[gstreamer-bugs] [Bug 151661] New: - Possible bug in mad plugin

[bugzilla-daemon at bugzilla.gnome.org]

http://bugzilla.gnome.org/show_bug.cgi?id=151661
GStreamer | gst-plugins | Ver: 0.8.2

           Summary: Possible bug in mad plugin
           Product: GStreamer
           Version: 0.8.2
          Platform: Other
        OS/Version: Linux
            Status: UNCONFIRMED
          Severity: normal
          Priority: Normal
         Component: gst-plugins
        AssignedTo: gstreamer-bugs at lists.sourceforge.net
        ReportedBy: francis_labonte at hotmail.com
         QAContact: gstreamer-bugs at lists.sourceforge.net


Hi,

I've got a problem when trying to decode a mp3 file (filesrc location=a.mp3 ! 
mad ! ...).  I've founded that the caps was'nt set correctly.  I think this 
problem come from the method use to recover from a "MAD_ERROR_LOST_SYNC".

Actually, when this happen and the data is a id3tag, then "mad_stream_skip" is 
used to move at the and of the current tag, and it is followed by a call 
to "mad_stream_sync".

However, I've founded that "mad_stream_sync" does'nt move the actual data 
pointer, but only set an internal offset/skip value.   

So, "mad_stram_sync" move the data pointer to the next frame and the next time 
you call mad_decode, the offset will be applied.  Of course, the position will 
now be wrong and will cause the decoder to return a sync error once again.

I think we should not call "mad_stram_sync" when "mad_stream_skip" has just 
been called.  OR, "mad_stream_skip" should consider the skip/offset if there 
is one.

------- You are receiving this mail because: -------
You are the assignee for the bug.
You are the QA contact for the bug.