[FFmpeg-devel] [PATCH 2/2] lavfi: add audio silencedetect filter.
Stefano Sabatini
stefasab at gmail.com
Sat Jan 7 21:37:09 CET 2012
On date Saturday 2012-01-07 15:41:05 +0100, Clément Bœsch encoded:
> On Thu, Jan 05, 2012 at 08:29:35PM +0100, Nicolas George wrote:
[...]
> From 2e7c6e75d225818b0eacf29f5599dfd308804406 Mon Sep 17 00:00:00 2001
> From: =?UTF-8?q?Cl=C3=A9ment=20B=C5=93sch?= <ubitux at gmail.com>
> Date: Tue, 3 Jan 2012 23:47:09 +0100
> Subject: [PATCH] lavfi: add audio silencedetect filter.
>
> ---
> Changelog | 1 +
> doc/filters.texi | 24 ++++++
> libavfilter/Makefile | 1 +
> libavfilter/af_silencedetect.c | 171 ++++++++++++++++++++++++++++++++++++++++
> libavfilter/allfilters.c | 1 +
> libavfilter/avfilter.h | 4 +-
> 6 files changed, 200 insertions(+), 2 deletions(-)
> create mode 100644 libavfilter/af_silencedetect.c
>
> diff --git a/Changelog b/Changelog
> index db4c7b3..d4ec698 100644
> --- a/Changelog
> +++ b/Changelog
> @@ -19,6 +19,7 @@ version next:
> - Avid 1:1 10-bit RGB Packer decoder
> - v308 Quicktime Uncompressed 4:4:4 encoder and decoder
> - yuv4 libquicktime packed 4:2:0 encoder and decoder
> +- silencedetect audio filter
>
>
> version 0.9:
> diff --git a/doc/filters.texi b/doc/filters.texi
> index de73e3f..3bc117c 100644
> --- a/doc/filters.texi
> +++ b/doc/filters.texi
> @@ -358,6 +358,30 @@ Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
> that should be preferred (see "-ac" option) unless you have very specific
> needs.
>
> + at section silencedetect
> +
> +Print a notification line when silence is detected, or when sound is back in
> +the given audio stream.
This needs more explanation. For example:
This filter logs a message when detect that the input audio volume is
less or equal to the noise tolerance value set with @option{noise} for
a duration greater or equal to the silence duration value set with
@option{duration}.
> +
> + at table @option
> + at item d, duration
> +Set silence duration until notification (default is 2 seconds).
> +
> + at item noise, n
> +Set noise tolerance. Can be specified in dB or amplitude ratio. Default is
in db (in case "dB" is appended to the specified value).
> +-60dB, or 0.001.
> + at end table
> +
> +Detect 5 seconds of silence with -50dB noise tolerance:
> + at example
> +silencedetect=n=-50dB:d=5
> + at end example
> +
> +Complete example with ffmpeg to detect silence with 0.0001 noise tolerance:
> + at example
> +ffmpeg -f lavfi -i amovie=silence.mp3,silencedetect=noise=0.0001 -f null -
> + at end example
> +
> @section volume
>
> Adjust the input audio volume.
> diff --git a/libavfilter/Makefile b/libavfilter/Makefile
> index 0d8f120..bb5748b 100644
> --- a/libavfilter/Makefile
> +++ b/libavfilter/Makefile
> @@ -35,6 +35,7 @@ OBJS-$(CONFIG_ASTREAMSYNC_FILTER) += af_astreamsync.o
> OBJS-$(CONFIG_EARWAX_FILTER) += af_earwax.o
> OBJS-$(CONFIG_PAN_FILTER) += af_pan.o
> OBJS-$(CONFIG_VOLUME_FILTER) += af_volume.o
> +OBJS-$(CONFIG_SILENCEDETECT_FILTER) += af_silencedetect.o
>
> OBJS-$(CONFIG_ABUFFER_FILTER) += asrc_abuffer.o
> OBJS-$(CONFIG_AEVALSRC_FILTER) += asrc_aevalsrc.o
> diff --git a/libavfilter/af_silencedetect.c b/libavfilter/af_silencedetect.c
> new file mode 100644
> index 0000000..fe798dc
> --- /dev/null
> +++ b/libavfilter/af_silencedetect.c
> @@ -0,0 +1,171 @@
> +/*
> + * This file is part of FFmpeg.
> + *
> + * Copyright (c) 2011 Clément Bœsch <ubitux at gmail.com>
> + *
> + * 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 "libavutil/opt.h"
> +#include "avfilter.h"
> +
> +typedef struct {
> + const AVClass *class;
> + char *noise_str; ///< noise option string
> + double noise; ///< noise amplitude ratio
> + int duration; ///< minimum duration to monitor in seconds
> + int64_t nb_null_samples; ///< current number of continuous zero samples
> + double start; ///< if silence is detected, this value contains the time of the first zero sample
> + int last_sample_rate; ///< last sample rate to check for sample rate changes
> +} SilenceDetectContext;
> +
> +#define OFFSET(x) offsetof(SilenceDetectContext, x)
> +static const AVOption silencedetect_options[] = {
> + { "n", "set noise tolerance", OFFSET(noise_str), AV_OPT_TYPE_STRING, {.str="-60dB"}, CHAR_MIN, CHAR_MAX },
> + { "noise", "set noise tolerance", OFFSET(noise_str), AV_OPT_TYPE_STRING, {.str="-60dB"}, CHAR_MIN, CHAR_MAX },
> + { "d", "set minimum duration in seconds", OFFSET(duration), AV_OPT_TYPE_INT, {.dbl=2}, 0, INT_MAX},
> + { "duration", "set minimum duration in seconds", OFFSET(duration), AV_OPT_TYPE_INT, {.dbl=2}, 0, INT_MAX},
> + { NULL },
> +};
> +
> +static const char *silencedetect_get_name(void *ctx)
> +{
> + return "silencedetect";
> +}
> +
> +static const AVClass silencedetect_class = {
> + .class_name = "SilenceDetectContext",
> + .item_name = silencedetect_get_name,
> + .option = silencedetect_options,
> +};
> +
> +static av_cold int init(AVFilterContext *ctx, const char *args, void *opaque)
> +{
> + int ret;
> + char *tail;
> + SilenceDetectContext *silence = ctx->priv;
> +
> + silence->class = &silencedetect_class;
> + av_opt_set_defaults(silence);
> +
> + if ((ret = av_set_options_string(silence, args, "=", ":")) < 0) {
> + av_log(ctx, AV_LOG_ERROR, "Error parsing options string: '%s'\n", args);
> + return ret;
> + }
> +
> + silence->noise = strtod(silence->noise_str, &tail);
> + if (!strcmp(tail, "dB"))
> + silence->noise = pow(10, silence->noise/20);
Alternatively you may want to create an internal function, to share
with af_volume(). Also this is missing a check, in case noise_str
contains a foobar value.
> + return 0;
> +}
> +
> +static void filter_samples(AVFilterLink *inlink, AVFilterBufferRef *insamples)
> +{
> + int i;
> + SilenceDetectContext *silence = inlink->dst->priv;
> + const int nb_channels = av_get_channel_layout_nb_channels(inlink->channel_layout);
> + const int nb_samples = insamples->audio->nb_samples * nb_channels;
> + const int64_t nb_samples_notify = inlink->sample_rate * silence->duration * nb_channels;
alternatively you can cache nb_samples_notify, and update it when
sample_rate changes.
> +
> + if (silence->last_sample_rate &&
> + silence->last_sample_rate != inlink->sample_rate) {
> + av_log(silence, AV_LOG_WARNING,
> + "Sample rate changed, detection might not be accurate\n");
> + silence->start = silence->nb_null_samples = 0;
> + }
> + silence->last_sample_rate = inlink->sample_rate;
> +
> + // TODO: support more sample formats
> + if (insamples->format == AV_SAMPLE_FMT_DBL) {
> + double *p = (double *)insamples->data[0];
> +
> + for (i = 0; i < nb_samples; i++, p++) {
> + if (*p < silence->noise && *p > -silence->noise) {
> + if (!silence->start) {
> + silence->nb_null_samples++;
> + if (silence->nb_null_samples == nb_samples_notify) {
> + silence->start = insamples->pts * av_q2d(inlink->time_base) - silence->duration;
> + av_log(silence, AV_LOG_INFO,
> + "Silence detected at %f sec\n", silence->start);
what about "silence_start:%f"
=> easier to grep
> + }
> + }
> + } else {
> + if (silence->start) {
> + double end = insamples->pts * av_q2d(inlink->time_base);
> + av_log(silence, AV_LOG_INFO,
> + "Silence ended at %f sec (duration: %f sec)\n",
same here:
"silence_end:%f silence_duration:%s"
[...]
--
FFmpeg = Fast and Frightening Multipurpose Pontificating Everlasting God
More information about the ffmpeg-devel
mailing list