RepoMirrors
/
ffmpeg
mirror of https://git.ffmpeg.org/ffmpeg.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

lavfi/vf_libplacebo: fix crop expr PTS calculation 

mix->timestamps is expressed relative to the source timebase, which is
possibly a different timescale from `base_pts`. We can't mix-and-match
here. The only reason this worked in my previous testing was because I
was testing on a source file which had an exactly matching timebase.

Fix it by always using the exact PTS as tagged on the AVFrame.
This commit is contained in:
Niklas Haas 2023-05-24 14:35:49 +02:00
parent d815584755
commit 0bce5590c3
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
libavfilter/vf_libplacebo.c
View File

@ -651,7 +651,7 @@ fail:
static void update_crops(AVFilterContext *ctx,
struct pl_frame_mix *mix, struct pl_frame *target,
uint64_t ref_sig, double base_pts)
uint64_t ref_sig, double target_pts)
{
LibplaceboContext *s = ctx->priv;
@ -659,11 +659,12 @@ static void update_crops(AVFilterContext *ctx,
// Mutate the `pl_frame.crop` fields in-place. This is fine because we
// own the entire pl_queue, and hence, the pointed-at frames.
struct pl_frame *image = (struct pl_frame *) mix->frames[i];
double image_pts = base_pts + mix->timestamps[i];
const AVFrame *src = pl_get_mapped_avframe(image);
double image_pts = src->pts * av_q2d(ctx->inputs[0]->time_base);
/* Update dynamic variables */
s->var_values[VAR_IN_T] = s->var_values[VAR_T] = image_pts;
s->var_values[VAR_OUT_T] = s->var_values[VAR_OT] = base_pts;
s->var_values[VAR_OUT_T] = s->var_values[VAR_OT] = target_pts;
s->var_values[VAR_N] = ctx->outputs[0]->frame_count_out;
/* Clear these explicitly to avoid leaking previous frames' state */