[FFmpeg-devel] [PATCH 3/5] Use int for num_faulty_pts/dts instead of int64_t
Mon Sep 13 05:05:19 CEST 2010
On Tue, Jul 27, 2010 at 9:04 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
> On Mon, Jul 26, 2010 at 01:16:08PM -0700, Alexander Strange wrote:
>> ?cmdutils.h | ? ?4 ++--
>> ?ffplay.c ? | ? ?2 +-
>> ?2 files changed, 3 insertions(+), 3 deletions(-)
> this change is wrong and unneeded, int64_t is needed to ensure they dont
> overflow. Consider audio with 1 sample per packet, it would overflow
> within less than a day. the code must be able to run for years without
> failing and it would fail for much more realistic audio packet sizes in
> that period.
Hmm, I haven't really thought about audio timestamps, they seem very
easy to "repair" by just ignoring them.
More information about the ffmpeg-devel