On Wed, Dec 15, 2021 at 08:36:25PM +0100, Anton Khirnov wrote:
> Quoting Michael Niedermayer (2021-12-14 22:13:43)
> > On Mon, Dec 13, 2021 at 04:20:41PM +0100, Anton Khirnov wrote:
> > > It features in limiting the number of output frames (-frames option) and
> > > currently can be incremented from two places:
> > > - for video encoding (not streamcopy), in do_video_out() after each
> > >   successful avcodec_send_frame() call
> > > - for all other cases, in of_submit_packet()
> > > 
> > > Not only is this inconsistent and confusing, but also the special
> > > treatment for video encoding is redundant, since
> > > AVCodecContext.frame_count stores the exact same value (number of
> > > successful avcodec_send_frame()) calls.
> > > 
> > > Replace the use of OutputStream.frame_count in do_video_out() with
> > > AVCodecContext.frame_count. Modify OutputStream.frame_count in the same
> > > way for all streams.
> > > ---
> > >  fftools/ffmpeg.c     |  9 ++++-----
> > >  fftools/ffmpeg_mux.c | 19 +++++++------------
> > >  2 files changed, 11 insertions(+), 17 deletions(-)
> > 
> > This results in differences
> > one is:
> > ./ffmpeg -i ~/tickets/4072/CoP\ 1\ 1\ 3\ -M.avi -vframes 2 -bitexact -vcodec huffyuv /tmp/4072-fic-old.avi
> > 
> > -rw-r----- 1 michael michael 3572792 Dez 14 22:09 /tmp/4072-fic-new.avi
> > -rw-r----- 1 michael michael 1908172 Dez 14 22:10 /tmp/4072-fic-old.avi
> > 
> > I see other differences caused by patches today i still need to investigate
> > what causes what
> 
> Fixing this will require larger restructuring first, so I am dropping
> this patch for now.
> 
> Please let me know if the rest of the set breaks anything else.

all failures from this patch set where from this patch i think

thx

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

I know you won't believe me, but the highest form of Human Excellence is
to question oneself and others. -- Socrates