FFmpeg
af_aecho.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2013 Paul B Mahol
3  *
4  * This file is part of FFmpeg.
5  *
6  * FFmpeg is free software; you can redistribute it and/or
7  * modify it under the terms of the GNU Lesser General Public
8  * License as published by the Free Software Foundation; either
9  * version 2.1 of the License, or (at your option) any later version.
10  *
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 GNU
14  * Lesser General Public License for more details.
15  *
16  * You should have received a copy of the GNU Lesser General Public
17  * License along with FFmpeg; if not, write to the Free Software
18  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19  */
20 
21 #include "libavutil/avassert.h"
22 #include "libavutil/avstring.h"
23 #include "libavutil/mem.h"
24 #include "libavutil/opt.h"
25 #include "libavutil/samplefmt.h"
26 #include "avfilter.h"
27 #include "audio.h"
28 #include "filters.h"
29 
30 typedef struct AudioEchoContext {
31  const AVClass *class;
32  float in_gain, out_gain;
33  char *delays, *decays;
34  float *delay, *decay;
35  int nb_echoes;
37  uint8_t **delayptrs;
39  int *samples;
40  int eof;
42 
43  void (*echo_samples)(struct AudioEchoContext *ctx, uint8_t **delayptrs,
44  uint8_t * const *src, uint8_t **dst,
45  int nb_samples, int channels);
47 
48 #define OFFSET(x) offsetof(AudioEchoContext, x)
49 #define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
50 
51 static const AVOption aecho_options[] = {
52  { "in_gain", "set signal input gain", OFFSET(in_gain), AV_OPT_TYPE_FLOAT, {.dbl=0.6}, 0, 1, A },
53  { "out_gain", "set signal output gain", OFFSET(out_gain), AV_OPT_TYPE_FLOAT, {.dbl=0.3}, 0, 1, A },
54  { "delays", "set list of signal delays", OFFSET(delays), AV_OPT_TYPE_STRING, {.str="1000"}, 0, 0, A },
55  { "decays", "set list of signal decays", OFFSET(decays), AV_OPT_TYPE_STRING, {.str="0.5"}, 0, 0, A },
56  { NULL }
57 };
58 
60 
61 static void count_items(char *item_str, int *nb_items)
62 {
63  char *p;
64 
65  *nb_items = 1;
66  for (p = item_str; *p; p++) {
67  if (*p == '|')
68  (*nb_items)++;
69  }
70 
71 }
72 
73 static void fill_items(char *item_str, int *nb_items, float *items)
74 {
75  char *p, *saveptr = NULL;
76  int i, new_nb_items = 0;
77 
78  p = item_str;
79  for (i = 0; i < *nb_items; i++) {
80  char *tstr = av_strtok(p, "|", &saveptr);
81  p = NULL;
82  if (tstr)
83  new_nb_items += av_sscanf(tstr, "%f", &items[new_nb_items]) == 1;
84  }
85 
86  *nb_items = new_nb_items;
87 }
88 
90 {
91  AudioEchoContext *s = ctx->priv;
92 
93  av_freep(&s->delay);
94  av_freep(&s->decay);
95  av_freep(&s->samples);
96 
97  if (s->delayptrs)
98  av_freep(&s->delayptrs[0]);
99  av_freep(&s->delayptrs);
100 }
101 
103 {
104  AudioEchoContext *s = ctx->priv;
105  int nb_delays, nb_decays, i;
106 
107  if (!s->delays || !s->decays) {
108  av_log(ctx, AV_LOG_ERROR, "Missing delays and/or decays.\n");
109  return AVERROR(EINVAL);
110  }
111 
112  count_items(s->delays, &nb_delays);
113  count_items(s->decays, &nb_decays);
114 
115  s->delay = av_realloc_f(s->delay, nb_delays, sizeof(*s->delay));
116  s->decay = av_realloc_f(s->decay, nb_decays, sizeof(*s->decay));
117  if (!s->delay || !s->decay)
118  return AVERROR(ENOMEM);
119 
120  fill_items(s->delays, &nb_delays, s->delay);
121  fill_items(s->decays, &nb_decays, s->decay);
122 
123  if (nb_delays != nb_decays) {
124  av_log(ctx, AV_LOG_ERROR, "Number of delays %d differs from number of decays %d.\n", nb_delays, nb_decays);
125  return AVERROR(EINVAL);
126  }
127 
128  s->nb_echoes = nb_delays;
129  if (!s->nb_echoes) {
130  av_log(ctx, AV_LOG_ERROR, "At least one decay & delay must be set.\n");
131  return AVERROR(EINVAL);
132  }
133 
134  s->samples = av_realloc_f(s->samples, nb_delays, sizeof(*s->samples));
135  if (!s->samples)
136  return AVERROR(ENOMEM);
137 
138  for (i = 0; i < nb_delays; i++) {
139  if (s->delay[i] <= 0 || s->delay[i] > 90000) {
140  av_log(ctx, AV_LOG_ERROR, "delay[%d]: %f is out of allowed range: (0, 90000]\n", i, s->delay[i]);
141  return AVERROR(EINVAL);
142  }
143  if (s->decay[i] <= 0 || s->decay[i] > 1) {
144  av_log(ctx, AV_LOG_ERROR, "decay[%d]: %f is out of allowed range: (0, 1]\n", i, s->decay[i]);
145  return AVERROR(EINVAL);
146  }
147  }
148 
149  s->next_pts = AV_NOPTS_VALUE;
150 
151  av_log(ctx, AV_LOG_DEBUG, "nb_echoes:%d\n", s->nb_echoes);
152  return 0;
153 }
154 
155 #define MOD(a, b) (((a) >= (b)) ? (a) - (b) : (a))
156 
157 #define ECHO(name, type, min, max) \
158 static void echo_samples_## name ##p(AudioEchoContext *ctx, \
159  uint8_t **delayptrs, \
160  uint8_t * const *src, uint8_t **dst, \
161  int nb_samples, int channels) \
162 { \
163  const double out_gain = ctx->out_gain; \
164  const double in_gain = ctx->in_gain; \
165  const int nb_echoes = ctx->nb_echoes; \
166  const int max_samples = ctx->max_samples; \
167  int i, j, chan, av_uninit(index); \
168  \
169  av_assert1(channels > 0); /* would corrupt delay_index */ \
170  \
171  for (chan = 0; chan < channels; chan++) { \
172  const type *s = (type *)src[chan]; \
173  type *d = (type *)dst[chan]; \
174  type *dbuf = (type *)delayptrs[chan]; \
175  \
176  index = ctx->delay_index; \
177  for (i = 0; i < nb_samples; i++, s++, d++) { \
178  double out, in; \
179  \
180  in = *s; \
181  out = in * in_gain; \
182  for (j = 0; j < nb_echoes; j++) { \
183  int ix = index + max_samples - ctx->samples[j]; \
184  ix = MOD(ix, max_samples); \
185  out += dbuf[ix] * ctx->decay[j]; \
186  } \
187  out *= out_gain; \
188  \
189  *d = av_clipd(out, min, max); \
190  dbuf[index] = in; \
191  \
192  index = MOD(index + 1, max_samples); \
193  } \
194  } \
195  ctx->delay_index = index; \
196 }
197 
198 ECHO(dbl, double, -1.0, 1.0 )
199 ECHO(flt, float, -1.0, 1.0 )
200 ECHO(s16, int16_t, INT16_MIN, INT16_MAX)
201 ECHO(s32, int32_t, INT32_MIN, INT32_MAX)
202 
203 static int config_output(AVFilterLink *outlink)
204 {
205  AVFilterContext *ctx = outlink->src;
206  AudioEchoContext *s = ctx->priv;
207  float volume = 1.0;
208  int i;
209 
210  for (i = 0; i < s->nb_echoes; i++) {
211  s->samples[i] = s->delay[i] * outlink->sample_rate / 1000.0;
212  s->max_samples = FFMAX(s->max_samples, s->samples[i]);
213  volume += s->decay[i];
214  }
215 
216  if (s->max_samples <= 0) {
217  av_log(ctx, AV_LOG_ERROR, "Nothing to echo - missing delay samples.\n");
218  return AVERROR(EINVAL);
219  }
220  s->fade_out = s->max_samples;
221 
222  if (volume * s->in_gain * s->out_gain > 1.0)
224  "out_gain %f can cause saturation of output\n", s->out_gain);
225 
226  switch (outlink->format) {
227  case AV_SAMPLE_FMT_DBLP: s->echo_samples = echo_samples_dblp; break;
228  case AV_SAMPLE_FMT_FLTP: s->echo_samples = echo_samples_fltp; break;
229  case AV_SAMPLE_FMT_S16P: s->echo_samples = echo_samples_s16p; break;
230  case AV_SAMPLE_FMT_S32P: s->echo_samples = echo_samples_s32p; break;
231  }
232 
233 
234  if (s->delayptrs)
235  av_freep(&s->delayptrs[0]);
236  av_freep(&s->delayptrs);
237 
238  return av_samples_alloc_array_and_samples(&s->delayptrs, NULL,
239  outlink->ch_layout.nb_channels,
240  s->max_samples,
241  outlink->format, 0);
242 }
243 
245 {
246  AVFilterContext *ctx = inlink->dst;
247  AudioEchoContext *s = ctx->priv;
248  AVFrame *out_frame;
249 
251  out_frame = frame;
252  } else {
253  out_frame = ff_get_audio_buffer(ctx->outputs[0], frame->nb_samples);
254  if (!out_frame) {
256  return AVERROR(ENOMEM);
257  }
258  av_frame_copy_props(out_frame, frame);
259  }
260 
261  s->echo_samples(s, s->delayptrs, frame->extended_data, out_frame->extended_data,
262  frame->nb_samples, inlink->ch_layout.nb_channels);
263 
264  s->next_pts = frame->pts + av_rescale_q(frame->nb_samples, (AVRational){1, inlink->sample_rate}, inlink->time_base);
265 
266  if (frame != out_frame)
268 
269  return ff_filter_frame(ctx->outputs[0], out_frame);
270 }
271 
272 static int request_frame(AVFilterLink *outlink)
273 {
274  AVFilterContext *ctx = outlink->src;
275  AudioEchoContext *s = ctx->priv;
276  int nb_samples = FFMIN(s->fade_out, 2048);
277  AVFrame *frame = ff_get_audio_buffer(outlink, nb_samples);
278 
279  if (!frame)
280  return AVERROR(ENOMEM);
281  s->fade_out -= nb_samples;
282 
283  av_samples_set_silence(frame->extended_data, 0,
284  frame->nb_samples,
285  outlink->ch_layout.nb_channels,
286  frame->format);
287 
288  s->echo_samples(s, s->delayptrs, frame->extended_data, frame->extended_data,
289  frame->nb_samples, outlink->ch_layout.nb_channels);
290 
291  frame->pts = s->next_pts;
292  if (s->next_pts != AV_NOPTS_VALUE)
293  s->next_pts += av_rescale_q(nb_samples, (AVRational){1, outlink->sample_rate}, outlink->time_base);
294 
295  return ff_filter_frame(outlink, frame);
296 }
297 
299 {
300  AVFilterLink *inlink = ctx->inputs[0];
301  AVFilterLink *outlink = ctx->outputs[0];
302  AudioEchoContext *s = ctx->priv;
303  AVFrame *in;
304  int ret, status;
305  int64_t pts;
306 
308 
310  if (ret < 0)
311  return ret;
312  if (ret > 0)
313  return filter_frame(inlink, in);
314 
315  if (!s->eof && ff_inlink_acknowledge_status(inlink, &status, &pts)) {
316  if (status == AVERROR_EOF)
317  s->eof = 1;
318  }
319 
320  if (s->eof && s->fade_out <= 0) {
321  ff_outlink_set_status(outlink, AVERROR_EOF, s->next_pts);
322  return 0;
323  }
324 
325  if (!s->eof)
327 
328  return request_frame(outlink);
329 }
330 
331 static const AVFilterPad aecho_outputs[] = {
332  {
333  .name = "default",
334  .config_props = config_output,
335  .type = AVMEDIA_TYPE_AUDIO,
336  },
337 };
338 
340  .name = "aecho",
341  .description = NULL_IF_CONFIG_SMALL("Add echoing to the audio."),
342  .priv_size = sizeof(AudioEchoContext),
343  .priv_class = &aecho_class,
344  .init = init,
345  .activate = activate,
346  .uninit = uninit,
351 };
aecho_outputs
static const AVFilterPad aecho_outputs[]
Definition: af_aecho.c:331
AudioEchoContext::max_samples
int max_samples
Definition: af_aecho.c:38
ff_get_audio_buffer
AVFrame * ff_get_audio_buffer(AVFilterLink *link, int nb_samples)
Request an audio samples buffer with a specific set of permissions.
Definition: audio.c:98
AV_SAMPLE_FMT_FLTP
@ AV_SAMPLE_FMT_FLTP
float, planar
Definition: samplefmt.h:66
AV_LOG_WARNING
#define AV_LOG_WARNING
Something somehow does not look correct.
Definition: log.h:215
AVERROR
Filter the word “frame” indicates either a video frame or a group of audio as stored in an AVFrame structure Format for each input and each output the list of supported formats For video that means pixel format For audio that means channel sample they are references to shared objects When the negotiation mechanism computes the intersection of the formats supported at each end of a all references to both lists are replaced with a reference to the intersection And when a single format is eventually chosen for a link amongst the remaining all references to the list are updated That means that if a filter requires that its input and output have the same format amongst a supported all it has to do is use a reference to the same list of formats query_formats can leave some formats unset and return AVERROR(EAGAIN) to cause the negotiation mechanism toagain later. That can be used by filters with complex requirements to use the format negotiated on one link to set the formats supported on another. Frame references ownership and permissions
AVFILTER_DEFINE_CLASS
AVFILTER_DEFINE_CLASS(aecho)
opt.h
ff_filter_frame
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:1061
AVERROR_EOF
#define AVERROR_EOF
End of file.
Definition: error.h:57
int64_t
long long int64_t
Definition: coverity.c:34
inlink
The exact code depends on how similar the blocks are and how related they are to the and needs to apply these operations to the correct inlink or outlink if there are several Macros are available to factor that when no extra processing is inlink
Definition: filter_design.txt:212
av_frame_free
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:162
FILTER_INPUTS
#define FILTER_INPUTS(array)
Definition: filters.h:262
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:389
av_samples_set_silence
int av_samples_set_silence(uint8_t *const *audio_data, int offset, int nb_samples, int nb_channels, enum AVSampleFormat sample_fmt)
Fill an audio buffer with silence.
Definition: samplefmt.c:246
AVOption
AVOption.
Definition: opt.h:429
OFFSET
#define OFFSET(x)
Definition: af_aecho.c:48
AV_SAMPLE_FMT_S32P
@ AV_SAMPLE_FMT_S32P
signed 32 bits, planar
Definition: samplefmt.h:65
AudioEchoContext::nb_echoes
int nb_echoes
Definition: af_aecho.c:35
AudioEchoContext::out_gain
float out_gain
Definition: af_aecho.c:32
FFMAX
#define FFMAX(a, b)
Definition: macros.h:47
AVFilter::name
const char * name
Filter name.
Definition: avfilter.h:205
AudioEchoContext::delayptrs
uint8_t ** delayptrs
Definition: af_aecho.c:37
AVChannelLayout::nb_channels
int nb_channels
Number of channels in this layout.
Definition: channel_layout.h:327
AudioEchoContext::eof
int eof
Definition: af_aecho.c:40
ff_af_aecho
const AVFilter ff_af_aecho
Definition: af_aecho.c:339
FF_FILTER_FORWARD_STATUS_BACK
#define FF_FILTER_FORWARD_STATUS_BACK(outlink, inlink)
Forward the status on an output link to an input link.
Definition: filters.h:434
ECHO
#define ECHO(name, type, min, max)
Definition: af_aecho.c:157
ff_inlink_consume_frame
int ff_inlink_consume_frame(AVFilterLink *link, AVFrame **rframe)
Take a frame from the link's FIFO and update the link's stats.
Definition: avfilter.c:1490
AudioEchoContext::decays
char * decays
Definition: af_aecho.c:33
samplefmt.h
init
static av_cold int init(AVFilterContext *ctx)
Definition: af_aecho.c:102
pts
static int64_t pts
Definition: transcode_aac.c:644
config_output
static int config_output(AVFilterLink *outlink)
Definition: af_aecho.c:203
AVFilterPad
A filter pad used for either input or output.
Definition: filters.h:38
avassert.h
AV_LOG_ERROR
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:209
av_cold
#define av_cold
Definition: attributes.h:90
FILTER_SAMPLEFMTS
#define FILTER_SAMPLEFMTS(...)
Definition: filters.h:250
ff_outlink_set_status
static void ff_outlink_set_status(AVFilterLink *link, int status, int64_t pts)
Set the status field of a link from the source filter.
Definition: filters.h:424
A
#define A
Definition: af_aecho.c:49
uninit
static av_cold void uninit(AVFilterContext *ctx)
Definition: af_aecho.c:89
s
#define s(width, name)
Definition: cbs_vp9.c:198
AVMEDIA_TYPE_AUDIO
@ AVMEDIA_TYPE_AUDIO
Definition: avutil.h:202
av_strtok
char * av_strtok(char *s, const char *delim, char **saveptr)
Split the string into several tokens which can be accessed by successive calls to av_strtok().
Definition: avstring.c:178
AudioEchoContext::echo_samples
void(* echo_samples)(struct AudioEchoContext *ctx, uint8_t **delayptrs, uint8_t *const *src, uint8_t **dst, int nb_samples, int channels)
Definition: af_aecho.c:43
filters.h
AV_LOG_DEBUG
#define AV_LOG_DEBUG
Stuff which is only useful for libav* developers.
Definition: log.h:230
ctx
AVFormatContext * ctx
Definition: movenc.c:49
channels
channels
Definition: aptx.h:31
av_rescale_q
int64_t av_rescale_q(int64_t a, AVRational bq, AVRational cq)
Rescale a 64-bit integer by 2 rational numbers.
Definition: mathematics.c:142
FILTER_OUTPUTS
#define FILTER_OUTPUTS(array)
Definition: filters.h:263
av_realloc_f
#define av_realloc_f(p, o, n)
Definition: tableprint_vlc.h:32
av_sscanf
int av_sscanf(const char *string, const char *format,...)
See libc sscanf manual for more information.
Definition: avsscanf.c:961
AVClass
Describe the class of an AVClass context structure.
Definition: log.h:75
NULL
#define NULL
Definition: coverity.c:32
av_frame_copy_props
int av_frame_copy_props(AVFrame *dst, const AVFrame *src)
Copy only "metadata" fields from src to dst.
Definition: frame.c:725
AVRational
Rational number (pair of numerator and denominator).
Definition: rational.h:58
ff_audio_default_filterpad
const AVFilterPad ff_audio_default_filterpad[1]
An AVFilterPad array whose only entry has name "default" and is of type AVMEDIA_TYPE_AUDIO.
Definition: audio.c:34
ff_inlink_acknowledge_status
int ff_inlink_acknowledge_status(AVFilterLink *link, int *rstatus, int64_t *rpts)
Test and acknowledge the change of status on the link.
Definition: avfilter.c:1437
fill_items
static void fill_items(char *item_str, int *nb_items, float *items)
Definition: af_aecho.c:73
aecho_options
static const AVOption aecho_options[]
Definition: af_aecho.c:51
AudioEchoContext
Definition: af_aecho.c:30
NULL_IF_CONFIG_SMALL
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
Definition: internal.h:94
dst
uint8_t ptrdiff_t const uint8_t ptrdiff_t int intptr_t intptr_t int int16_t * dst
Definition: dsp.h:83
AudioEchoContext::samples
int * samples
Definition: af_aecho.c:39
AV_NOPTS_VALUE
#define AV_NOPTS_VALUE
Undefined timestamp value.
Definition: avutil.h:248
av_frame_is_writable
int av_frame_is_writable(AVFrame *frame)
Check if the frame data is writable.
Definition: frame.c:661
AudioEchoContext::delays
char * delays
Definition: af_aecho.c:33
FF_FILTER_FORWARD_WANTED
FF_FILTER_FORWARD_WANTED(outlink, inlink)
AudioEchoContext::fade_out
int fade_out
Definition: af_aecho.c:38
AV_SAMPLE_FMT_S16P
@ AV_SAMPLE_FMT_S16P
signed 16 bits, planar
Definition: samplefmt.h:64
AV_OPT_TYPE_FLOAT
@ AV_OPT_TYPE_FLOAT
Underlying C type is float.
Definition: opt.h:271
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:256
AudioEchoContext::next_pts
int64_t next_pts
Definition: af_aecho.c:41
AVFrame::extended_data
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:450
FFMIN
#define FFMIN(a, b)
Definition: macros.h:49
AVFilterPad::name
const char * name
Pad name.
Definition: filters.h:44
AVFilter
Filter definition.
Definition: avfilter.h:201
ret
ret
Definition: filter_design.txt:187
frame
these buffered frames must be flushed immediately if a new input produces new the filter must not call request_frame to get more It must just process the frame or queue it The task of requesting more frames is left to the filter s request_frame method or the application If a filter has several the filter must be ready for frames arriving randomly on any input any filter with several inputs will most likely require some kind of queuing mechanism It is perfectly acceptable to have a limited queue and to drop frames when the inputs are too unbalanced request_frame For filters that do not use the this method is called when a frame is wanted on an output For a it should directly call filter_frame on the corresponding output For a if there are queued frames already one of these frames should be pushed If the filter should request a frame on one of its repeatedly until at least one frame has been pushed Return or at least make progress towards producing a frame
Definition: filter_design.txt:264
request_frame
static int request_frame(AVFilterLink *outlink)
Definition: af_aecho.c:272
count_items
static void count_items(char *item_str, int *nb_items)
Definition: af_aecho.c:61
status
ov_status_e status
Definition: dnn_backend_openvino.c:100
avfilter.h
filter_frame
static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
Definition: af_aecho.c:244
AV_SAMPLE_FMT_DBLP
@ AV_SAMPLE_FMT_DBLP
double, planar
Definition: samplefmt.h:67
AudioEchoContext::decay
float * decay
Definition: af_aecho.c:34
AVFilterContext
An instance of a filter.
Definition: avfilter.h:457
AudioEchoContext::delay
float * delay
Definition: af_aecho.c:34
activate
static int activate(AVFilterContext *ctx)
Definition: af_aecho.c:298
mem.h
audio.h
av_samples_alloc_array_and_samples
int av_samples_alloc_array_and_samples(uint8_t ***audio_data, int *linesize, int nb_channels, int nb_samples, enum AVSampleFormat sample_fmt, int align)
Allocate a data pointers array, samples buffer for nb_samples samples, and fill data pointers and lin...
Definition: samplefmt.c:207
AudioEchoContext::delay_index
int delay_index
Definition: af_aecho.c:36
av_freep
#define av_freep(p)
Definition: tableprint_vlc.h:34
int32_t
int32_t
Definition: audioconvert.c:56
av_log
#define av_log(a,...)
Definition: tableprint_vlc.h:27
avstring.h
AV_OPT_TYPE_STRING
@ AV_OPT_TYPE_STRING
Underlying C type is a uint8_t* that is either NULL or points to a C string allocated with the av_mal...
Definition: opt.h:276
AudioEchoContext::in_gain
float in_gain
Definition: af_aecho.c:32
src
#define src
Definition: vp8dsp.c:248