Commit 1181461c authored by Stefano Sabatini's avatar Stefano Sabatini

lavfi/asetpts,setpts: add variables T, STARTT, PREV_INT and PREV_OUTT

parent 3d6d59ef
......@@ -4129,9 +4129,15 @@ audio sample rate
@item STARTPTS
the PTS of the first frame
@item STARTT
the time in seconds of the first frame
@item INTERLACED
tell if the current frame is interlaced
@item T
the time in seconds of the current frame
@item TB
the time base
......@@ -4142,9 +4148,14 @@ for the current frame
@item PREV_INPTS
previous input PTS
@item PREV_INT
previous input time in seconds
@item PREV_OUTPTS
previous output PTS
@item PREV_OUTT
previous output time in seconds
@end table
@subsection Examples
......
......@@ -41,10 +41,14 @@ static const char *const var_names[] = {
"NB_SAMPLES", ///< number of samples in the current frame (only audio)
"POS", ///< original position in the file of the frame
"PREV_INPTS", ///< previous input PTS
"PREV_INT", ///< previous input time in seconds
"PREV_OUTPTS", ///< previous output PTS
"PREV_OUTT", ///< previous output time in seconds
"PTS", ///< original pts in the file of the frame
"SAMPLE_RATE", ///< sample rate (only audio)
"STARTPTS", ///< PTS at start of movie
"STARTT", ///< time at start of movie
"T", ///< original time in the file of the frame
"TB", ///< timebase
NULL
};
......@@ -56,10 +60,14 @@ enum var_name {
VAR_NB_SAMPLES,
VAR_POS,
VAR_PREV_INPTS,
VAR_PREV_INT,
VAR_PREV_OUTPTS,
VAR_PREV_OUTT,
VAR_PTS,
VAR_SAMPLE_RATE,
VAR_STARTPTS,
VAR_STARTT,
VAR_T,
VAR_TB,
VAR_VARS_NB
};
......@@ -82,9 +90,9 @@ static av_cold int init(AVFilterContext *ctx, const char *args)
}
setpts->var_values[VAR_N ] = 0.0;
setpts->var_values[VAR_PREV_INPTS ] = NAN;
setpts->var_values[VAR_PREV_OUTPTS] = NAN;
setpts->var_values[VAR_STARTPTS ] = NAN;
setpts->var_values[VAR_PREV_INPTS ] = setpts->var_values[VAR_PREV_INT ] = NAN;
setpts->var_values[VAR_PREV_OUTPTS] = setpts->var_values[VAR_PREV_OUTT] = NAN;
setpts->var_values[VAR_STARTPTS ] = setpts->var_values[VAR_STARTT ] = NAN;
return 0;
}
......@@ -106,6 +114,7 @@ static int config_input(AVFilterLink *inlink)
#define D2TS(d) (isnan(d) ? AV_NOPTS_VALUE : (int64_t)(d))
#define TS2D(ts) ((ts) == AV_NOPTS_VALUE ? NAN : (double)(ts))
#define TS2T(ts, tb) ((ts) == AV_NOPTS_VALUE ? NAN : (double)(ts)*av_q2d(tb))
static int filter_frame(AVFilterLink *inlink, AVFilterBufferRef *inpicref)
{
......@@ -116,9 +125,12 @@ static int filter_frame(AVFilterLink *inlink, AVFilterBufferRef *inpicref)
if (!outpicref)
return AVERROR(ENOMEM);
if (isnan(setpts->var_values[VAR_STARTPTS]))
if (isnan(setpts->var_values[VAR_STARTPTS])) {
setpts->var_values[VAR_STARTPTS] = TS2D(inpicref->pts);
setpts->var_values[VAR_STARTT ] = TS2T(inpicref->pts, inlink->time_base);
}
setpts->var_values[VAR_PTS ] = TS2D(inpicref->pts);
setpts->var_values[VAR_T ] = TS2T(inpicref->pts, inlink->time_base);
setpts->var_values[VAR_POS ] = inpicref->pos == -1 ? NAN : inpicref->pos;
switch (inlink->type) {
......@@ -150,7 +162,9 @@ static int filter_frame(AVFilterLink *inlink, AVFilterBufferRef *inpicref)
setpts->var_values[VAR_N] += 1.0;
setpts->var_values[VAR_PREV_INPTS ] = TS2D(inpicref ->pts);
setpts->var_values[VAR_PREV_INT ] = TS2T(inpicref ->pts, inlink->time_base);
setpts->var_values[VAR_PREV_OUTPTS] = TS2D(outpicref->pts);
setpts->var_values[VAR_PREV_OUTT] = TS2T(outpicref->pts, inlink->time_base);
if (setpts->type == AVMEDIA_TYPE_AUDIO) {
setpts->var_values[VAR_NB_CONSUMED_SAMPLES] += inpicref->audio->nb_samples;
......
......@@ -30,7 +30,7 @@
#define LIBAVFILTER_VERSION_MAJOR 3
#define LIBAVFILTER_VERSION_MINOR 13
#define LIBAVFILTER_VERSION_MICRO 100
#define LIBAVFILTER_VERSION_MICRO 101
#define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
LIBAVFILTER_VERSION_MINOR, \
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment