Hi all,

Thanks to Giles' feedback, the issue should be fixed now!

Romain

2013/6/20 Gilou <[email protected]>:
> Le 19/06/2013 19:34, Romain Beauxis a écrit :
>> 2013/6/19 Gilou <[email protected]>:
>>> Le 19/06/2013 14:06, Gilou a écrit :
>>>> Le 19/06/2013 04:44, David Baelde a écrit :
>>>>> Hi Gilou,
>>>>>
>>>>> Thanks for investigating more. I wonder what it could be. If you're
>>>>> using a regular playlist (not playlist.safe) then the files are not
>>>>> checked in advance, so really the only thing that changes with the
>>>>> size of the playlist is parsing time. What you said is consistent with
>>>>> this: you have the same problem with a smaller playlist. After that
>>>>> point, one possible source of slowdown is if a file takes forever to
>>>>> check before loading into the queue. Perhaps can you identify
>>>>> problematic files and see how long "liquidsoap -r <file>" takes on it.
>>>>>
>>>>> Good luck,
>>>
>>> OK.. Found it again.. It's a bug in faad, that can't read some m4a
>>> files. Not sure why. If I do faad -i $file, it makes the CPU cry. The
>>> XML is not the problem, as some normally-read files have the same kind
>>> of metadata.
>>>
>>> I need to investigate this further, or ban .m4a (maybe not a bad idea,
>>> but kinda annoying)..
>>> Lead: http://bugs.musicpd.org/print_bug_page.php?bug_id=2414
>>>
>>
>> Thanks for looking at that, Gilou! I'll do my best to look at it ASAP.
>>
>> Romain
>
> So !
>
> The bug seems to indeed be in libmp4ff, that can't read some of the
> files.. As I told Romain, if I try to use faad -i, it uses 100% CPU, and
> the backtrace gives:
>
> #12 0x00000000004056dd in parse_sub_atoms (f=0x60e220, total_size=37434,
>     meta_only=0) at mp4ff.c:173
> #13 0x000000000040568a in parse_sub_atoms (f=0x60e220, total_size=125657,
>     meta_only=0) at mp4ff.c:171
> #14 0x0000000000405799 in parse_atoms (f=0x60e220, meta_only=0) at
> mp4ff.c:214
> #15 0x00000000004058ba in mp4ff_open_read (f=0x60d090) at mp4ff.c:43
> #16 0x00000000004026d7 in decodeMP4file (song_length=<synthetic pointer>,
>     adts_out=0, infoOnly=1, noGapless=0, downMatrix=0, fileType=1,
>     outputFormat=1, to_stdout=0, adts_fn=0x0,
>     sndfile=0x60d050 "01 Tenderly.wav",
>     mp4file=0x60d010 "01 Tenderly.m4a")
>     at main.c:856
> #17 main (argc=<optimized out>, argv=<optimized out>) at main.c:1364
>
> I'm not exactly an expert on libmp4ff, nor on m4a format.. So I thought
> to myself, "hey, maybe ffmpeg can help, and you could use add_decoder
> with avconv". No luck, but another lead, maybe..
>
> % ffprobe /tmp/01\ Tenderly.m4a
> avprobe version 0.8.6-6:0.8.6-1ubuntu2, Copyright (c) 2007-2013 the
> Libav developers
>   built on Mar 30 2013 22:20:06 with gcc 4.7.2
> [mov,mp4,m4a,3gp,3g2,mj2 @ 0x1af4b20] error reading header: -1094995529
> /tmp/01 Tenderly.m4a: Invalid data found when processing input
>
> Well.. That's something I can parse (though ffprobe returns 0 after
> this.. yey..), but I'd rather just have liq play those files..
>
> vlc reads them, and libmediainfo is able to get the metadatas out..
> without using 100% cpu.. so.. dunno.. Help ? :D
>
> mencoder and mplayer chokes on it as well, but it uses libav.. and spits
> this:
>
> Playing /tmp/01 Tenderly.m4a.
> libavformat version 53.21.1 (external)
> Mismatching header version 53.19.0
> libavformat file format detected.
> [mov,mp4,m4a,3gp,3g2,mj2 @ 0x7faa7b6758c0]error reading header: -1094995529
> LAVF_header: av_open_input_stream() failed
> ISO: File Type Major Brand: Apple iTunes AAC-LC Audio
> Quicktime/MOV file format detected.
> [mov] Audio stream found, -aid 0
> Load subtitles in /tmp/
> ==========================================================================
> Opening audio decoder: [ffmpeg] FFmpeg/libavcodec audio decoders
> libavcodec version 53.35.0 (external)
> Mismatching header version 53.32.2
> AUDIO: 44100 Hz, 2 ch, s16le, 1411.2 kbit/100.00% (ratio: 176400->176400)
> Selected audio codec: [ffaac] afm: ffmpeg (FFmpeg AAC (MPEG-2/MPEG-4 Audio))
> ==========================================================================
> AO: [pulse] 44100Hz 2ch s16le (2 bytes per sample)
> Video: no video
> Starting playback...
>
> So it works in the end, but it cries again about a header... Maybe it
> expects video ?
>
> Cheers
> Gilou
>
>
> ------------------------------------------------------------------------------
> This SF.net email is sponsored by Windows:
>
> Build for Windows Store.
>
> http://p.sf.net/sfu/windows-dev2dev
> _______________________________________________
> Savonet-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/savonet-users

------------------------------------------------------------------------------
This SF.net email is sponsored by Windows:

Build for Windows Store.

http://p.sf.net/sfu/windows-dev2dev
_______________________________________________
Savonet-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/savonet-users

Reply via email to