RepoMirrors
/
mpv
mirror of https://github.com/mpv-player/mpv
1
0
Fork 0
Code Issues Releases Wiki Activity

vo: add new frame drop detection 

For the vo-delayed-frame-count property.

Slightly less dumb than the previous one (which was removed earlier),
but still pretty dumb. But this also seems to be relatively robust, even
with strong vsync jittering.
This commit is contained in:
wm4 2015-11-25 22:11:30 +01:00
parent b8bcf0f466
commit 44376d2d9b
1 changed files with 15 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
video/out/vo.c
View File

@ -143,6 +143,7 @@ struct vo_internal {
int64_t *vsync_samples;
int num_vsync_samples;
int64_t prev_vsync;
int64_t base_vsync;
double estimated_vsync_interval;
double estimated_vsync_jitter;
bool expecting_vsync;
@ -326,6 +327,7 @@ static void reset_vsync_timings(struct vo *vo)
in->prev_vsync = 0;
in->estimated_vsync_interval = 0;
in->estimated_vsync_jitter = -1;
in->base_vsync = 0;
in->expecting_vsync = false;
}
@ -344,6 +346,8 @@ static void update_vsync_timing_after_swap(struct vo *vo)
in->num_vsync_samples -= 1;
MP_TARRAY_INSERT_AT(in, in->vsync_samples, in->num_vsync_samples, 0,
now - in->prev_vsync);
if (!in->base_vsync)
in->base_vsync = now;
in->prev_vsync = now;
double avg = 0, jitter = 0;
@ -357,6 +361,17 @@ static void update_vsync_timing_after_swap(struct vo *vo)
in->estimated_vsync_jitter = sqrt(jitter / in->num_vsync_samples);
MP_STATS(vo, "value %f jitter", in->estimated_vsync_jitter);
if (llabs(in->base_vsync - now) > in->vsync_interval * 2 / 3) {
// Assume a drop. An underflow can technically speaking not be a drop
// (it's up to the driver what this is supposed to mean), but no reason
// to treat it differently.
in->base_vsync = now;
in->delayed_count += 1;
MP_STATS(vo, "vo-delayed");
}
in->base_vsync += in->vsync_interval;
}
// to be called from VO thread only