[Bug 793959] qtmux: round to nearest when computing mehd.fragment_duration

[GStreamer (GNOME Bugzilla)]

https://bugzilla.gnome.org/show_bug.cgi?id=793959

--- Comment #3 from Alicia Boya García <aboya at igalia.com> ---
(In reply to Sebastian Dröge (slomo) from comment #2)
> Comment on attachment 369152 [details] [review]
> qtmux: round to nearest when computing mehd.fragment_duration
> 
> Makes sense but I'm worried about this breaking some gst-validate tests. Did
> you check? :)

Yes.

> Also the same problem should exist at the places where the duration in the
> tkhd and mdhd are calculated. They're basically the same calculations. We
> should keep them consistent. Can you check?

mdhd needs no rounding (it defines its own timescale).

tkhd is computed from mdhd, but rounding was also not used (incorrect). I just
updated the patch to fix that too (and ran the tests again).

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.