[Ffmpeg-devel] [PATCH] DVR-MS probe audio format - was DVR-MS bug (MP2 decoding)

John Donaghy johnfdonaghy
Mon Apr 9 03:17:24 CEST 2007

> > +    int stream_needs_parsing[MAX_STREAMS]={0};
> what is this good for?

This is there to remember if the audio streams with an id of CODEC_ID_NONE
have their 'need_parsing' flag set - but I suppose that they always do (now
that you mention it) - I'll remove it.

> >
> >      for(i=0;i<MAX_STREAMS;i++){
> >          last_dts[i]= AV_NOPTS_VALUE;
> > +        probe_data[i].filename = "";
> > +        probe_data[i].buf = NULL;
> > +        probe_data[i].buf_size = 0;
> >      }
> memset(0) is simpler

will do.

> code duplication, also the max score should win not some >1 or other
> random
> test

Both ac3_probe and mp3_read_probe often return 1. What should I do then?  -
in all the samples tested the audio is always ac3 when both probe functions
return 1.

actually, you should rather use read_probe() instead of this

Do you mean read_probe() as it is used in av_probe_input_format2 - ie loop
through the formats looking for ac3 and mp3 and then call read_probe thus
avoiding the need to make the probe functions non-static?

Also, just to clarify - I'm passing the concatenated raw data found in the
audio packets to the probe functions. Is that definitely the right thing to
do? I'm not certain that the data doesnt need to be inserted into some kind
of container to make it suitable for consumption by the probe functions. I'd
just like clarification that the way I'm doing it is right.



> --
> Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
> Observe your enemies, for they first find out your faults. -- Antisthenes
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at mplayerhq.hu
> http://lists.mplayerhq.hu/mailman/listinfo/ffmpeg-devel

More information about the ffmpeg-devel mailing list