Baptiste Coudurier <[email protected]> added the comment:
On 07/10/2009 02:55 PM, Jason Garrett-Glaser wrote: > > New submission from Jason Garrett-Glaser<[email protected]>: > > Another Facebook-reported bug; delete the original file once the problem is > solved. > > "facebook_svq3_watermark_failure.mov" in incoming plays fine on my version of > mplayer (using lavc for decoding) but fails on ffmpeg as follows (perhaps a > recent regression)? Mplayer, by the way, doesn't display the watermark; > Quicktime does. There was an issue on 64bit. Displaying the watermark would require to blend the picture, so this has to be done in libavfilter, I guess patch welcome. > ffmpeg -y -i facebook_svq3_watermark_failure.mov -f yuv4mpegpipe -an /dev/null > FFmpeg version SVN-r19398, Copyright (c) 2000-2009 Fabrice Bellard, et al. > configuration: --disable-devices > libavutil 50. 3. 0 / 50. 3. 0 > libavcodec 52.32. 0 / 52.32. 0 > libavformat 52.36. 0 / 52.36. 0 > libavdevice 52. 2. 0 / 52. 2. 0 > libswscale 0. 7. 1 / 0. 7. 1 > built on Jul 10 2009 14:29:53, gcc: 4.0.1 20050727 (Red Hat 4.0.1-5) > [mov,mp4,m4a,3gp,3g2,mj2 @ 0xf42370]multiple edit list entries, a/v desync > might > occur, patch welcome > [svq3 @ 0xf4b720]could not uncompress watermark logo > > Seems stream 0 codec frame rate differs from container frame rate: 29970.00 > (29970/1) -> 29.97 (30000/1001) > Input #0, mov,mp4,m4a,3gp,3g2,mj2, from 'facebook_svq3_watermark_failure.mov': > Duration: 00:01:20.09, start: 0.000000, bitrate: 1074 kb/s > Stream #0.0(eng): Video: svq3, yuvj420p, 320x240, 29.97 tbr, 29970 tbn, > 29970 tbc > Stream #0.1(eng): Audio: mp3, 44100 Hz, 2 channels, s16 > Stream #0.2(eng): Data: rtp / 0x20707472 > Stream #0.3(eng): Data: rtp / 0x20707472 > [svq3 @ 0xf4b720]could not uncompress watermark logo > Output #0, yuv4mpegpipe, to '/dev/null': > Stream #0.0(eng): Video: rawvideo, yuvj420p, 320x240, q=2-31, 200 kb/s, > 90k > tbn, 29.97 tbc > Stream mapping: > Stream #0.0 -> #0.0 > Error while opening decoder for input stream #0.0 > > This also shows another potential problem; a watermark decoding failure > shouldn't be a critical failure IMO unless there's a good reason. There is a _very_ good reason the watermark is xored with the bistream. no watermark -> no bitstream. ________________________________________________ FFmpeg issue tracker <[email protected]> <https://roundup.ffmpeg.org/issue1263> ________________________________________________
