[Ffmpeg-devel] [bug?] evo sample

[Kostya]

On Fri, Feb 16, 2007 at 12:51:44AM +0100, Nico Sabbi wrote:
> Kostya wrote:
> >On Thu, Feb 15, 2007 at 03:48:05AM +0000, Compn wrote:
> >
> >>ffvc1_decode_broken_very_badly.evo in /incoming
> >>
> >>produce some messages in mplayer svn like:
> >>[vc1 @ 00B18864]Bits overconsumption: 540819 > 540728
> >>and displays a bad frame...
> >>
> >>can someone check if its a bug? sample is 5mb 
> >>
> >>-compn
> >
> > 
> >Looks like it's MPlayer demuxer tricks.
> 
> syncing packets to sequence/frames boundaries?
> 
> >This sample was cut from the middle of the stream and
> >contains some frames before sequence header (I had to
> >chop ~200k from the beginning to handle it in FFmpeg).
> >
> >And then MPlayer and FFplay handle this file slightly
> >differently - and the sequences are different too!
> >
> 
> sorry to disappoint you ;) , but I have the same error with mplayer 
> -demuxer lavf:
> [ASPECT] Warning: No suitable new res found!
> [vc1 @ 0xb74b7ac8]Bits overconsumption: 320092 > 320016 at 72x41
> V:   6.5 159/159 141%  6%  0.0% 0 0

Is it _really_ the same?
It does not play in FFplay and mplayer -demuxer lavf
If you chop 204800 bytes from the beginning (so it may be handled by
lavc demuxer) it gives such results to me (MacOSX/PPC)
     MPlayer:
[vc1 @ 0x5abfb0]Bits overconsumption: 540903 > 540728
[vc1 @ 0x5abfb0]Bits overconsumption: 184212 > 183392 at 84x25

     MPlayer -demuxer lavf:
[vc1 @ 0x5abfb0]Bits overconsumption: 194040 > 193824 at 71x26

Note that the last error is the usual thing when the frame is truncated.

> -- 
> "Without a frontend, mplayer is useless" - someone in mplayer-users
>