[FFmpeg-devel] Fix ffmpeg -re behaviour
Sun Nov 23 14:58:39 CET 2008
On date Sunday 2008-11-23 13:16:50 +0100, Michael Niedermayer encoded:
> On Sun, Nov 23, 2008 at 11:41:39AM +0100, Stefano Sabatini wrote:
> > On date Saturday 2008-11-22 22:36:29 +0100, Michael Niedermayer encoded:
> > > On Sat, Nov 22, 2008 at 07:17:00PM +0100, Stefano Sabatini wrote:
> > [...]
> > > > Check the patch, I also changed the name pts -> next, I think it's
> > > > less prone to confusion since "pts" may be confused with the meaning
> > > > of pts in ist->next_pts (which uses another time unit).
> > >
> > > so the patch contains 3 unrelated changes now ...
> > Sorry for that.
> > Patches:
> > * fixrate-emu-pts.patch
> > Use ist->next_pts rather than use frame_nb * codec_time_base as the
> > timestamp for the next frame to print.
> > This assumption is not always true and only works for video codecs anyway.
> I dont think next_pts is the correct variable to use, but then what exactly
> where the use cases of rate_emu anyway?
> next_pts is a guessed value that may have no relation to the next pts
-re is used mainly for real-time streaming, for example:
ffmpeg -re -i test.mpg -an -vcodec copy -f rtp rtp://localhost:5004
(currently it works only with video).
Without -re ffmpeg send packets as fast as possible, filling the UDP
buffer on the receiving end, resulting in damaged/corrupted data.
> > * rename-pts.patch
> > Rename the pts variable, which may be confused with the meaning
> > of pts in ist->next_pts (which uses another time unit).
> makes no sense to me
Then I'll drop it.
> > * remove-ist-frame
> > Remove the variable ist->frame, not anymore used.
FFmpeg = Funny and Free Mean Picky Ecletic Guru
More information about the ffmpeg-devel