/*
* Copyright (c) 1998 Juergen Mueller And Sundry Contributors
* This source code is freely redistributable and may be used for
* any purpose. This copyright notice must be maintained.
* Juergen Mueller And Sundry Contributors are not responsible for
* the consequences of using this software.
*
* Copyright (c) 2015 Paul B Mahol
*
* 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
* chorus audio filter
*/
#include "libavutil/avstring.h"
#include "libavutil/opt.h"
#include "audio.h"
#include "avfilter.h"
#include "internal.h"
#include "generate_wave_table.h"
typedef struct ChorusContext {
const AVClass *class;
float in_gain, out_gain;
char *delays_str;
char *decays_str;
char *speeds_str;
char *depths_str;
float *delays;
float *decays;
float *speeds;
float *depths;
uint8_t **chorusbuf;
int **phase;
int *length;
int32_t **lookup_table;
int *counter;
int num_chorus;
int max_samples;
int channels;
int modulation;
int fade_out;
int64_t next_pts;
} ChorusContext;
#define OFFSET(x) offsetof(ChorusContext, x)
#define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
static const AVOption chorus_options[] = {
{ "in_gain", "set input gain", OFFSET(in_gain), AV_OPT_TYPE_FLOAT, {.dbl=.4}, 0, 1, A },
{ "out_gain", "set output gain", OFFSET(out_gain), AV_OPT_TYPE_FLOAT, {.dbl=.4}, 0, 1, A },
{ "delays", "set delays", OFFSET(delays_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, A },
{ "decays", "set decays", OFFSET(decays_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, A },
{ "speeds", "set speeds", OFFSET(speeds_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, A },
{ "depths", "set depths", OFFSET(depths_str), AV_OPT_TYPE_STRING, {.str=NULL}, 0, 0, A },
{ NULL }
};
AVFILTER_DEFINE_CLASS(chorus);
static void count_items(char *item_str, int *nb_items)
{
char *p;
*nb_items = 1;
for (p = item_str; *p; p++) {
if (*p == '|')
(*nb_items)++;
}
}
static void fill_items(char *item_str, int *nb_items, float *items)
{
char *p, *saveptr = NULL;
int i, new_nb_items = 0;
p = item_str;
for (i = 0; i < *nb_items; i++) {
char *tstr = av_strtok(p, "|", &saveptr);
p = NULL;
new_nb_items += sscanf(tstr, "%f", &items[i]) == 1;
}
*nb_items = new_nb_items;
}
static av_cold int init(AVFilterContext *ctx)
{
ChorusContext *s = ctx->priv;
int nb_delays, nb_decays, nb_speeds, nb_depths;
if (!s->delays_str || !s->decays_str || !s->speeds_str || !s->depths_str) {
av_log(ctx, AV_LOG_ERROR, "Both delays & decays & speeds & depths must be set.\n");
return AVERROR(EINVAL);
}
count_items(s->delays_str, &nb_delays);
count_items(s->decays_str, &nb_decays);
count_items(s->speeds_str, &nb_speeds);
count_items(s->depths_str, &nb_depths);
s->delays = av_realloc_f(s->delays, nb_delays, sizeof(*s->delays));
s->decays = av_realloc_f(s->decays, nb_decays, sizeof(*s->decays));
s->speeds = av_realloc_f(s->speeds, nb_speeds, sizeof(*s->speeds));
s->depths = av_realloc_f(s->depths, nb_depths, sizeof(*s->depths));
if (!s->delays || !s->decays || !s->speeds || !s->depths)
return AVERROR(ENOMEM);
fill_items(s->delays_str, &nb_delays, s->delays);
fill_items(s->decays_str, &nb_decays, s->decays);
fill_items(s->speeds_str, &nb_speeds, s->speeds);
fill_items(s->depths_str, &nb_depths, s->depths);
if (nb_delays != nb_decays && nb_delays != nb_speeds && nb_delays != nb_depths) {
av_log(ctx, AV_LOG_ERROR, "Number of delays & decays & speeds & depths given must be same.\n");
return AVERROR(EINVAL);
}
s->num_chorus = nb_delays;
if (s->num_chorus < 1) {
av_log(ctx, AV_LOG_ERROR, "At least one delay & decay & speed & depth must be set.\n");
return AVERROR(EINVAL);
}
s->length = av_calloc(s->num_chorus, sizeof(*s->length));
s->lookup_table = av_calloc(s->num_chorus, sizeof(*s->lookup_table));
if (!s->length || !s->lookup_table)
return AVERROR(ENOMEM);
s->next_pts = AV_NOPTS_VALUE;
return 0;
}
static int query_formats(AVFilterContext *ctx)
{
AVFilterFormats *formats;
AVFilterChannelLayouts *layouts;
static const enum AVSampleFormat sample_fmts[] = {
AV_SAMPLE_FMT_FLTP, AV_SAMPLE_FMT_NONE
};
int 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_make_format_list(sample_fmts);
if (!formats)
return AVERROR(ENOMEM);
ret = ff_set_common_formats(ctx, formats);
if (ret < 0)
return ret;
formats = ff_all_samplerates();
if (!formats)
return AVERROR(ENOMEM);
return ff_set_common_samplerates(ctx, formats);
}
static int config_output(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
ChorusContext *s = ctx->priv;
float sum_in_volume = 1.0;
int n;
s->channels = outlink->channels;
for (n = 0; n < s->num_chorus; n++) {
int samples = (int) ((s->delays[n] + s->depths[n]) * outlink->sample_rate / 1000.0);
int depth_samples = (int) (s->depths[n] * outlink->sample_rate / 1000.0);
s->length[n] = outlink->sample_rate / s->speeds[n];
s->lookup_table[n] = av_malloc(sizeof(int32_t) * s->length[n]);
if (!s->lookup_table[n])
return AVERROR(ENOMEM);
ff_generate_wave_table(WAVE_SIN, AV_SAMPLE_FMT_S32, s->lookup_table[n],
s->length[n], 0., depth_samples, 0);
s->max_samples = FFMAX(s->max_samples, samples);
}
for (n = 0; n < s->num_chorus; n++)
sum_in_volume += s->decays[n];
if (s->in_gain * (sum_in_volume) > 1.0 / s->out_gain)
av_log(ctx, AV_LOG_WARNING, "output gain can cause saturation or clipping of output\n");
s->counter = av_calloc(outlink->channels, sizeof(*s->counter));
if (!s->counter)
return AVERROR(ENOMEM);
s->phase = av_calloc(outlink->channels, sizeof(*s->phase));
if (!s->phase)
return AVERROR(ENOMEM);
for (n = 0; n < outlink->channels; n++) {
s->phase[n] = av_calloc(s->num_chorus, sizeof(int));
if (!s->phase[n])
return AVERROR(ENOMEM);
}
s->fade_out = s->max_samples;
return av_samples_alloc_array_and_samples(&s->chorusbuf, NULL,
outlink->channels,
s->max_samples,
outlink->format, 0);
}
#define MOD(a, b) (((a) >= (b)) ? (a) - (b) : (a))
static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
{
AVFilterContext *ctx = inlink->dst;
ChorusContext *s = ctx->priv;
AVFrame *out_frame;
int c, i, n;
if (av_frame_is_writable(frame)) {
out_frame = frame;
} else {
out_frame = ff_get_audio_buffer(inlink, frame->nb_samples);
if (!out_frame) {
av_frame_free(&frame);
return AVERROR(ENOMEM);
}
av_frame_copy_props(out_frame, frame);
}
for (c = 0; c < inlink->channels; c++) {
const float *src = (const float *)frame->extended_data[c];
float *dst = (float *)out_frame->extended_data[c];
float *chorusbuf = (float *)s->chorusbuf[c];
int *phase = s->phase[c];
for (i = 0; i < frame->nb_samples; i++) {
float out, in = src[i];
out = in * s->in_gain;
for (n = 0; n < s->num_chorus; n++) {
out += chorusbuf[MOD(s->max_samples + s->counter[c] -
s->lookup_table[n][phase[n]],
s->max_samples)] * s->decays[n];
phase[n] = MOD(phase[n] + 1, s->length[n]);
}
out *= s->out_gain;
dst[i] = out;
chorusbuf[s->counter[c]] = in;
s->counter[c] = MOD(s->counter[c] + 1, s->max_samples);
}
}
s->next_pts = frame->pts + av_rescale_q(frame->nb_samples, (AVRational){1, inlink->sample_rate}, inlink->time_base);
if (frame != out_frame)
av_frame_free(&frame);
return ff_filter_frame(ctx->outputs[0], out_frame);
}
static int request_frame(AVFilterLink *outlink)
{
AVFilterContext *ctx = outlink->src;
ChorusContext *s = ctx->priv;
int ret;
ret = ff_request_frame(ctx->inputs[0]);
if (ret == AVERROR_EOF && !ctx->is_disabled && s->fade_out) {
int nb_samples = FFMIN(s->fade_out, 2048);
AVFrame *frame;
frame = ff_get_audio_buffer(outlink, nb_samples);
if (!frame)
return AVERROR(ENOMEM);
s->fade_out -= nb_samples;
av_samples_set_silence(frame->extended_data, 0,
frame->nb_samples,
outlink->channels,
frame->format);
frame->pts = s->next_pts;
if (s->next_pts != AV_NOPTS_VALUE)
s->next_pts += av_rescale_q(nb_samples, (AVRational){1, outlink->sample_rate}, outlink->time_base);
ret = filter_frame(ctx->inputs[0], frame);
}
return ret;
}
static av_cold void uninit(AVFilterContext *ctx)
{
ChorusContext *s = ctx->priv;
int n;
av_freep(&s->delays);
av_freep(&s->decays);
av_freep(&s->speeds);
av_freep(&s->depths);
if (s->chorusbuf)
av_freep(&s->chorusbuf[0]);
av_freep(&s->chorusbuf);
if (s->phase)
for (n = 0; n < s->channels; n++)
av_freep(&s->phase[n]);
av_freep(&s->phase);
av_freep(&s->counter);
av_freep(&s->length);
if (s->lookup_table)
for (n = 0; n < s->num_chorus; n++)
av_freep(&s->lookup_table[n]);
av_freep(&s->lookup_table);
}
static const AVFilterPad chorus_inputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_AUDIO,
.filter_frame = filter_frame,
},
{ NULL }
};
static const AVFilterPad chorus_outputs[] = {
{
.name = "default",
.type = AVMEDIA_TYPE_AUDIO,
.request_frame = request_frame,
.config_props = config_output,
},
{ NULL }
};
AVFilter ff_af_chorus = {
.name = "chorus",
.description = NULL_IF_CONFIG_SMALL("Add a chorus effect to the audio."),
.query_formats = query_formats,
.priv_size = sizeof(ChorusContext),
.priv_class = &chorus_class,
.init = init,
.uninit = uninit,
.inputs = chorus_inputs,
.outputs = chorus_outputs,
};