2 * Copyright (c) 2012 Nicolas George
4 * This file is part of FFmpeg.
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public License
8 * as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU Lesser General Public License for more details.
16 * You should have received a copy of the GNU Lesser General Public License
17 * along with FFmpeg; if not, write to the Free Software Foundation, Inc.,
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
21 #include "libavutil/channel_layout.h"
22 #include "libavutil/avassert.h"
27 typedef struct VolDetectContext {
29 * Number of samples at each PCM value.
30 * histogram[0x8000 + i] is the number of samples at value i.
31 * The extra element is there for symmetry.
33 uint64_t histogram[0x10001];
36 static int query_formats(AVFilterContext *ctx)
38 static const enum AVSampleFormat sample_fmts[] = {
43 AVFilterFormats *formats;
44 AVFilterChannelLayouts *layouts;
47 if (!(formats = ff_make_format_list(sample_fmts)))
48 return AVERROR(ENOMEM);
50 layouts = ff_all_channel_counts();
52 return AVERROR(ENOMEM);
53 ret = ff_set_common_channel_layouts(ctx, layouts);
57 return ff_set_common_formats(ctx, formats);
60 static int filter_frame(AVFilterLink *inlink, AVFrame *samples)
62 AVFilterContext *ctx = inlink->dst;
63 VolDetectContext *vd = ctx->priv;
64 int nb_samples = samples->nb_samples;
65 int nb_channels = samples->channels;
66 int nb_planes = nb_channels;
70 if (!av_sample_fmt_is_planar(samples->format)) {
71 nb_samples *= nb_channels;
74 for (plane = 0; plane < nb_planes; plane++) {
75 pcm = (int16_t *)samples->extended_data[plane];
76 for (i = 0; i < nb_samples; i++)
77 vd->histogram[pcm[i] + 0x8000]++;
80 return ff_filter_frame(inlink->dst->outputs[0], samples);
85 static inline double logdb(uint64_t v)
87 double d = v / (double)(0x8000 * 0x8000);
90 return -log10(d) * 10;
93 static void print_stats(AVFilterContext *ctx)
95 VolDetectContext *vd = ctx->priv;
96 int i, max_volume, shift;
97 uint64_t nb_samples = 0, power = 0, nb_samples_shift = 0, sum = 0;
98 uint64_t histdb[MAX_DB + 1] = { 0 };
100 for (i = 0; i < 0x10000; i++)
101 nb_samples += vd->histogram[i];
102 av_log(ctx, AV_LOG_INFO, "n_samples: %"PRId64"\n", nb_samples);
106 /* If nb_samples > 1<<34, there is a risk of overflow in the
107 multiplication or the sum: shift all histogram values to avoid that.
108 The total number of samples must be recomputed to avoid rounding
110 shift = av_log2(nb_samples >> 33);
111 for (i = 0; i < 0x10000; i++) {
112 nb_samples_shift += vd->histogram[i] >> shift;
113 power += (i - 0x8000) * (i - 0x8000) * (vd->histogram[i] >> shift);
115 if (!nb_samples_shift)
117 power = (power + nb_samples_shift / 2) / nb_samples_shift;
118 av_assert0(power <= 0x8000 * 0x8000);
119 av_log(ctx, AV_LOG_INFO, "mean_volume: %.1f dB\n", -logdb(power));
122 while (max_volume > 0 && !vd->histogram[0x8000 + max_volume] &&
123 !vd->histogram[0x8000 - max_volume])
125 av_log(ctx, AV_LOG_INFO, "max_volume: %.1f dB\n", -logdb(max_volume * max_volume));
127 for (i = 0; i < 0x10000; i++)
128 histdb[(int)logdb((i - 0x8000) * (i - 0x8000))] += vd->histogram[i];
129 for (i = 0; i <= MAX_DB && !histdb[i]; i++);
130 for (; i <= MAX_DB && sum < nb_samples / 1000; i++) {
131 av_log(ctx, AV_LOG_INFO, "histogram_%ddb: %"PRId64"\n", i, histdb[i]);
136 static av_cold void uninit(AVFilterContext *ctx)
141 static const AVFilterPad volumedetect_inputs[] = {
144 .type = AVMEDIA_TYPE_AUDIO,
145 .filter_frame = filter_frame,
150 static const AVFilterPad volumedetect_outputs[] = {
153 .type = AVMEDIA_TYPE_AUDIO,
158 AVFilter ff_af_volumedetect = {
159 .name = "volumedetect",
160 .description = NULL_IF_CONFIG_SMALL("Detect audio volume."),
161 .priv_size = sizeof(VolDetectContext),
162 .query_formats = query_formats,
164 .inputs = volumedetect_inputs,
165 .outputs = volumedetect_outputs,