2016-02-23 19:41 GMT+01:00 Stefano Sabatini <stefa...@gmail.com>:
> On date Friday 2016-02-19 14:04:12 +0100, Martin Vignali encoded:
> > 2016-02-15 2:35 GMT+01:00 Stefano Sabatini <stefa...@gmail.com>:
> [...]
> > > > + err = av_timecode_init_from_string(&tc, rate,
> > > tcr->value, s);
> > > > + if (err < 0) {
> > > > + av_log(s, AV_LOG_WARNING, "Could not
> increment
> > > timecode, error occured during timecode creation.");
> > > > + break;
> > > > + }
> > >
> > > > + tc.start += (int)
> > > av_q2d(av_mul_q(av_make_q(seg->time, 1000000 ), rate));/* increment
> > > timecode */
> > >
> > > nit: 1000000_) => 1000000)
> > >
> > > Also, reading from the timecode.c code it looks like the increment
> > > must be expressed in fps units.
> > >
> >
> > Sorry don't understand what's wrong now. In my tests, the timecode
> > incrementation is correct (but maybe i missed a special case)
>
> Not necessarily a problem, assuming the fps is consistent with the
> rate.
>
Ok, so i think there is no problem, because timecode is created with the
rate (av_timecode_init_from_string set the fps from rate.),
and the increment is made with the rate.
Both values are consistent.
>
> > > Finally, you are using the segment->time, which is different from the
> > > effective segment duration. Could you set the metadata in segment_end,
> > > where the exact duration is known?
> > >
> >
> > Do you know how can i get the exact segment duration ?
>
> What about setting the metadata when you're *closing* the segment
> (e.g. in segment_end())?
>
> This way you can know the exact duration.
> --
>
>
Ok for that, but in fact my question is more about the way to have a more
precise duration than segment->time.
Thanks,
Martin
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
