[MPlayer-users] mencoder bug reporting question

[Corey Hickey]

No, I didn't use b-frames. To be sure, I just tried one of my known-bad 
cases with -lavcopts vmax_b_frames=0 (which should default to 0 anyway), 
and got the same results as earlier.

I guess I'll go ahead and describe the problem.

When the video should show a very sudden white flash, decoding with 
labavcodec displays the flash as it should. When using DivX5 for 
decoding, though, the effect is similar to that of a bad frame - the 
flash shows but an instant later the picture gets blocky and garbled, 
until the next keyframe.

Apparently, lavc is encoding something that DivX5 can't properly read, 
but I don't know which codec is to blame.

Encoding with these commands works fine:

mencoder -dvd 1 -lavcopts vcodec=mpeg4:vhq -frames 2500

mencoder -dvd 1 -lavcopts vcodec=mpeg4 -frames 2500 -ss 38

...while these do not:

mencoder -dvd 1 -lavcopts vcodec=mpeg4 -frames 2500 -ss 40

mencoder -dvd 1 -lavcopts vcodec=mpeg4:vhq -frames 2500

mencoder -dvd 1 -lavcopts vcodec=mpeg4 -vop crop=720:370:0:52\
  -frames 2500

mencoder -dvd 1 -lavcopts vcodec=mpeg4 -vop scale=720:312\
  -frames 2500

If anyone is curious at this point, I'll upload a few files:
clh-stream.dump was created with
$ mplayer -dvd 1 -sb 60000000 -dumpvideo
and clh-bug.avi was then encodec by
$ mencoder stream.dump -lavcopts vcodec=mpeg4:vhq

Someone tell me whether a full bugreport is necessary. :)

-Corey

Michael Niedermayer wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> Hi
> 
> On Monday 20 May 2002 10:31, Corey Hickey wrote:
> 
>>[Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
>>I am preparing a complete bugreport regarding a visual artifact glitch
>>that is reproducible (under certain conditions) when encoding with
>>libavcodec and then playing back with the DivX 5 library (under both
>>Linux and Windows).
> 
> ill bet its another? divx5 bug, did u use b-frames? divx5 is buggy with IBI 
> type sequences ...
> 
> 
>>For this report, what mencoder verbosity level should I use (how many
>>-v's)? Bugreports.html doesn't say, and I don't know what level would be
>>useful for debugging what seems to be an encoding problem. I plan to
> 
> hmm dunno (i doubt the -v log does matter in that case ...)
> [...]
> 
> Michael
> 
> _______________________________________________
> RTFM!!!  http://www.MPlayerHQ.hu/DOCS
> Search:  http://www.MPlayerHQ.hu/cgi-bin/htsearch
> http://mplayerhq.hu/mailman/listinfo/mplayer-users
>