uvDC,uvAC intra: 1.0% 1.5% 1.5% inter: 0.0% 0.0% 0.0% [libx264 @ 00000292ae6e2700] i16 v,h,dc,p: 98% 0% 2% 0% [libx264 @ 00000292ae6e2700] i8 v,h,dc,ddl,ddr,vr,hd,vl,hu: 55% 4% 41% 0% 0% 0% 0% 0% 0% [libx264 @ 00000292ae6e2700] i4 v,h,dc,ddl,ddr,vr,hd,vl,hu: 43% 18% 14% 3% 4% 6% 5% 4% 3% [libx264 @ 00000292ae6e2700] i8c dc,h,v,p: 99% 1% 0% 0% [libx264 @ 00000292ae6e2700] Weighted P-Frames: Y:0.0% UV:0.0% [libx264 @ 00000292ae6e2700] ref P L0: 90.5% 6.0% 3.6% [libx264 @ 00000292ae6e2700] ref B L0: 75.5% 24.5% [libx264 @ 00000292ae6e2700] ref B L1: 98.5% 1.5% [libx264 @ 00000292ae6e2700] kb/s:0.00 }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/7117> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 5

#8007(avfilter:new): bsf_dump_extra repeat tables
by FFmpeg 14 Jul '19

14 Jul '19

#8007: bsf_dump_extra repeat tables -------------------------------------+------------------------------------- Reporter: atorp | Type: defect Status: new | Priority: normal Component: avfilter | Version: git- Keywords: | master bsf_dump_extra | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: When using the "bsf_dump_extra" some tables are repeated. How to reproduce: {{{ % ffmpeg -i input.ts -c:v mpeg2_qsv -bsf:v dump_extra out.ts }}} With this example, the resulting "out.ts" file has an incorrect header in the first I frame. In this frame the PES tables "SEQUENCE HEADER" and "SEQUENCE EXTENSION" are repeated. And that's a clear bug! The problem is based on this: The MPEG-2 QSV encoder doesn't repeat these tables. So they only appear in the bitstream in the first I frame. Then every new GOP lacks the tables, and as result the bitstream is not suitable for Broadcasting: If you start to play after the first GOP you can't reproduce the stream. The solution is then to use the bitstream filter "dump_extra". This filter repeats the required tables at every GOP. But..., the current implementation doesn't check for the existence of a table, and always inserts it... even if they already exists. So, please can you review the code and fix this bug? The idea is to check the existence of a table before inserting it. Thank you! -- Ticket URL: <https://trac.ffmpeg.org/ticket/8007> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 2

#7915(undetermined:new): [ffmpeg-vaapi] jpeg encode failed when build with gcc 9.1.1 on Fedora 30
by FFmpeg 14 Jul '19

14 Jul '19

#7915: [ffmpeg-vaapi] jpeg encode failed when build with gcc 9.1.1 on Fedora 30 -------------------------------------+------------------------------------- Reporter: zchrzhou | Type: defect Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: vaapi | Blocked By: encode jpeg | Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: On Fedora 30, The output of ffmpeg vaapi jpeg encode is not correct both with iHD and i965 driver, it gets green output when playback with ffplay. The same codes (including libva,msdk,gmmlib,iHD/i965 driver and ffmpeg) work well on Fedora 29 built with gcc 8.3.1 gcc version is 9.1.1 on Fedora 30 gcc version is 8.3.1 on Fedora 29 libva 520b9e54f95d libva-utils 7be46f017cae intel-vaapi-driver 7b859a9b143f gmmlib 8feb8385a3c9 intel-media-driver 37f9922543d7 msdk 98ff579a9c4f ffmpeg 58d167bcd517 How to reproduce: {{{ % ffmpeg -hwaccel vaapi -vaapi_device /dev/dri/renderD128 -v verbose -f rawvideo -pix_fmt yuv420p -s:v 1280x720 -i /opt/media/src/assets/otc- media/yuv/stockholm_1280x720_604_1280x720_I420.yuv -vf 'format=nv12,hwupload' -c:v mjpeg_vaapi -global_quality 75 -vframes 10 -y /opt/media/src/vaapi-fits-smoke/results/smoke/6ed45286-7853-11e9-b16d- 0242ac110002_0/_0.test.ffmpeg-vaapi.encode.jpeg/cqp/stockholm-cqp- baseline-75.mjpeg % ffplay stockholm-cqp-baseline-75.mjpeg % ffmpeg -v ffmpeg version N-93863-g58d167bcd5 Copyright (c) 2000-2019 the FFmpeg developers built with gcc 8 (GCC) configuration: --prefix=/opt/media/install --disable-static --enable- shared --enable-libdrm --enable-vaapi --enable-libmfx --disable-amf --disable-audiotoolbox --disable-cuda --disable-cuda-sdk --disable-cuvid --disable-d3d11va --disable-dxva2 --disable-libnpp --disable-mmal --disable-nvdec --disable-nvenc --disable-omx --disable-omx-rpi --disable- rkmpp --disable-v4l2-m2m --disable-vdpau --disable-videotoolbox --enable- gpl --enable-libx264 --enable-libx265 libavutil 56. 27.100 / 56. 27.100 libavcodec 58. 52.101 / 58. 52.101 libavformat 58. 27.103 / 58. 27.103 libavdevice 58. 7.100 / 58. 7.100 libavfilter 7. 53.100 / 7. 53.100 libswscale 5. 4.101 / 5. 4.101 libswresample 3. 4.100 / 3. 4.100 libpostproc 55. 4.100 / 55. 4.100 Missing argument for option 'v'. Error splitting the argument list: Invalid argument }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/7915> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 10

#8019(avformat:new): Failure to properly find syncword in ADTS after desync
by FFmpeg 14 Jul '19

14 Jul '19

#8019: Failure to properly find syncword in ADTS after desync ----------------------------------+-------------------------------------- Reporter: ePirat | Type: defect Status: new | Priority: normal Component: avformat | Version: git-master Keywords: ADTS | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | ----------------------------------+-------------------------------------- Summary of the bug: How to reproduce: {{{ % ffplay aac-denbosch.aac ffmpeg version 4.1.3 Copyright (c) 2000-2019 the FFmpeg developers built with Apple LLVM version 10.0.1 (clang-1001.0.46.3) }}} Leave it play for a while (about 25 seconds) and it will fail to successfully play further, even though the file contains more data. Thats because a de-sync happens and ffmpeg does not properly syncs again on a syncword, so it get garbage data for ADTS headers. -- Ticket URL: <https://trac.ffmpeg.org/ticket/8019> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 2

#7963(ffmpeg:new): Concat demuxer adds time to video duration
by FFmpeg 13 Jul '19

13 Jul '19

#7963: Concat demuxer adds time to video duration -------------------------------------+------------------------------------- Reporter: bgx90 | Type: defect Status: new | Priority: normal Component: ffmpeg | Version: 4.1 Keywords: concat | Blocked By: demuxer protocol concatenate | duration mkv | Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- I concatenate some ts files with the concat demuxer and create a mkv file using stream copying with the following command: {{{ ffmpeg -f concat -safe 0 -fflags +genpts -i file_list.txt -c copy output.mkv }}} The result is a playable mkv, but the duration is 8 seconds longer than the duration of the original ts stream. Also, the picture periodically freezes for a very small fraction of a second, but I don't detect any pause in the audio; I suspect these stutters are the source of the extra 8 seconds. I have tried the command without the +genpts flag as well, but there is no change in the resulting mkv. Concatenating the ts files with the concat protocol results in an mkv with the correct duration and no pauses in the picture during playback. This command works as expected: {{{ ffmpeg -fflags +genpts -i "concat:$file_list" -c copy output.mkv }}} Running ''ffprobe -show_format -show_streams'' on both of the files and comparing the outputs shows only the following differences: Concat demuxer mkv {{{ duration=1449.811000 size=280861150 bit_rate=1549780 }}} Concat protocol mkv {{{ duration=1441.727000 size=280861151 bit_rate=1558470 }}} I read in ticket [https://trac.ffmpeg.org/ticket/2801 2801] that the concat demuxer needs the files to be split cleanly on keyframes. Could this problem be caused by bad splitting of the ts files? I don't mind using the concat protocol instead of the concat demuxer. I only worry about the limit on command length. It was suggested in ticket [https://trac.ffmpeg.org/ticket/5686 5686] that the cat command could be used in cases where the file list is too long and the concat protocol works but the concat demuxer does not. Is using the cat command to concatenate ts files and giving the resulting file as input to ffmpeg the same as using the concat protocol, or does the concat protocol do something more than just concatenating files as the cat command does? -- Ticket URL: <https://trac.ffmpeg.org/ticket/7963> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#8016(undetermined:new): ismindex doesn't handle input filenames or urls containing '&' characters correctly.
by FFmpeg 13 Jul '19

13 Jul '19

#8016: ismindex doesn't handle input filenames or urls containing '&' characters correctly. --------------------------------------+---------------------------------- Reporter: jsyrjala | Type: defect Status: new | Priority: normal Component: undetermined | Version: 4.1 Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | --------------------------------------+---------------------------------- `ismindex` doesn't handle input filenames or urls containing `&` characters correctly. Running command: {{{ ismindex -n foobar 'foo&bar.ismv' }}} Creates this 'foobar.ism' file with following contents: {{{ <?xml version="1.0" encoding="utf-8"?> <smil xmlns="http://www.w3.org/2001/SMIL20/Language"> <head> <meta name="clientManifestRelativePath" content="foobar.ismc" /> </head> <body> <switch> <video src="foo&bar.ismv" systemBitrate="307228"> <param name="trackID" value="1" valueType="data" /> </video> </switch> </body> </smil> }}} Notice `src="foo&bar.ismv"` in `<video>` tag. This is not valid XML. `&` and some other characters should be escaped. See: https://stackoverflow.com/questions/5709232/how-do-i-include-etc-in- xml-attribute-values So the value should be `src="foo&bar.ism"` according to XML specification. This causes issues in real world apps when you are using URLs as input. For example this kind of command would also create invalid `.ism` file. {{{ ismindex -n foobar https://example.com/something?file=foobar&type=ismv }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/8016> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#6368(avformat:new): audio_preload broken
by FFmpeg 12 Jul '19

12 Jul '19

#6368: audio_preload broken -------------------------------------+------------------------------------- Reporter: alex3d | Type: defect Status: new | Priority: normal Component: avformat | Version: git- Keywords: | master audio_preload | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- When audio_preload option is set output media stream a/v interleaving is broken because of "wrong sign issue" in interleave_compare_dts. I'm willing to provide patch solving this issue and also enabling audio stream shift in both directions (currently audio_preload is 0-int_max). Probably audio_preload can be renamed to something like audio_shift? (it seems no one uses audio_preload option because of this issue) libavformat/mux.c: (master branch) {{{ static int interleave_compare_dts(AVFormatContext *s, AVPacket *next, AVPacket *pkt) { AVStream *st = s->streams[pkt->stream_index]; AVStream *st2 = s->streams[next->stream_index]; int comp = av_compare_ts(next->dts, st2->time_base, pkt->dts, st->time_base); if (s->audio_preload && ((st->codecpar->codec_type == AVMEDIA_TYPE_AUDIO) != (st2->codecpar->codec_type == AVMEDIA_TYPE_AUDIO))) { int64_t ts = av_rescale_q(pkt ->dts, st ->time_base, AV_TIME_BASE_Q) - s->audio_preload*(st ->codecpar->codec_type == AVMEDIA_TYPE_AUDIO); int64_t ts2= av_rescale_q(next->dts, st2->time_base, AV_TIME_BASE_Q) - s->audio_preload*(st2->codecpar->codec_type == AVMEDIA_TYPE_AUDIO); if (ts == ts2) { ts= ( pkt ->dts* st->time_base.num*AV_TIME_BASE - s->audio_preload*(int64_t)(st ->codecpar->codec_type == AVMEDIA_TYPE_AUDIO)* st->time_base.den)*st2->time_base.den -( next->dts*st2->time_base.num*AV_TIME_BASE - s->audio_preload*(int64_t)(st2->codecpar->codec_type == AVMEDIA_TYPE_AUDIO)*st2->time_base.den)* st->time_base.den; ts2=0; } comp= (ts>ts2) - (ts<ts2); ///////// there should be (ts2>ts) - (ts2<ts) obviously } if (comp == 0) return pkt->stream_index < next->stream_index; return comp > 0; } }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/6368> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#8014(ffmpeg:new): Piping mjpeg drops frames
by FFmpeg 12 Jul '19

12 Jul '19

#8014: Piping mjpeg drops frames -------------------------------------+------------------------------------- Reporter: sig | Type: defect Status: new | Priority: normal Component: ffmpeg | Version: git- Keywords: h264 mjpeg | master pipe dropped frames | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- It seems switching output from jpeg image files to a pipe results in frames not being outputted to the pipe. Attached is 100-frames.mp4 which is a video of 100 sequentially numbered frames. As expected running the below outputs all 100 frames correctly. {{{ ffmpeg -i 100-frames.mp4 -f image2 -vcodec mjpeg "out%05d.jpg" }}} However running this next command results in only 53 frames being detected: {{{ ffmpeg -i 100-frames.mp4 -f image2 -vcodec mjpeg -update 1 pipe:1 | xxd | grep "ffd8" | wc -l }}} This one uses xxd to convert the outputted binary into hex then uses grep to search for the jpeg start bytes ("ffd8"). Finally wc counts the search hits and prints the total at the end. So the number output at the end is the number of JPEGs seen in the stream. -- Ticket URL: <https://trac.ffmpeg.org/ticket/8014> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 7

#5752(avformat:new): Crash when muxing webm_chunk
by FFmpeg 12 Jul '19

12 Jul '19

#5752: Crash when muxing webm_chunk -------------------------------------+------------------------------------- Reporter: cehoyos | Owner: Type: defect | Status: new Priority: important | Component: avformat Version: git- | Keywords: crash master | SIGSEGV mkv Blocked By: | Blocking: Reproduced by developer: 0 | Analyzed by developer: 0 -------------------------------------+------------------------------------- http://ffmpeg.org/pipermail/ffmpeg-user/2016-August/033006.html A user provided ac3 audio that crashes the webm_chunk muxer. {{{ (gdb) r -i out.ac3 -map 0 -strict -2 -acodec vorbis -ac 2 -header chk/header chk/out%d.chk Starting program: ffmpeg_g -i out.ac3 -map 0 -strict -2 -acodec vorbis -ac 2 -header chk/header chk/out%d.chk [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". ffmpeg version N-81244-g8916ad9 Copyright (c) 2000-2016 the FFmpeg developers built with gcc 4.7 (SUSE Linux) configuration: --enable-libvpx --enable-libvorbis libavutil 55. 28.100 / 55. 28.100 libavcodec 57. 51.100 / 57. 51.100 libavformat 57. 44.100 / 57. 44.100 libavdevice 57. 0.102 / 57. 0.102 libavfilter 6. 49.100 / 6. 49.100 libswscale 4. 1.100 / 4. 1.100 libswresample 2. 1.100 / 2. 1.100 [ac3 @ 0x1e31560] Format ac3 detected only with low score of 25, misdetection possible! [ac3 @ 0x1e31560] Estimating duration from bitrate, this may be inaccurate Input #0, ac3, from 'out.ac3': Duration: 00:00:26.68, start: 0.000000, bitrate: 384 kb/s Stream #0:0: Audio: ac3, 48000 Hz, 5.1(side), fltp, 384 kb/s [New Thread 0x7ffff11fb700 (LWP 30131)] [New Thread 0x7ffff09fa700 (LWP 30132)] [New Thread 0x7ffff01f9700 (LWP 30133)] [New Thread 0x7fffef9f8700 (LWP 30134)] [New Thread 0x7fffef1f7700 (LWP 30135)] [New Thread 0x7fffee9f6700 (LWP 30136)] [New Thread 0x7fffee1f5700 (LWP 30137)] [New Thread 0x7fffed9f4700 (LWP 30138)] [New Thread 0x7fffed1f3700 (LWP 30139)] [webm_chunk @ 0x1e33aa0] Using AVStream.codec to pass codec parameters to muxers is deprecated, use AVStream.codecpar instead. Output #0, webm_chunk, to 'chk/out%d.chk': Metadata: encoder : Lavf57.44.100 Stream #0:0: Audio: vorbis, 48000 Hz, stereo, fltp Metadata: encoder : Lavc57.51.100 vorbis Stream mapping: Stream #0:0 -> #0:0 (ac3 (native) -> vorbis (native)) Press [q] to stop, [?] for help [ac3 @ 0x1e33320] frame sync error Error while decoding stream #0:0: Invalid data found when processing input Program received signal SIGSEGV, Segmentation fault. 0x0000000000600980 in mkv_write_flush_packet (s=0x1eade20, pkt=0x7fffffffd2b0) at libavformat/matroskaenc.c:2111 2111 if (s->pb->seekable) (gdb) bt #0 0x0000000000600980 in mkv_write_flush_packet (s=0x1eade20, pkt=0x7fffffffd2b0) at libavformat/matroskaenc.c:2111 #1 0x00000000006c6cb6 in webm_chunk_write_packet (s=0x1e33aa0, pkt=0x7fffffffd2b0) at libavformat/webm_chunk.c:210 #2 0x000000000063df2d in write_packet (s=s@entry=0x1e33aa0, pkt=pkt@entry=0x7fffffffd2b0) at libavformat/mux.c:732 #3 0x000000000064022e in av_interleaved_write_frame (s=s@entry=0x1e33aa0, pkt=0x0, pkt@entry=0x7fffffffd610) at libavformat/mux.c:1184 #4 0x0000000000493b0d in write_frame (s=s@entry=0x1e33aa0, pkt=pkt@entry=0x7fffffffd610, ost=ost@entry=0x1e68d20) at ffmpeg.c:762 #5 0x0000000000498807 in do_audio_out (frame=0x1e34e60, ost=0x1e68d20, s=0x1e33aa0) at ffmpeg.c:840 #6 reap_filters (flush=flush@entry=0) at ffmpeg.c:1376 #7 0x000000000049be1a in transcode_step () at ffmpeg.c:4119 #8 transcode () at ffmpeg.c:4163 #9 0x000000000047e36b in main (argc=<optimized out>, argv=0x7fffffffdcc8) at ffmpeg.c:4356 (gdb) disass $pc-32,$pc+32 Dump of assembler code from 0x600960 to 0x6009a0: 0x0000000000600960 <mkv_write_flush_packet+16>: and $0xe0,%al 0x0000000000600962 <mkv_write_flush_packet+18>: mov %r13,-0x18(%rsp) 0x0000000000600967 <mkv_write_flush_packet+23>: mov %rsi,%r12 0x000000000060096a <mkv_write_flush_packet+26>: mov %r14,-0x10(%rsp) 0x000000000060096f <mkv_write_flush_packet+31>: mov %r15,-0x8(%rsp) 0x0000000000600974 <mkv_write_flush_packet+36>: sub $0x58,%rsp 0x0000000000600978 <mkv_write_flush_packet+40>: mov 0x18(%rdi),%rbx 0x000000000060097c <mkv_write_flush_packet+44>: mov 0x20(%rdi),%rdi => 0x0000000000600980 <mkv_write_flush_packet+48>: mov 0x90(%rdi),%edx 0x0000000000600986 <mkv_write_flush_packet+54>: mov %rdi,%r13 0x0000000000600989 <mkv_write_flush_packet+57>: test %edx,%edx 0x000000000060098b <mkv_write_flush_packet+59>: jne 0x600991 <mkv_write_flush_packet+65> 0x000000000060098d <mkv_write_flush_packet+61>: mov 0x10(%rbx),%r13 0x0000000000600991 <mkv_write_flush_packet+65>: test %r12,%r12 0x0000000000600994 <mkv_write_flush_packet+68>: je 0x600bb0 <mkv_write_flush_packet+608> 0x000000000060099a <mkv_write_flush_packet+74>: movslq 0x24(%r12),%rax 0x000000000060099f <mkv_write_flush_packet+79>: mov 0x30(%rbp),%rcx End of assembler dump. (gdb) info register rax 0x16a3d00 23739648 rbx 0x1eae440 32171072 rcx 0x5 5 rdx 0x5dc0 24000 rsi 0x7fffffffd2b0 140737488343728 rdi 0x0 0 rbp 0x1eade20 0x1eade20 rsp 0x7fffffffd110 0x7fffffffd110 r8 0x1eaef10 32173840 r9 0x5dc0 24000 r10 0x1 1 r11 0x8000000000000001 -9223372036854775807 r12 0x7fffffffd2b0 140737488343728 r13 0x1eade20 32169504 r14 0x20 32 r15 0x600 1536 rip 0x600980 0x600980 <mkv_write_flush_packet+48> eflags 0x10202 [ IF RF ] cs 0x33 51 ss 0x2b 43 ds 0x0 0 es 0x0 0 fs 0x0 0 gs 0x0 0 }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/5752> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 4