"-t" and "-filter_complex" are incompatible

[nichot20]

Summary of the bug:

If using a commandline containing "-filter_complex" on a large file, attempting to use -t to limit the length of the output file (say in order to test the command line) results in the required number of frames being processed, and then the buffers continuing to fill up until either all frames have been read, or memory runs out.

In the first case the resultant file is OK (in my tests), but time is wasted as the process reads to the end of the file, in the second case the process is terminated without the file being written fully, making it unusable (in my tests).

How to reproduce:

% ffmpeg -i LTA01631701.mxf -filter_complex "
> [0:1] [0:2] amerge [audio]" \
> -map 0:0 -map '[audio]' \
> -c:v dvvideo -pix_fmt yuv420p \
> -c:a pcm_s16le -ar 48k -ac 2 \
> -t 10 -y ./LTA01631701-duff.mov
ffmpeg version N-41976-g734aeeb-by_Tim Copyright (c) 2000-2012 the FFmpeg developers
  built on Jun 27 2012 08:52:54 with gcc 4.6.2
  configuration: --extra-version=by_Tim --enable-static --disable-shared --enable-gpl --enable-nonfree --enable-version3 --prefix=/mnt/msds-store-0/tim/ffmpeg-tux/usr/local --libdir=/mnt/msds-store-0/tim/ffmpeg-tux/usr/local/lib64 --samples=../fate-suite/ --enable-runtime-cpudetect --extra-cflags='-static -I/mnt/msds-store-0/tim/ffmpeg-tux/usr/local/include' --extra-ldflags='-static -L/mnt/msds-store-0/tim/ffmpeg-tux/usr/local/lib64' --progs-suffix=_Jun-27 --enable-libfaac --enable-libx264 --enable-libfreetype --disable-ffplay --disable-ffserver
  libavutil      51. 63.100 / 51. 63.100
  libavcodec     54. 29.101 / 54. 29.101
  libavformat    54. 11.100 / 54. 11.100
  libavdevice    54.  0.100 / 54.  0.100
  libavfilter     3.  0.100 /  3.  0.100
  libswscale      2.  1.100 /  2.  1.100
  libswresample   0. 15.100 /  0. 15.100
  libpostproc    52.  0.100 / 52.  0.100
[mxf @ 0x180c960] could not resolve material track sequence strong ref
Guessed Channel Layout for  Input Stream #0.1 : mono               
Guessed Channel Layout for  Input Stream #0.2 : mono 
Guessed Channel Layout for  Input Stream #0.3 : mono 
Guessed Channel Layout for  Input Stream #0.4 : mono 
Input #0, mxf, from 'LTA01631701.mxf': 
  Metadata:
    timecode        : 00:00:00:00
  Duration: 00:17:44.44, start: 0.000000, bitrate: 170532 kb/s
    Stream #0:0: Video: rawvideo (UYVY / 0x59565955), uyvy422, 720x576, 25 fps, 25 tbr, 25 tbn, 25 tbc
    Stream #0:1: Audio: pcm_s24le, 48000 Hz, mono, s32, 1152 kb/s
    Stream #0:2: Audio: pcm_s24le, 48000 Hz, mono, s32, 1152 kb/s
    Stream #0:3: Audio: pcm_s24le, 48000 Hz, mono, s32, 1152 kb/s
    Stream #0:4: Audio: pcm_s24le, 48000 Hz, mono, s32, 1152 kb/s
[graph 0 input from stream 0:1 @ 0x183cac0] tb:1/25 samplefmt:s32 samplerate:48000 chlayout:0x4
[graph 0 input from stream 0:2 @ 0x183cb20] tb:1/25 samplefmt:s32 samplerate:48000 chlayout:0x4
[Parsed_amerge_0 @ 0x1812cc0] Inputs overlap: output layout will be meaningless
[Parsed_amerge_0 @ 0x1812cc0] in0:mono + in1:mono -> out:stereo                 
[auto-inserted resampler 0 @ 0x182cf60] chl:stereo fmt:s32 r:48000Hz -> chl:stereo fmt:s16 r:48000Hz
[graph 1 input from stream 0:0 @ 0x182c420] w:720 h:576 pixfmt:uyvy422 tb:1/25 fr:25/1 sar:0/1 sws_param:flags=2
[output stream 0:0 @ 0x182cec0] No opaque field provided
[auto-inserted scaler 0 @ 0x182c540] w:720 h:576 fmt:uyvy422 sar:0/1 -> w:720 h:576 fmt:yuv420p sar:0/1 flags:0x4 
Output #0, mov, to './LTA01631701-duff.mov':
  Metadata:
    timecode        : 00:00:00:00
    encoder         : Lavf54.11.100
    Stream #0:0: Video: dvvideo (dvcp / 0x70637664), yuv420p, 720x576, q=2-31, 200 kb/s, 25 tbn, 25 tbc
    Stream #0:1: Audio: pcm_s16le (sowt / 0x74776F73), 48000 Hz, stereo, s16, 1536 kb/s
Stream mapping:
  Stream #0:1 (pcm_s24le) -> amerge: (graph 0)
  Stream #0:2 (pcm_s24le) -> amerge::sample_fmt=s32:channel_layout=0x4 (graph 0)
  Stream #0:0 -> #0:0 (rawvideo -> dvvideo)
  amerge (graph 0) -> Stream #0:1 (pcm_s16le)
Press [q] to stop, [?] for help
frame=  250 fps= 28 q=0.0 Lsize=   37035kB time=00:00:10.00 bitrate=30339.1kbits/s
8424 Killed                  ffmpeg -i.......

The input file need to be larger than available memory to see the process being killed.

[nichot20]

8069db86 by Nicolas now adds an AV_LOG_WARNING about this issue.

[cehoyos]

This was fixed before 1.0 and works fine with current git head.