On Tue, Oct 01, 2019 at 09:34:48PM +0200, Marton Balint wrote:
>
>
> On Mon, 30 Sep 2019, lance.lmw...@gmail.com wrote:
>
> >From: Limin Wang <lance.lmw...@gmail.com>
> >
> >How to tested it, please try with the following command:
> >./ffmpeg -f lavfi -i
> >"smptebars=duration=5:size=1280x720:rate=30,freezedetect=d=1:f=0" -f null -
> >frame= 150 fps=0.0 q=-0.0 Lsize=N/A time=00:00:05.00 bitrate=N/A speed= 232x
> >
> >./ffmpeg -f lavfi -i
> >"smptebars=duration=5:size=1280x720:rate=30,freezedetect=d=1:f=-1" -f null -
> >frame= 120 fps=0.0 q=-0.0 Lsize=N/A time=00:00:04.00 bitrate=N/A speed= 211x
> >
> >./ffmpeg -f lavfi -i
> >"smptebars=duration=5:size=1280x720:rate=30,freezedetect=d=1:f=1" -f null -
> >frame= 30 fps=0.0 q=-0.0 Lsize=N/A time=00:00:01.00 bitrate=N/A speed=93.9x
> >
> >Signed-off-by: Limin Wang <lance.lmw...@gmail.com>
> >---
> >doc/filters.texi | 10 ++++++++++
> >libavfilter/vf_freezedetect.c | 23 ++++++++++++++++++++++-
> >2 files changed, 32 insertions(+), 1 deletion(-)
> >
> >diff --git a/doc/filters.texi b/doc/filters.texi
> >index 6ed1c8fa75..2be8b93c53 100644
> >--- a/doc/filters.texi
> >+++ b/doc/filters.texi
> >@@ -10678,6 +10678,8 @@ timestamp of the first frame of the freeze. The
> >@code{lavfi.freezedetect.freeze_duration} and
> >@code{lavfi.freezedetect.freeze_end} metadata keys are set on the first frame
> >after the freeze.
> >+In addition, you can choose to discard the freeze/non-freezee frames
> >instead of
> >+report by metadata.
> >
> >The filter accepts the following options:
> >
> >@@ -10689,6 +10691,14 @@ specified value) or as a difference ratio between 0
> >and 1. Default is -60dB, or
> >
> >@item duration, d
> >Set freeze duration until notification (default is 2 seconds).
> >+
> >+@item force_discard, f
> >+Set force to discard or keep freeze frame.
> >+ 0: do nothing
> >+-1: discard non-freeze frame
> >+ 1: discard freeze frame
>
> I don't see how discarding non-frozen frames is useful in any way,
> so that option should be removed unless you have something sensible
> in mind.
When the input file is big, I like to extract the freeze frame
after the running sometimes, it's useful to help check whether
it's expect for the detecting result.
>
> I would also like to point out that freezedetect does not buffer any
> frames, so the first few frames (the detection interval) of a freeze
> won't be dropped, which is not what the user might expect when it
> wants to drop frozen frames. At least you should document this.
OK, will update the document to note it.
>
> >+
> >+Default is 0
> >@end table
> >
> >@anchor{frei0r}
> >diff --git a/libavfilter/vf_freezedetect.c b/libavfilter/vf_freezedetect.c
> >index cc086afee6..fc08c235f7 100644
> >--- a/libavfilter/vf_freezedetect.c
> >+++ b/libavfilter/vf_freezedetect.c
> >@@ -45,6 +45,9 @@ typedef struct FreezeDetectContext {
> >
> > double noise;
> > int64_t duration; ///< minimum duration of frozen frame until
> > notification
> >+
> >+ int force_discard; ///< 0: no discard, -1: discard non-freeze
> >frame, 1: discard freeze frame
> >+ int drop_count;
>
> Why don't you simply use s->frozen to see if the current frame needs
> to be dropped or not? That would be much easier to follow.
OK, I'll consider it how to simplify it by your advise.
>
> >} FreezeDetectContext;
> >
> >#define OFFSET(x) offsetof(FreezeDetectContext, x)
> >@@ -56,6 +59,8 @@ static const AVOption freezedetect_options[] = {
> > { "noise", "set noise tolerance",
> > OFFSET(noise), AV_OPT_TYPE_DOUBLE, {.dbl=0.001}, 0, 1.0, V|F },
> > { "d", "set minimum duration in seconds",
> > OFFSET(duration), AV_OPT_TYPE_DURATION, {.i64=2000000}, 0, INT64_MAX,
> > V|F },
> > { "duration", "set minimum duration in seconds",
> > OFFSET(duration), AV_OPT_TYPE_DURATION, {.i64=2000000}, 0, INT64_MAX,
> > V|F },
> >+ { "f", "set frame discard",
> >OFFSET(force_discard), AV_OPT_TYPE_INT, {.i64=0}, -1, 1, V|F
> >},
> >+ { "force_discard", "set frame discard",
> >OFFSET(force_discard), AV_OPT_TYPE_INT, {.i64=0}, -1, 1, V|F
> >},
>
> A name like "discard" should be enough, no need for short version,
> it is only used for consistency with silencedetect.
OK, will update it.
>
> >
> > {NULL}
> >};
> >@@ -115,6 +120,7 @@ static int config_input(AVFilterLink *inlink)
> > if (!s->sad)
> > return AVERROR(EINVAL);
> >
> >+ s->drop_count = 0;
> > return 0;
> >}
> >
> >@@ -184,10 +190,22 @@ static int activate(AVFilterContext *ctx)
> > set_meta(s, frame, "lavfi.freezedetect.freeze_end",
> > av_ts2timestr(frame->pts, &inlink->time_base));
> > }
> > s->frozen = frozen;
> >+ if ( s->force_discard > 0 && frozen)
> >+ s->drop_count++;
> >+ else if ( s->force_discard < 0 && frozen && s->drop_count <
> >0) {
> >+ s->drop_count = 0;
> >+ }
> >+ } else {
> >+ if ( s->force_discard < 0)
> >+ s->drop_count--;
> > }
> > }
> >
> > if (!frozen) {
> >+ if (s->force_discard > 0) {
> >+ s->drop_count = 0;
> >+ } else if ( s->force_discard < 0)
> >+ s->drop_count--;
> > av_frame_free(&s->reference_frame);
> > s->reference_frame = av_frame_clone(frame);
> > s->reference_n = s->n;
> >@@ -196,7 +214,10 @@ static int activate(AVFilterContext *ctx)
> > return AVERROR(ENOMEM);
> > }
> > }
> >- return ff_filter_frame(outlink, frame);
> >+ if (s->drop_count > 0 || s->drop_count < 0) {
> >+ av_frame_free(&frame);
> >+ } else
> >+ return ff_filter_frame(outlink, frame);
> > }
> >
> > FF_FILTER_FORWARD_STATUS(inlink, outlink);
> >--
> >2.21.0
> >
>
> Regards,
> Marton
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
