lavfi: add audio silencedetect filter.

Changelog

@@ -20,6 +20,7 @@ version next:
 - v308 Quicktime Uncompressed 4:4:4 encoder and decoder
 - yuv4 libquicktime packed 4:2:0 encoder and decoder
 - ffprobe -show_frames option
+- silencedetect audio filter
 
 
 version 0.9:

doc/filters.texi

@@ -358,6 +358,36 @@ 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.
 
+@section silencedetect
+
+Detect silence in an audio stream.
+
+This filter logs a message when it detects that the input audio volume is less
+or equal to a noise tolerance value for a duration greater or equal to the
+minimum detected noise duration.
+
+The printed times and duration are expressed in seconds.
+
+@table @option
+@item duration, d
+Set silence duration until notification (default is 2 seconds).
+
+@item noise, n
+Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
+specified value) or amplitude ratio. Default is -60dB, or 0.001.
+@end table
+
+Detect 5 seconds of silence with -50dB noise tolerance:
+@example
+silencedetect=n=-50dB:d=5
+@end example
+
+Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
+tolerance in @file{silence.mp3}:
+@example
+ffmpeg -f lavfi -i amovie=silence.mp3,silencedetect=noise=0.0001 -f null -
+@end example
+
 @section volume
 
 Adjust the input audio volume.

libavfilter/Makefile

@@ -34,6 +34,7 @@ OBJS-$(CONFIG_ASPLIT_FILTER)                 += af_asplit.o
 OBJS-$(CONFIG_ASTREAMSYNC_FILTER)            += af_astreamsync.o
 OBJS-$(CONFIG_EARWAX_FILTER)                 += af_earwax.o
 OBJS-$(CONFIG_PAN_FILTER)                    += af_pan.o
+OBJS-$(CONFIG_SILENCEDETECT_FILTER)          += af_silencedetect.o
 OBJS-$(CONFIG_VOLUME_FILTER)                 += af_volume.o
 
 OBJS-$(CONFIG_ABUFFER_FILTER)                += asrc_abuffer.o

libavfilter/af_silencedetect.c

@@ -0,0 +1,175 @@
+/*
+ * Copyright (c) 2012 Clément Bœsch <ubitux@gmail.com>
+ *
+ * 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 "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 of silence until notification
+    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);
+    } else if (*tail) {
+        av_log(ctx, AV_LOG_ERROR, "Invalid value '%s' for noise parameter.\n",
+               silence->noise_str);
+        return AVERROR(EINVAL);
+    }
+
+    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 srate                 = inlink->sample_rate;
+    const int nb_samples            = insamples->audio->nb_samples * nb_channels;
+    const int64_t nb_samples_notify = srate * silence->duration    * nb_channels;
+
+    // scale number of null samples to the new sample rate
+    if (silence->last_sample_rate && silence->last_sample_rate != srate)
+        silence->nb_null_samples =
+            srate * silence->nb_null_samples / silence->last_sample_rate;
+    silence->last_sample_rate = srate;
+
+    // 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_start: %f\n", silence->start);
+                    }
+                }
+            } else {
+                if (silence->start) {
+                    double end = insamples->pts * av_q2d(inlink->time_base);
+                    av_log(silence, AV_LOG_INFO,
+                           "silence_end: %f | silence_duration: %f\n",
+                           end, end - silence->start);
+                }
+                silence->nb_null_samples = silence->start = 0;
+            }
+        }
+    }
+
+    avfilter_filter_samples(inlink->dst->outputs[0], insamples);
+}
+
+static int query_formats(AVFilterContext *ctx)
+{
+    AVFilterFormats *formats = NULL;
+    enum AVSampleFormat sample_fmts[] = {
+        AV_SAMPLE_FMT_DBL,
+        AV_SAMPLE_FMT_NONE
+    };
+    int packing_fmts[] = { AVFILTER_PACKED, -1 };
+
+    formats = avfilter_make_all_channel_layouts();
+    if (!formats)
+        return AVERROR(ENOMEM);
+    avfilter_set_common_channel_layouts(ctx, formats);
+
+    formats = avfilter_make_format_list(sample_fmts);
+    if (!formats)
+        return AVERROR(ENOMEM);
+    avfilter_set_common_sample_formats(ctx, formats);
+
+    formats = avfilter_make_format_list(packing_fmts);
+    if (!formats)
+        return AVERROR(ENOMEM);
+    avfilter_set_common_packing_formats(ctx, formats);
+
+    return 0;
+}
+
+AVFilter avfilter_af_silencedetect = {
+    .name          = "silencedetect",
+    .description   = NULL_IF_CONFIG_SMALL("Detect silence."),
+    .priv_size     = sizeof(SilenceDetectContext),
+    .init          = init,
+    .query_formats = query_formats,
+
+    .inputs = (const AVFilterPad[]) {
+        { .name             = "default",
+          .type             = AVMEDIA_TYPE_AUDIO,
+          .get_audio_buffer = avfilter_null_get_audio_buffer,
+          .filter_samples   = filter_samples, },
+        { .name = NULL }
+    },
+    .outputs = (const AVFilterPad[]) {
+        { .name = "default",
+          .type = AVMEDIA_TYPE_AUDIO, },
+        { .name = NULL }
+    },
+};

libavfilter/allfilters.c

@@ -44,6 +44,7 @@ void avfilter_register_all(void)
     REGISTER_FILTER (ASTREAMSYNC, astreamsync, af);
     REGISTER_FILTER (EARWAX,      earwax,      af);
     REGISTER_FILTER (PAN,         pan,         af);
+    REGISTER_FILTER (SILENCEDETECT, silencedetect, af);
     REGISTER_FILTER (VOLUME,      volume,      af);
 
     REGISTER_FILTER (ABUFFER,     abuffer,     asrc);

libavfilter/avfilter.h

@@ -30,8 +30,8 @@
 #include "libavcodec/avcodec.h"
 
 #define LIBAVFILTER_VERSION_MAJOR  2
-#define LIBAVFILTER_VERSION_MINOR 57
+#define LIBAVFILTER_VERSION_MINOR 58
-#define LIBAVFILTER_VERSION_MICRO 101
+#define LIBAVFILTER_VERSION_MICRO 100
 
 #define LIBAVFILTER_VERSION_INT AV_VERSION_INT(LIBAVFILTER_VERSION_MAJOR, \
                                                LIBAVFILTER_VERSION_MINOR, \