238 lines
8.0 KiB
C
238 lines
8.0 KiB
C
/*
|
|
* Copyright (C) 2001-2010 Krzysztof Foltman, Markus Schmidt, Thor Harald Johansen, Damien Zammit
|
|
*
|
|
* 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
|
|
*/
|
|
|
|
#include "libavutil/channel_layout.h"
|
|
#include "libavutil/opt.h"
|
|
#include "avfilter.h"
|
|
#include "audio.h"
|
|
#include "formats.h"
|
|
#include "hermite.h"
|
|
|
|
typedef struct AudioGateContext {
|
|
const AVClass *class;
|
|
|
|
double level_in;
|
|
double attack;
|
|
double release;
|
|
double threshold;
|
|
double ratio;
|
|
double knee;
|
|
double makeup;
|
|
double range;
|
|
int link;
|
|
int detection;
|
|
|
|
double thres;
|
|
double knee_start;
|
|
double lin_knee_stop;
|
|
double knee_stop;
|
|
double lin_slope;
|
|
double attack_coeff;
|
|
double release_coeff;
|
|
} AudioGateContext;
|
|
|
|
#define OFFSET(x) offsetof(AudioGateContext, x)
|
|
#define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
|
|
|
|
static const AVOption agate_options[] = {
|
|
{ "level_in", "set input level", OFFSET(level_in), AV_OPT_TYPE_DOUBLE, {.dbl=1}, 0.015625, 64, A },
|
|
{ "range", "set max gain reduction", OFFSET(range), AV_OPT_TYPE_DOUBLE, {.dbl=0.06125}, 0, 1, A },
|
|
{ "threshold", "set threshold", OFFSET(threshold), AV_OPT_TYPE_DOUBLE, {.dbl=0.125}, 0, 1, A },
|
|
{ "ratio", "set ratio", OFFSET(ratio), AV_OPT_TYPE_DOUBLE, {.dbl=2}, 1, 9000, A },
|
|
{ "attack", "set attack", OFFSET(attack), AV_OPT_TYPE_DOUBLE, {.dbl=20}, 0.01, 9000, A },
|
|
{ "release", "set release", OFFSET(release), AV_OPT_TYPE_DOUBLE, {.dbl=250}, 0.01, 9000, A },
|
|
{ "makeup", "set makeup gain", OFFSET(makeup), AV_OPT_TYPE_DOUBLE, {.dbl=1}, 1, 64, A },
|
|
{ "knee", "set knee", OFFSET(knee), AV_OPT_TYPE_DOUBLE, {.dbl=2.828427125}, 1, 8, A },
|
|
{ "detection", "set detection", OFFSET(detection), AV_OPT_TYPE_INT, {.i64=0}, 0, 1, A, "detection" },
|
|
{ "peak", 0, 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, A, "detection" },
|
|
{ "rms", 0, 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, A, "detection" },
|
|
{ "link", "set link", OFFSET(link), AV_OPT_TYPE_INT, {.i64=0}, 0, 1, A, "link" },
|
|
{ "average", 0, 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, A, "link" },
|
|
{ "maximum", 0, 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, A, "link" },
|
|
{ NULL }
|
|
};
|
|
|
|
AVFILTER_DEFINE_CLASS(agate);
|
|
|
|
static int query_formats(AVFilterContext *ctx)
|
|
{
|
|
AVFilterFormats *formats = NULL;
|
|
AVFilterChannelLayouts *layouts;
|
|
int ret;
|
|
|
|
ff_add_format(&formats, AV_SAMPLE_FMT_DBL);
|
|
ret = ff_set_common_formats(ctx, formats);
|
|
if (ret < 0)
|
|
return ret;
|
|
|
|
layouts = ff_all_channel_counts();
|
|
if (!layouts)
|
|
return AVERROR(ENOMEM);
|
|
ret = ff_set_common_channel_layouts(ctx, layouts);
|
|
if (ret < 0)
|
|
return ret;
|
|
|
|
formats = ff_all_samplerates();
|
|
if (!formats)
|
|
return AVERROR(ENOMEM);
|
|
|
|
return ff_set_common_samplerates(ctx, formats);
|
|
}
|
|
|
|
static int config_input(AVFilterLink *inlink)
|
|
{
|
|
AVFilterContext *ctx = inlink->dst;
|
|
AudioGateContext *s = ctx->priv;
|
|
double lin_threshold = s->threshold;
|
|
double lin_knee_sqrt = sqrt(s->knee);
|
|
double lin_knee_start;
|
|
|
|
if (s->detection)
|
|
lin_threshold *= lin_threshold;
|
|
|
|
s->attack_coeff = FFMIN(1., 1. / (s->attack * inlink->sample_rate / 4000.));
|
|
s->release_coeff = FFMIN(1., 1. / (s->release * inlink->sample_rate / 4000.));
|
|
s->lin_knee_stop = lin_threshold * lin_knee_sqrt;
|
|
lin_knee_start = lin_threshold / lin_knee_sqrt;
|
|
s->thres = log(lin_threshold);
|
|
s->knee_start = log(lin_knee_start);
|
|
s->knee_stop = log(s->lin_knee_stop);
|
|
|
|
return 0;
|
|
}
|
|
|
|
// A fake infinity value (because real infinity may break some hosts)
|
|
#define FAKE_INFINITY (65536.0 * 65536.0)
|
|
|
|
// Check for infinity (with appropriate-ish tolerance)
|
|
#define IS_FAKE_INFINITY(value) (fabs(value-FAKE_INFINITY) < 1.0)
|
|
|
|
static double output_gain(double lin_slope, double ratio, double thres,
|
|
double knee, double knee_start, double knee_stop,
|
|
double lin_knee_stop, double range)
|
|
{
|
|
if (lin_slope < lin_knee_stop) {
|
|
double slope = log(lin_slope);
|
|
double tratio = ratio;
|
|
double gain = 0.;
|
|
double delta = 0.;
|
|
|
|
if (IS_FAKE_INFINITY(ratio))
|
|
tratio = 1000.;
|
|
gain = (slope - thres) * tratio + thres;
|
|
delta = tratio;
|
|
|
|
if (knee > 1. && slope > knee_start) {
|
|
gain = hermite_interpolation(slope, knee_start, knee_stop, ((knee_start - thres) * tratio + thres), knee_stop, delta, 1.);
|
|
}
|
|
return FFMAX(range, exp(gain - slope));
|
|
}
|
|
|
|
return 1.;
|
|
}
|
|
|
|
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
|
|
{
|
|
AVFilterContext *ctx = inlink->dst;
|
|
AVFilterLink *outlink = ctx->outputs[0];
|
|
AudioGateContext *s = ctx->priv;
|
|
const double *src = (const double *)in->data[0];
|
|
const double makeup = s->makeup;
|
|
const double attack_coeff = s->attack_coeff;
|
|
const double release_coeff = s->release_coeff;
|
|
const double level_in = s->level_in;
|
|
AVFrame *out;
|
|
double *dst;
|
|
int n, c;
|
|
|
|
if (av_frame_is_writable(in)) {
|
|
out = in;
|
|
} else {
|
|
out = ff_get_audio_buffer(inlink, in->nb_samples);
|
|
if (!out) {
|
|
av_frame_free(&in);
|
|
return AVERROR(ENOMEM);
|
|
}
|
|
av_frame_copy_props(out, in);
|
|
}
|
|
dst = (double *)out->data[0];
|
|
|
|
for (n = 0; n < in->nb_samples; n++, src += inlink->channels, dst += inlink->channels) {
|
|
double abs_sample = FFABS(src[0]), gain = 1.0;
|
|
|
|
for (c = 0; c < inlink->channels; c++)
|
|
dst[c] = src[c] * level_in;
|
|
|
|
if (s->link == 1) {
|
|
for (c = 1; c < inlink->channels; c++)
|
|
abs_sample = FFMAX(FFABS(src[c]), abs_sample);
|
|
} else {
|
|
for (c = 1; c < inlink->channels; c++)
|
|
abs_sample += FFABS(src[c]);
|
|
|
|
abs_sample /= inlink->channels;
|
|
}
|
|
|
|
if (s->detection)
|
|
abs_sample *= abs_sample;
|
|
|
|
s->lin_slope += (abs_sample - s->lin_slope) * (abs_sample > s->lin_slope ? attack_coeff : release_coeff);
|
|
if (s->lin_slope > 0.0)
|
|
gain = output_gain(s->lin_slope, s->ratio, s->thres,
|
|
s->knee, s->knee_start, s->knee_stop,
|
|
s->lin_knee_stop, s->range);
|
|
|
|
for (c = 0; c < inlink->channels; c++)
|
|
dst[c] *= gain * makeup;
|
|
}
|
|
|
|
if (out != in)
|
|
av_frame_free(&in);
|
|
return ff_filter_frame(outlink, out);
|
|
}
|
|
|
|
static const AVFilterPad inputs[] = {
|
|
{
|
|
.name = "default",
|
|
.type = AVMEDIA_TYPE_AUDIO,
|
|
.filter_frame = filter_frame,
|
|
.config_props = config_input,
|
|
},
|
|
{ NULL }
|
|
};
|
|
|
|
static const AVFilterPad outputs[] = {
|
|
{
|
|
.name = "default",
|
|
.type = AVMEDIA_TYPE_AUDIO,
|
|
},
|
|
{ NULL }
|
|
};
|
|
|
|
AVFilter ff_af_agate = {
|
|
.name = "agate",
|
|
.description = NULL_IF_CONFIG_SMALL("Audio gate."),
|
|
.query_formats = query_formats,
|
|
.priv_size = sizeof(AudioGateContext),
|
|
.priv_class = &agate_class,
|
|
.inputs = inputs,
|
|
.outputs = outputs,
|
|
};
|