[MPlayer-dev-eng] rgselk's .rm patch

[Michael Niedermayer]

Hi

On Friday 05 December 2003 21:15, rgselk wrote:
> --- Nico <nsabbi at tiscali.it> wrote:
> > rgselk wrote:
> > > I guess your .rm file is the result of mplayer
> > >-dumpstream? At least that is the reason why the first patch put
> > >keyframe markers in the dumped stream...
> >
> > yes, it's a dump.
> > I also used -idx and -forceidx, but I couldn't seek anyway.
> > How can I do?
>
> Once you have dumped it without keyframe tags it is too late for me to
> find the keyframes :-(
> The first part of the patch finds keyframes in the rtsp stream and sets
> the right tag (0x02).
>
> Try to dump the stream again with my patch applied and try -idx on the
> resulting dump. Without tags I cannot build an index and you cannot
> seek...
>
> If someone knows how to reliably detect a new keyframe without the
> keyframe tag being available, I would like to know.
> And trying to seek to a non-key frame results in 'funny' noise...
i didnt follow that thread and dunno what exactly is dumped and what not ...
but maybe rv20_decode_picture_header() in libavcodec/rv10.c could help in 
figureing out if a video packet belongs to a keyfame (pict_type=I_TYPE) the 
code can currently parse rv10-rv30 headers, for rv40 look at pict_type in 
demux_real.c

[...]
-- 
Michael
level[i]= get_vlc(); i+=get_vlc();		(violates patent EP0266049)
median(mv[y-1][x], mv[y][x-1], mv[y+1][x+1]);	(violates patent #5,905,535)
buf[i]= qp - buf[i-1];				(violates patent #?)
for more examples, see http://mplayerhq.hu/~michael/patent.html
stop it, see http://petition.eurolinux.org & http://petition.ffii.org/eubsa/en