[FFmpeg-devel] [PATCH 5/5] lavfi/select: add support for concatdec_select option

[Nicolas George]

Le decadi 20 brumaire, an CCXXIV, Marton Balint a écrit :
> This option can be used to select useful frames from an ffconcat file which is
> using inpoints and outpoints but where the source files are not intra frame
> only.
> 
> Signed-off-by: Marton Balint <cus at passwd.hu>
> ---
>  doc/filters.texi       | 17 +++++++++++++++++
>  libavfilter/f_select.c | 27 +++++++++++++++++++++++++++
>  2 files changed, 44 insertions(+)
> 
> diff --git a/doc/filters.texi b/doc/filters.texi
> index 471ec3f..ade571d 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -13196,6 +13196,16 @@ value between 0 and 1 to indicate a new scene; a low value reflects a low
>  probability for the current frame to introduce a new scene, while a higher
>  value means the current frame is more likely to be one (see the example below)
>  
> + at item concatdec_select
> +The concat demuxer can set the @var{lavf.concat.start_time} and the
> + at var{lavf.concat.duration} packet metadata values which are also present in the
> +decoded frames.
> +
> +The @var{concatdec_select} variable is -1 if the frame pts is at least
> +start_time and either the duration metadata is missing or the frame pts is less
> +than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
> +missing.
> +
>  @end table
>  
>  The default value of the select expression is "1".
> @@ -13270,6 +13280,13 @@ Send even and odd frames to separate outputs, and compose them:
>  @example
>  select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
>  @end example
> +
> + at item
> +Select useful frames from an ffconcat file which is using inpoints and
> +outpoints but where the source files are not intra frame only.
> + at example
> +ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
> + at end example
>  @end itemize
>  
>  @section selectivecolor
> diff --git a/libavfilter/f_select.c b/libavfilter/f_select.c
> index 2b926e1..14a23d2 100644
> --- a/libavfilter/f_select.c
> +++ b/libavfilter/f_select.c
> @@ -82,6 +82,8 @@ static const char *const var_names[] = {
>  
>      "scene",
>  
> +    "concatdec_select",  ///< frame usefulness based on pts and frame metadata originating from the concat demuxer
> +
>      NULL
>  };
>  
> @@ -132,6 +134,8 @@ enum var_name {
>  
>      VAR_SCENE,
>  
> +    VAR_CONCATDEC_SELECT,
> +
>      VAR_VARS_NB
>  };
>  
> @@ -278,6 +282,28 @@ static double get_scene_score(AVFilterContext *ctx, AVFrame *frame)
>      return ret;
>  }
>  
> +static double get_concatdec_select(AVFrame *frame, int64_t pts)
> +{
> +    AVDictionary *metadata = av_frame_get_metadata(frame);
> +    AVDictionaryEntry *e1 = av_dict_get(metadata, "lavf.concatdec.start_time", NULL, 0);
> +    AVDictionaryEntry *e2 = av_dict_get(metadata, "lavf.concatdec.duration", NULL, 0);

> +    if (e1 && e1->value) {

Is the "e1->value" test necessary? 

> +        int64_t start_time = strtoll(e1->value, NULL, 10);
> +        if (pts >= start_time) {
> +            if (e2 && e2->value) {
> +              int64_t duration = strtoll(e2->value, NULL, 10);
> +              if (pts < start_time + duration)
> +                  return -1;
> +              else
> +                  return 0;
> +            }

> +            return -1;

Did you mean +1 here?

> +        }
> +        return 0;
> +    }
> +    return NAN;
> +}
> +
>  #define D2TS(d)  (isnan(d) ? AV_NOPTS_VALUE : (int64_t)(d))
>  #define TS2D(ts) ((ts) == AV_NOPTS_VALUE ? NAN : (double)(ts))
>  
> @@ -297,6 +323,7 @@ static void select_frame(AVFilterContext *ctx, AVFrame *frame)
>      select->var_values[VAR_T  ] = TS2D(frame->pts) * av_q2d(inlink->time_base);
>      select->var_values[VAR_POS] = av_frame_get_pkt_pos(frame) == -1 ? NAN : av_frame_get_pkt_pos(frame);
>      select->var_values[VAR_KEY] = frame->key_frame;
> +    select->var_values[VAR_CONCATDEC_SELECT] = get_concatdec_select(frame, av_rescale_q(frame->pts, inlink->time_base, AV_TIME_BASE_Q));
>  
>      switch (inlink->type) {
>      case AVMEDIA_TYPE_AUDIO:

LGTM apart from that, but I do not maintain this filter.

Regards,

-- 
  Nicolas George
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://ffmpeg.org/pipermail/ffmpeg-devel/attachments/20151113/0e53bf9d/attachment.sig>