January 2020 - FFmpeg-trac - lists.ffmpeg.org

#7997(undetermined:new): undefined-behavior at libavcodec/proresenc_anatoliy.c
by FFmpeg 09 Jan '20

09 Jan '20

#7997: undefined-behavior at libavcodec/proresenc_anatoliy.c -------------------------------------+------------------------------------- Reporter: Suhwan | Type: defect Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: ubsan | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: There's am undefined-behavior at libavcodec/proresenc_anatoliy.c:250:12 How to reproduce: {{{ % ffmpeg_g -y -r 8 -i tmp.avi -map 0 -c:v snow -c:v prores_aw -disposition:a:79 prores_aw -disposition:a ffv1 -vframes 42 -aframes 83 -r 9 -level 3 -b:v 507k tmp_.adx ffmpeg version : N-94163-g664a27ea40 built with clang version 9.0.0 }}} negative value "prev_dc" is left shifted {{{ #define TO_GOLOMB(val) (((val) << 1) ^ ((val) >> 31)) 242 static void encode_dc_coeffs(PutBitContext *pb, int16_t *in, 243 int blocks_per_slice, int *qmat) 244 { 245 int prev_dc, code; 246 int i, sign, idx; 247 int new_dc, delta, diff_sign, new_code; 248 249 prev_dc = QSCALE(qmat, 0, in[0] - 16384); 250 code = TO_GOLOMB(prev_dc); 251 encode_codeword(pb, code, FIRST_DC_CB); 252 253 code = 5; sign = 0; idx = 64; 254 for (i = 1; i < blocks_per_slice; i++, idx += 64) { 255 new_dc = QSCALE(qmat, 0, in[idx] - 16384); 256 delta = new_dc - prev_dc; 257 diff_sign = DIFF_SIGN(delta, sign); 258 new_code = TO_GOLOMB2(get_level(delta), diff_sign); 259 260 encode_codeword(pb, new_code, dc_codebook[FFMIN(code, 6)]); 261 262 code = new_code; 263 sign = delta >> 31; 264 prev_dc = new_dc; 265 } 266 } }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/7997> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 8

#7930(undetermined:new): avfoundation input device does not find attached firewire/dv source
by FFmpeg 08 Jan '20

08 Jan '20

#7930: avfoundation input device does not find attached firewire/dv source -----------------------------------+-------------------------------------- Reporter: dericed | Owner: Type: defect | Status: new Priority: normal | Component: undetermined Version: 4.1 | Keywords: avfoundation Blocked By: | Blocking: Reproduced by developer: 0 | Analyzed by developer: 0 -----------------------------------+-------------------------------------- Summary of the bug: I have a DV tape player attached to a Mac via firewire. When I open QuickTime X, I can select the DV stream as an input for QuickTime X (I'm presuming from this behavior that avfoundation is accepting the dv stream as an input); however when I query avfoundation inputs via ffmpeg it only lists screen capture and no dv input available. Is it feasible to access the dv input via the avfoundation input device? How to reproduce: Attach a DV tape player source to a Mac via firewire, verify that the DV is a source input to QuickTime X and other avfoundation-based apps. Quit QuickTime X. Then try to access the DV input via ffmpeg's avfoundation device. {{{ ffmpeg -f avfoundation -list_devices 1 -i dummy ffmpeg version 4.1.3 Copyright (c) 2000-2019 the FFmpeg developers built with Apple LLVM version 10.0.1 (clang-1001.0.46.4) configuration: --prefix=/usr/local/Cellar/ffmpeg/4.1.3_1 --enable-shared --enable-pthreads --enable-version3 --enable-hardcoded-tables --enable- avresample --cc=clang --host- cflags='-I/Library/Java/JavaVirtualMachines/adoptopenjdk-11.0.2.jdk/Contents/Home/include -I/Library/Java/JavaVirtualMachines/adoptopenjdk-11.0.2.jdk/Contents/Home/include/darwin' --host-ldflags= --enable-ffplay --enable-gnutls --enable-gpl --enable- libaom --enable-libbluray --enable-libmp3lame --enable-libopus --enable- librubberband --enable-libsnappy --enable-libtesseract --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-libx265 --enable-libxvid --enable-lzma --enable-libfontconfig --enable-libfreetype --enable-frei0r --enable-libass --enable-libopencore-amrnb --enable- libopencore-amrwb --enable-libopenjpeg --enable-librtmp --enable-libspeex --enable-videotoolbox --disable-libjack --disable-indev=jack --enable- libaom --enable-libsoxr libavutil 56. 22.100 / 56. 22.100 libavcodec 58. 35.100 / 58. 35.100 libavformat 58. 20.100 / 58. 20.100 libavdevice 58. 5.100 / 58. 5.100 libavfilter 7. 40.101 / 7. 40.101 libavresample 4. 0. 0 / 4. 0. 0 libswscale 5. 3.100 / 5. 3.100 libswresample 3. 3.100 / 3. 3.100 libpostproc 55. 3.100 / 55. 3.100 [AVFoundation input device @ 0x7fd212400380] AVFoundation video devices: [AVFoundation input device @ 0x7fd212400380] [0] Capture screen 0 [AVFoundation input device @ 0x7fd212400380] AVFoundation audio devices: dummy: Input/output error }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/7930> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 5

#8314(undetermined:new): memory leaks in wtvfile_open_sector()
by FFmpeg 08 Jan '20

08 Jan '20

#8314: memory leaks in wtvfile_open_sector() -------------------------------------+------------------------------------- Reporter: Suhwan | Type: defect Status: new | Priority: important Component: | Version: git- undetermined | master Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: There're memory leaks detected in wtvfile_open_sector() How to reproduce: {{{ % ffmpeg_g -i $PoC tmp.lxf ffmpeg version N-95464-g7056ddc0e0 Copyright (c) 2000-2019 the FFmpeg developers built with clang version 6.0.0-1ubuntu2 (tags/RELEASE_600/final) configuration: --cc=clang --cxx=clang++ --ld=clang --enable-debug }}} Here's Valgrind log {{{ ==10445== HEAP SUMMARY: ==10445== in use at exit: 5,236 bytes in 38 blocks ==10445== total heap usage: 102 allocs, 64 frees, 79,475 bytes allocated ==10445== ==10445== 4,412 (264 direct, 4,148 indirect) bytes in 1 blocks are definitely lost in loss record 33 of 33 ==10445== at 0x9FE3E76: memalign (in /usr/lib/valgrind /vgpreload_memcheck-amd64-linux.so) ==10445== by 0x9FE3F91: posix_memalign (in /usr/lib/valgrind /vgpreload_memcheck-amd64-linux.so) ==10445== by 0x592EC7D: av_malloc (mem.c:87) ==10445== by 0x14731F6: avio_alloc_context (aviobuf.c:140) ==10445== by 0x1AB7BA6: wtvfile_open_sector (wtvdec.c:238) ==10445== by 0x1AB7BA6: wtvfile_open2 (wtvdec.c:289) ==10445== by 0x1AB3355: read_header (wtvdec.c:989) ==10445== by 0x1A145FC: avformat_open_input (utils.c:633) ==10445== by 0x42FFF7: open_input_file (ffmpeg_opt.c:1105) ==10445== by 0x42DE5E: open_files (ffmpeg_opt.c:3283) ==10445== by 0x42DB4F: ffmpeg_parse_options (ffmpeg_opt.c:3323) ==10445== by 0x487BB3: main (ffmpeg.c:4862) ==10445== ==10445== LEAK SUMMARY: ==10445== definitely lost: 264 bytes in 1 blocks ==10445== indirectly lost: 4,148 bytes in 3 blocks ==10445== possibly lost: 0 bytes in 0 blocks ==10445== still reachable: 824 bytes in 34 blocks ==10445== suppressed: 0 bytes in 0 blocks ==10445== Reachable blocks (those to which a pointer was found) are not shown. ==10445== To see them, rerun with: --leak-check=full --show-leak-kinds=all ==10445== ==10445== For counts of detected and suppressed errors, rerun with: -v ==10445== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 0 from 0) }}} Please confirm. Thanks -- Ticket URL: <https://trac.ffmpeg.org/ticket/8314> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#8072(ffplay:new): ffplay does not scale video to window size when exiting fullscreen
by FFmpeg 08 Jan '20

08 Jan '20

#8072: ffplay does not scale video to window size when exiting fullscreen -----------------------------------+--------------------------------------- Reporter: jimbo1qaz | Type: defect Status: new | Priority: normal Component: ffplay | Version: unspecified Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -----------------------------------+--------------------------------------- Summary of the bug: How to reproduce: {{{ % ffplay file.mp4 ffplay version N-94500-g2828f5b0d8 Copyright (c) 2003-2019 the FFmpeg developers built with gcc 9.1.1 (GCC) 20190807 }}} * While video is playing, rress F to enter fullscreen. The video will resize to fit the entire screen. * Press F to exit fullscreen. The video will not resize to fit the window, and only the top-left portion is visible until the window is resized. This bug happens on both NUT holding raw video frames from stdin, and x264 MP4 files. OS: Windows 10 x64. I tested historical builds from https://ffmpeg.zeranoe.com/builds/win64/shared/ : Unaffected: ffmpeg-4.1. Affected: ffmpeg-4.1.1, 20190101, and 20190811. -- Ticket URL: <https://trac.ffmpeg.org/ticket/8072> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 13

#7673(undetermined:new): Fails to play long (30+ hours) hls playlist
by FFmpeg 07 Jan '20

07 Jan '20

#7673: Fails to play long (30+ hours) hls playlist -------------------------------------+------------------------------------- Reporter: Jasdf923 | Type: defect Status: new | Priority: normal Component: | Version: git- undetermined | master Keywords: hls m3u | Blocked By: m3u8 playlist | Reproduced by developer: 0 Blocking: | Analyzed by developer: 0 | -------------------------------------+------------------------------------- Summary of the bug: ffmpeg (ffplay) fails to play a long hls playlist. Trying to seek to somewhere after roughly after 27 hours does not work. How to reproduce: First, generate a long playlist. I am using the following command. It creates a lot of files totalling about 750MiB, so run it in a new directory. {{{ ffmpeg -t 165600 -f lavfi -i color=c=black:s=2x2 -c:v libx264 -tune stillimage -pix_fmt yuv420p -hls_time 10 -hls_playlist_type vod playlist.m3u8 }}} This is 46 hours of blank video split into 10 second segments. {{{ ffmpeg -v 9 -loglevel 99 -i playlist.m3u8 ffmpeg version n4.1 Copyright (c) 2000-2018 the FFmpeg developers built with gcc 8.2.1 (GCC) 20180831 configuration: --prefix=/usr --disable-debug --disable-static --disable- stripping --enable-fontconfig --enable-gmp --enable-gnutls --enable-gpl --enable-ladspa --enable-libaom --enable-libass --enable-libbluray --enable-libdrm --enable-libfreetype --enable-libfribidi --enable-libgsm --enable-libiec61883 --enable-libjack --enable-libmodplug --enable- libmp3lame --enable-libopencore_amrnb --enable-libopencore_amrwb --enable- libopenjpeg --enable-libopus --enable-libpulse --enable-libsoxr --enable- libspeex --enable-libssh --enable-libtheora --enable-libv4l2 --enable- libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable- libx264 --enable-libx265 --enable-libxcb --enable-libxml2 --enable-libxvid --enable-nvdec --enable-nvenc --enable-omx --enable-shared --enable- version3 libavutil 56. 22.100 / 56. 22.100 libavcodec 58. 35.100 / 58. 35.100 libavformat 58. 20.100 / 58. 20.100 libavdevice 58. 5.100 / 58. 5.100 libavfilter 7. 40.101 / 7. 40.101 libswscale 5. 3.100 / 5. 3.100 libswresample 3. 3.100 / 3. 3.100 libpostproc 55. 3.100 / 55. 3.100 Splitting the commandline. Reading option '-v' ... matched as option 'v' (set logging level) with argument '9'. Reading option '-loglevel' ... matched as option 'loglevel' (set logging level) with argument '99'. Reading option '-i' ... matched as input url with argument 'playlist.m3u8'. Finished splitting the commandline. Parsing a group of options: global . Applying option v (set logging level) with argument 9. Successfully parsed a group of options. Parsing a group of options: input url playlist.m3u8. Successfully parsed a group of options. Opening an input file: playlist.m3u8. [NULL @ 0x5610dd18edc0] Opening 'playlist.m3u8' for reading [file @ 0x5610dd18f8c0] Setting default whitelist 'file,crypto' Probing hls,applehttp score:100 size:2048 [hls,applehttp @ 0x5610dd18edc0] Format hls,applehttp probed with size=2048 and score=100 [hls,applehttp @ 0x5610dd18edc0] new_program: id=0x0000 [hls,applehttp @ 0x5610dd18edc0] HLS request for url 'playlist0.ts', offset 0, playlist 0 [hls,applehttp @ 0x5610dd18edc0] Opening 'playlist0.ts' for reading Probing mpeg score:25 size:2048 Probing mpegts score:50 size:2048 Format mpegts probed with size=2048 and score=50 score: 44, dvhs_score: -2, fec_score: -2 [mpegts @ 0x5610dd4a6880] Filter: pid=0x11 type=1 [mpegts @ 0x5610dd4a6880] Filter: pid=0x0 type=1 [mpegts @ 0x5610dd4a6880] SDT: [mpegts @ 0x5610dd4a6880] tag: 0x48 len=18 [mpegts @ 0x5610dd4a6880] new_program: id=0x0001 [mpegts @ 0x5610dd4a6880] PAT: [mpegts @ 0x5610dd4a6880] sid=0x1 pid=0x1000 [mpegts @ 0x5610dd4a6880] new_program: id=0x0001 [mpegts @ 0x5610dd4a6880] Filter: pid=0x1000 type=1 [mpegts @ 0x5610dd4a6880] PMT: len 21 [mpegts @ 0x5610dd4a6880] sid=0x1 sec_num=0/0 version=0 tid=2 [mpegts @ 0x5610dd4a6880] pcr_pid=0x100 [mpegts @ 0x5610dd4a6880] Filter: pid=0x100 type=0 [mpegts @ 0x5610dd4a6880] stream=0 stream_type=1b pid=100 prog_reg_desc= [mpegts @ 0x5610dd4a6880] tuning done [hls,applehttp @ 0x5610dd18edc0] Before avformat_find_stream_info() pos: 585164 bytes read:585164 seeks:0 nb_streams:1 [mpegts @ 0x5610dd4a6880] Skipping after seek [mpegts @ 0x5610dd4a6880] SDT: [mpegts @ 0x5610dd4a6880] tag: 0x48 len=18 [mpegts @ 0x5610dd4a6880] new_program: id=0x0001 [mpegts @ 0x5610dd4a6880] PAT: [mpegts @ 0x5610dd4a6880] sid=0x1 pid=0x1000 [mpegts @ 0x5610dd4a6880] new_program: id=0x0001 [mpegts @ 0x5610dd4a6880] PMT: len 21 [mpegts @ 0x5610dd4a6880] sid=0x1 sec_num=0/0 version=0 tid=2 [mpegts @ 0x5610dd4a6880] pcr_pid=0x100 [mpegts @ 0x5610dd4a6880] pid=100 pes_code=0x1e0 Last message repeated 2 times [AVBSFContext @ 0x5610dd4bf340] nal_unit_type: 9(AUD), nal_ref_idc: 0 [AVBSFContext @ 0x5610dd4bf340] nal_unit_type: 7(SPS), nal_ref_idc: 3 [AVBSFContext @ 0x5610dd4bf340] nal_unit_type: 8(PPS), nal_ref_idc: 3 [AVBSFContext @ 0x5610dd4bf340] nal_unit_type: 6(SEI), nal_ref_idc: 0 [AVBSFContext @ 0x5610dd4bf340] nal_unit_type: 5(IDR), nal_ref_idc: 3 [h264 @ 0x5610dd4bd300] nal_unit_type: 9(AUD), nal_ref_idc: 0 [h264 @ 0x5610dd4bd300] nal_unit_type: 7(SPS), nal_ref_idc: 3 [h264 @ 0x5610dd4bd300] nal_unit_type: 8(PPS), nal_ref_idc: 3 [h264 @ 0x5610dd4bd300] nal_unit_type: 6(SEI), nal_ref_idc: 0 [h264 @ 0x5610dd4bd300] nal_unit_type: 5(IDR), nal_ref_idc: 3 [h264 @ 0x5610dd4bd300] Format yuv420p chosen by get_format(). [h264 @ 0x5610dd4bd300] Reinit context to 16x16, pix_fmt: yuv420p [h264 @ 0x5610dd4bd300] no picture [mpegts @ 0x5610dd4a6880] pid=100 pes_code=0x1e0 Last message repeated 126 times [hls,applehttp @ 0x5610dd18edc0] max_analyze_duration 5000000 reached at 5000000 microseconds st:0 [hls,applehttp @ 0x5610dd18edc0] Setting avg frame rate based on r frame rate [hls,applehttp @ 0x5610dd18edc0] stream 0: start_time: 1.480 duration: -102481911520608.625 [hls,applehttp @ 0x5610dd18edc0] format: start_time: 1.480 duration: 165600.000 bitrate=0 kb/s [hls,applehttp @ 0x5610dd18edc0] After avformat_find_stream_info() pos: 585164 bytes read:585164 seeks:0 frames:127 Input #0, hls,applehttp, from 'playlist.m3u8': Duration: 46:00:00.00, start: 1.480000, bitrate: 0 kb/s Program 0 Metadata: variant_bitrate : 0 Stream #0:0, 127, 1/90000: Video: h264 (High), 1 reference frame ([27][0][0][0] / 0x001B), yuv420p(left), 2x2 (16x16) [SAR 1:1 DAR 1:1], 0/1, 25 fps, 25 tbr, 90k tbn, 50 tbc Metadata: variant_bitrate : 0 Successfully opened the file. At least one output file must be specified [AVIOContext @ 0x5610dd4b8040] Statistics: 32768 bytes read, 0 seeks [AVIOContext @ 0x5610dd197d40] Statistics: 585164 bytes read, 0 seeks }}} Note that the log above was created with release version 4.1 but I have reproduced the same problem using the development version. The error can be reproduced with: {{{ ffplay playlist.m3u8 }}} Then seek to somewhere later than 30 hours. Earlier than 25 hours works fine but I am do not know exactly at which time between 25 and 30 hours the problem starts to occur. I use the right mouse button to seek in ffplay. Here is an example log output {{{ Seek to 31% (14:22:30) of total duration (46:00:00) B f=0/0 [hls,applehttp @ 0x7fc81c000b80] Opening 'playlist5175.ts' for reading [hls,applehttp @ 0x7fc81c000b80] Opening 'playlist5176.ts' for reading Seek to 75% (34:34:18) of total duration (46:00:00) 0B f=0/0 [hls,applehttp @ 0x7fc81c000b80] Opening 'playlist12445.ts' for reading [hls,applehttp @ 0x7fc81c000b80] Opening 'playlist12446.ts' for reading [hls,applehttp @ 0x7fc81c000b80] Opening 'playlist12447.ts' for reading ... }}} The first seek works fine and the file plays correctly, after the second seek the console is filled with the above messages, showing all the individual segments in quick succession. The file no longer plays at all. I can also reproduce this with the mpv media player which uses ffmpeg internally. This makes seeking a bit easier and it shows the same log messages. {{{ mpv --no-config --msg-level=all=debug playlist.m3u8 }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/7673> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 3

#8462(ffmpeg:new): some sort of bash completion for ffmpeg
by FFmpeg 07 Jan '20

07 Jan '20

#8462: some sort of bash completion for ffmpeg ---------------------------------+--------------------------------------- Reporter: shirish | Type: enhancement Status: new | Priority: normal Component: ffmpeg | Version: 4.2 Keywords: | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | ---------------------------------+--------------------------------------- Dear Maintainers, It would be nice if ffmpeg has some sort of command-completion so something like - ffmpeg -err_detect ignore_err -i video.mkv -c copy video_fixed.mkv I am thinking of something like bash-completion https://github.com/scop /bash-completion but in ffmpeg . It would be ok and probably nicer if it was a separate ffmpeg-completion package so overtime people could add the commands and swtiches that could be autocompleted after writing the first few letters. -- Ticket URL: <https://trac.ffmpeg.org/ticket/8462> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 2

#8456(ffmpeg:new): Copy codec doubles video stream
by FFmpeg 05 Jan '20

05 Jan '20

#8456: Copy codec doubles video stream -------------------------------------+------------------------------------- Reporter: MarianD | Type: defect Status: new | Priority: normal Component: ffmpeg | Version: git- Keywords: copy muxer | master demuxer | Blocked By: Blocking: | Reproduced by developer: 0 Analyzed by developer: 0 | -------------------------------------+------------------------------------- Copy codec doubles video stream (audio stream is OK): How to reproduce: {{{ % ffmpeg -i Bubble_burst.mp4 -c copy Bubble.mp4 ffmpeg version: git-2020-01-02-81172b5 built on Windows 7 64-bit (Zeranoe's build) }}} FFprobe for input and output files: {{{ >ffprobe -hide_banner Bubble_burst.mp4 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'Bubble_burst.mp4': Metadata: major_brand : mp42 minor_version : 0 compatible_brands: mp42isom encoder : mp4creator 1.6.1d Duration: 00:00:06.33, start: 0.000000, bitrate: 556 kb/s Stream #0:0(eng): Video: h264 (Constrained Baseline) (avc1 / 0x31637661), yuv420p, 376x360 [SAR 1:1 DAR 47:45], 554 kb/s, 15 fps, 15 tbr, 90k tbn, 30 tbc (default) Metadata: creation_time : 2020-01-04T18:54:36.000000Z encoder : VC Coding Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 24000 Hz, stereo, fltp, 2 kb/s (default) Metadata: creation_time : 2020-01-04T18:54:36.000000Z >ffprobe -hide_banner Bubble.mp4 Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'Bubble.mp4': Metadata: major_brand : isom minor_version : 512 compatible_brands: isomiso2avc1mp41 encoder : Lavf58.35.101 Duration: 00:00:12.67, start: 0.000000, bitrate: 555 kb/s Stream #0:0(eng): Video: h264 (Constrained Baseline) (avc1 / 0x31637661), yuv420p, 376x360 [SAR 1:1 DAR 47:45], 554 kb/s, 15 fps, 15 tbr, 90k tbn, 30 tbc (default) Metadata: handler_name : VideoHandler Stream #0:1(eng): Audio: aac (LC) (mp4a / 0x6134706D), 24000 Hz, stereo, fltp, 2 kb/s (default) Metadata: handler_name : SoundHandler }}} -- Ticket URL: <https://trac.ffmpeg.org/ticket/8456> FFmpeg <https://ffmpeg.org> FFmpeg issue tracker

1 9