1
0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2024-12-12 19:18:44 +02:00
FFmpeg/libavfilter/af_silencedetect.c

270 lines
12 KiB
C
Raw Normal View History

2012-01-04 01:47:09 +03:00
/*
* Copyright (c) 2012 Clément Bœsch <u pkh me>
2012-01-04 01:47:09 +03:00
*
* This file is part of FFmpeg.
*
* FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
/**
* @file
* Audio silence detector
*/
#include <float.h> /* DBL_MAX */
#include "libavutil/mem.h"
2012-01-04 01:47:09 +03:00
#include "libavutil/opt.h"
#include "libavutil/timestamp.h"
#include "audio.h"
2012-01-04 01:47:09 +03:00
#include "avfilter.h"
#include "internal.h"
2012-01-04 01:47:09 +03:00
typedef struct SilenceDetectContext {
2012-01-04 01:47:09 +03:00
const AVClass *class;
double noise; ///< noise amplitude ratio
int64_t duration; ///< minimum duration of silence until notification
int mono; ///< mono mode : check each channel separately (default = check when ALL channels are silent)
int channels; ///< number of channels
int independent_channels; ///< number of entries in following arrays (always 1 in mono mode)
int64_t *nb_null_samples; ///< (array) current number of continuous zero samples
int64_t *start; ///< (array) if silence is detected, this value contains the time of the first zero sample (default/unset = INT64_MIN)
int64_t frame_end; ///< pts of the end of the current frame (used to compute duration of silence at EOS)
2012-01-04 01:47:09 +03:00
int last_sample_rate; ///< last sample rate to check for sample rate changes
AVRational time_base; ///< time_base
void (*silencedetect)(struct SilenceDetectContext *s, AVFrame *insamples,
int nb_samples, int64_t nb_samples_notify,
AVRational time_base);
2012-01-04 01:47:09 +03:00
} SilenceDetectContext;
#define MAX_DURATION (24*3600*1000000LL)
2012-01-04 01:47:09 +03:00
#define OFFSET(x) offsetof(SilenceDetectContext, x)
#define FLAGS AV_OPT_FLAG_FILTERING_PARAM|AV_OPT_FLAG_AUDIO_PARAM
2012-01-04 01:47:09 +03:00
static const AVOption silencedetect_options[] = {
{ "n", "set noise tolerance", OFFSET(noise), AV_OPT_TYPE_DOUBLE, {.dbl=0.001}, 0, DBL_MAX, FLAGS },
{ "noise", "set noise tolerance", OFFSET(noise), AV_OPT_TYPE_DOUBLE, {.dbl=0.001}, 0, DBL_MAX, FLAGS },
{ "d", "set minimum duration in seconds", OFFSET(duration), AV_OPT_TYPE_DURATION, {.i64=2000000}, 0, MAX_DURATION,FLAGS },
{ "duration", "set minimum duration in seconds", OFFSET(duration), AV_OPT_TYPE_DURATION, {.i64=2000000}, 0, MAX_DURATION,FLAGS },
{ "mono", "check each channel separately", OFFSET(mono), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
{ "m", "check each channel separately", OFFSET(mono), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, FLAGS },
{ NULL }
2012-01-04 01:47:09 +03:00
};
AVFILTER_DEFINE_CLASS(silencedetect);
2012-01-04 01:47:09 +03:00
static void set_meta(AVFrame *insamples, int channel, const char *key, char *value)
{
char key2[128];
if (channel)
snprintf(key2, sizeof(key2), "lavfi.%s.%d", key, channel);
else
snprintf(key2, sizeof(key2), "lavfi.%s", key);
av_dict_set(&insamples->metadata, key2, value, 0);
}
static av_always_inline void update(SilenceDetectContext *s, AVFrame *insamples,
int is_silence, int current_sample, int64_t nb_samples_notify,
AVRational time_base)
{
int channel = current_sample % s->independent_channels;
if (is_silence) {
if (s->start[channel] == INT64_MIN) {
s->nb_null_samples[channel]++;
if (s->nb_null_samples[channel] >= nb_samples_notify) {
s->start[channel] = insamples->pts + av_rescale_q(current_sample / s->channels + 1 - nb_samples_notify * s->independent_channels / s->channels,
(AVRational){ 1, s->last_sample_rate }, time_base);
set_meta(insamples, s->mono ? channel + 1 : 0, "silence_start",
av_ts2timestr(s->start[channel], &time_base));
if (s->mono)
av_log(s, AV_LOG_INFO, "channel: %d | ", channel);
av_log(s, AV_LOG_INFO, "silence_start: %s\n",
av_ts2timestr(s->start[channel], &time_base));
}
}
} else {
if (s->start[channel] > INT64_MIN) {
int64_t end_pts = insamples ? insamples->pts + av_rescale_q(current_sample / s->channels,
(AVRational){ 1, s->last_sample_rate }, time_base)
: s->frame_end;
int64_t duration_ts = end_pts - s->start[channel];
if (insamples) {
set_meta(insamples, s->mono ? channel + 1 : 0, "silence_end",
av_ts2timestr(end_pts, &time_base));
set_meta(insamples, s->mono ? channel + 1 : 0, "silence_duration",
av_ts2timestr(duration_ts, &time_base));
}
if (s->mono)
av_log(s, AV_LOG_INFO, "channel: %d | ", channel);
av_log(s, AV_LOG_INFO, "silence_end: %s | silence_duration: %s\n",
av_ts2timestr(end_pts, &time_base),
av_ts2timestr(duration_ts, &time_base));
}
s->nb_null_samples[channel] = 0;
s->start[channel] = INT64_MIN;
}
}
#define SILENCE_DETECT(name, type) \
static void silencedetect_##name(SilenceDetectContext *s, AVFrame *insamples, \
int nb_samples, int64_t nb_samples_notify, \
AVRational time_base) \
{ \
const type *p = (const type *)insamples->data[0]; \
const type noise = s->noise; \
int i; \
\
for (i = 0; i < nb_samples; i++, p++) \
update(s, insamples, *p < noise && *p > -noise, i, \
nb_samples_notify, time_base); \
}
#define SILENCE_DETECT_PLANAR(name, type) \
static void silencedetect_##name(SilenceDetectContext *s, AVFrame *insamples, \
int nb_samples, int64_t nb_samples_notify, \
AVRational time_base) \
{ \
const int channels = insamples->ch_layout.nb_channels; \
const type noise = s->noise; \
\
nb_samples /= channels; \
for (int i = 0; i < nb_samples; i++) { \
for (int ch = 0; ch < insamples->ch_layout.nb_channels; ch++) { \
const type *p = (const type *)insamples->extended_data[ch]; \
update(s, insamples, p[i] < noise && p[i] > -noise, \
channels * i + ch, \
nb_samples_notify, time_base); \
} \
} \
}
SILENCE_DETECT(dbl, double)
SILENCE_DETECT(flt, float)
SILENCE_DETECT(s32, int32_t)
SILENCE_DETECT(s16, int16_t)
SILENCE_DETECT_PLANAR(dblp, double)
SILENCE_DETECT_PLANAR(fltp, float)
SILENCE_DETECT_PLANAR(s32p, int32_t)
SILENCE_DETECT_PLANAR(s16p, int16_t)
static int config_input(AVFilterLink *inlink)
{
AVFilterContext *ctx = inlink->dst;
SilenceDetectContext *s = ctx->priv;
int c;
s->channels = inlink->ch_layout.nb_channels;
s->duration = av_rescale(s->duration, inlink->sample_rate, AV_TIME_BASE);
s->independent_channels = s->mono ? s->channels : 1;
s->nb_null_samples = av_calloc(s->independent_channels,
sizeof(*s->nb_null_samples));
if (!s->nb_null_samples)
return AVERROR(ENOMEM);
s->start = av_malloc_array(sizeof(*s->start), s->independent_channels);
if (!s->start)
return AVERROR(ENOMEM);
for (c = 0; c < s->independent_channels; c++)
s->start[c] = INT64_MIN;
switch (inlink->format) {
case AV_SAMPLE_FMT_DBL: s->silencedetect = silencedetect_dbl; break;
case AV_SAMPLE_FMT_FLT: s->silencedetect = silencedetect_flt; break;
case AV_SAMPLE_FMT_S32:
s->noise *= INT32_MAX;
s->silencedetect = silencedetect_s32;
break;
case AV_SAMPLE_FMT_S16:
s->noise *= INT16_MAX;
s->silencedetect = silencedetect_s16;
break;
case AV_SAMPLE_FMT_DBLP: s->silencedetect = silencedetect_dblp; break;
case AV_SAMPLE_FMT_FLTP: s->silencedetect = silencedetect_fltp; break;
case AV_SAMPLE_FMT_S32P:
s->noise *= INT32_MAX;
s->silencedetect = silencedetect_s32p;
break;
case AV_SAMPLE_FMT_S16P:
s->noise *= INT16_MAX;
s->silencedetect = silencedetect_s16p;
break;
default:
return AVERROR_BUG;
}
return 0;
}
Merge commit '7e350379f87e7f74420b4813170fe808e2313911' * commit '7e350379f87e7f74420b4813170fe808e2313911': lavfi: switch to AVFrame. Conflicts: doc/filters.texi libavfilter/af_ashowinfo.c libavfilter/audio.c libavfilter/avfilter.c libavfilter/avfilter.h libavfilter/buffersink.c libavfilter/buffersrc.c libavfilter/buffersrc.h libavfilter/f_select.c libavfilter/f_setpts.c libavfilter/fifo.c libavfilter/split.c libavfilter/src_movie.c libavfilter/version.h libavfilter/vf_aspect.c libavfilter/vf_bbox.c libavfilter/vf_blackframe.c libavfilter/vf_delogo.c libavfilter/vf_drawbox.c libavfilter/vf_drawtext.c libavfilter/vf_fade.c libavfilter/vf_fieldorder.c libavfilter/vf_fps.c libavfilter/vf_frei0r.c libavfilter/vf_gradfun.c libavfilter/vf_hqdn3d.c libavfilter/vf_lut.c libavfilter/vf_overlay.c libavfilter/vf_pad.c libavfilter/vf_scale.c libavfilter/vf_showinfo.c libavfilter/vf_transpose.c libavfilter/vf_vflip.c libavfilter/vf_yadif.c libavfilter/video.c libavfilter/vsrc_testsrc.c libavfilter/yadif.h Following are notes about the merge authorship and various technical details. Michael Niedermayer: * Main merge operation, notably avfilter.c and video.c * Switch to AVFrame: - afade - anullsrc - apad - aresample - blackframe - deshake - idet - il - mandelbrot - mptestsrc - noise - setfield - smartblur - tinterlace * various merge changes and fixes in: - ashowinfo - blackdetect - field - fps - select - testsrc - yadif Nicolas George: * Switch to AVFrame: - make rawdec work with refcounted frames. Adapted from commit 759001c534287a96dc96d1e274665feb7059145d by Anton Khirnov. Also, fix the use of || instead of | in a flags check. - make buffer sink and src, audio and video work all together Clément Bœsch: * Switch to AVFrame: - aevalsrc - alphaextract - blend - cellauto - colormatrix - concat - earwax - ebur128 - edgedetect - geq - histeq - histogram - hue - kerndeint - life - movie - mp (with the help of Michael) - overlay - pad - pan - pp - pp - removelogo - sendcmd - showspectrum - showwaves - silencedetect - stereo3d - subtitles - super2xsai - swapuv - thumbnail - tile Hendrik Leppkes: * Switch to AVFrame: - aconvert - amerge - asetnsamples - atempo - biquads Matthieu Bouron: * Switch to AVFrame - alphamerge - decimate - volumedetect Stefano Sabatini: * Switch to AVFrame: - astreamsync - flite - framestep Signed-off-by: Michael Niedermayer <michaelni@gmx.at> Signed-off-by: Nicolas George <nicolas.george@normalesup.org> Signed-off-by: Clément Bœsch <ubitux@gmail.com> Signed-off-by: Hendrik Leppkes <h.leppkes@gmail.com> Signed-off-by: Matthieu Bouron <matthieu.bouron@gmail.com> Signed-off-by: Stefano Sabatini <stefasab@gmail.com> Merged-by: Michael Niedermayer <michaelni@gmx.at>
2013-03-10 03:30:30 +03:00
static int filter_frame(AVFilterLink *inlink, AVFrame *insamples)
2012-01-04 01:47:09 +03:00
{
SilenceDetectContext *s = inlink->dst->priv;
const int nb_channels = inlink->ch_layout.nb_channels;
2012-01-04 01:47:09 +03:00
const int srate = inlink->sample_rate;
Merge commit '7e350379f87e7f74420b4813170fe808e2313911' * commit '7e350379f87e7f74420b4813170fe808e2313911': lavfi: switch to AVFrame. Conflicts: doc/filters.texi libavfilter/af_ashowinfo.c libavfilter/audio.c libavfilter/avfilter.c libavfilter/avfilter.h libavfilter/buffersink.c libavfilter/buffersrc.c libavfilter/buffersrc.h libavfilter/f_select.c libavfilter/f_setpts.c libavfilter/fifo.c libavfilter/split.c libavfilter/src_movie.c libavfilter/version.h libavfilter/vf_aspect.c libavfilter/vf_bbox.c libavfilter/vf_blackframe.c libavfilter/vf_delogo.c libavfilter/vf_drawbox.c libavfilter/vf_drawtext.c libavfilter/vf_fade.c libavfilter/vf_fieldorder.c libavfilter/vf_fps.c libavfilter/vf_frei0r.c libavfilter/vf_gradfun.c libavfilter/vf_hqdn3d.c libavfilter/vf_lut.c libavfilter/vf_overlay.c libavfilter/vf_pad.c libavfilter/vf_scale.c libavfilter/vf_showinfo.c libavfilter/vf_transpose.c libavfilter/vf_vflip.c libavfilter/vf_yadif.c libavfilter/video.c libavfilter/vsrc_testsrc.c libavfilter/yadif.h Following are notes about the merge authorship and various technical details. Michael Niedermayer: * Main merge operation, notably avfilter.c and video.c * Switch to AVFrame: - afade - anullsrc - apad - aresample - blackframe - deshake - idet - il - mandelbrot - mptestsrc - noise - setfield - smartblur - tinterlace * various merge changes and fixes in: - ashowinfo - blackdetect - field - fps - select - testsrc - yadif Nicolas George: * Switch to AVFrame: - make rawdec work with refcounted frames. Adapted from commit 759001c534287a96dc96d1e274665feb7059145d by Anton Khirnov. Also, fix the use of || instead of | in a flags check. - make buffer sink and src, audio and video work all together Clément Bœsch: * Switch to AVFrame: - aevalsrc - alphaextract - blend - cellauto - colormatrix - concat - earwax - ebur128 - edgedetect - geq - histeq - histogram - hue - kerndeint - life - movie - mp (with the help of Michael) - overlay - pad - pan - pp - pp - removelogo - sendcmd - showspectrum - showwaves - silencedetect - stereo3d - subtitles - super2xsai - swapuv - thumbnail - tile Hendrik Leppkes: * Switch to AVFrame: - aconvert - amerge - asetnsamples - atempo - biquads Matthieu Bouron: * Switch to AVFrame - alphamerge - decimate - volumedetect Stefano Sabatini: * Switch to AVFrame: - astreamsync - flite - framestep Signed-off-by: Michael Niedermayer <michaelni@gmx.at> Signed-off-by: Nicolas George <nicolas.george@normalesup.org> Signed-off-by: Clément Bœsch <ubitux@gmail.com> Signed-off-by: Hendrik Leppkes <h.leppkes@gmail.com> Signed-off-by: Matthieu Bouron <matthieu.bouron@gmail.com> Signed-off-by: Stefano Sabatini <stefasab@gmail.com> Merged-by: Michael Niedermayer <michaelni@gmx.at>
2013-03-10 03:30:30 +03:00
const int nb_samples = insamples->nb_samples * nb_channels;
const int64_t nb_samples_notify = s->duration * (s->mono ? 1 : nb_channels);
int c;
2012-01-04 01:47:09 +03:00
// scale number of null samples to the new sample rate
if (s->last_sample_rate && s->last_sample_rate != srate)
for (c = 0; c < s->independent_channels; c++) {
s->nb_null_samples[c] = srate * s->nb_null_samples[c] / s->last_sample_rate;
}
s->last_sample_rate = srate;
s->time_base = inlink->time_base;
s->frame_end = insamples->pts + av_rescale_q(insamples->nb_samples,
(AVRational){ 1, s->last_sample_rate }, inlink->time_base);
2012-01-04 01:47:09 +03:00
s->silencedetect(s, insamples, nb_samples, nb_samples_notify,
inlink->time_base);
2012-01-04 01:47:09 +03:00
return ff_filter_frame(inlink->dst->outputs[0], insamples);
2012-01-04 01:47:09 +03:00
}
static av_cold void uninit(AVFilterContext *ctx)
{
SilenceDetectContext *s = ctx->priv;
int c;
for (c = 0; c < s->independent_channels; c++)
if (s->start[c] > INT64_MIN)
update(s, NULL, 0, c, 0, s->time_base);
av_freep(&s->nb_null_samples);
av_freep(&s->start);
}
static const AVFilterPad silencedetect_inputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_AUDIO,
.config_props = config_input,
.filter_frame = filter_frame,
},
};
const AVFilter ff_af_silencedetect = {
2012-01-04 01:47:09 +03:00
.name = "silencedetect",
.description = NULL_IF_CONFIG_SMALL("Detect silence."),
.priv_size = sizeof(SilenceDetectContext),
.uninit = uninit,
2021-08-12 13:05:31 +02:00
FILTER_INPUTS(silencedetect_inputs),
FILTER_OUTPUTS(ff_audio_default_filterpad),
FILTER_SAMPLEFMTS(AV_SAMPLE_FMT_DBL, AV_SAMPLE_FMT_DBLP,
AV_SAMPLE_FMT_FLT, AV_SAMPLE_FMT_FLTP,
AV_SAMPLE_FMT_S32, AV_SAMPLE_FMT_S32P,
AV_SAMPLE_FMT_S16, AV_SAMPLE_FMT_S16P),
.priv_class = &silencedetect_class,
.flags = AVFILTER_FLAG_METADATA_ONLY,
2012-01-04 01:47:09 +03:00
};