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 #include "internal.h"
30 
31 typedef struct AudioEchoContext {
32  const AVClass *class;
33  float in_gain, out_gain;
34  char *delays, *decays;
35  float *delay, *decay;
36  int nb_echoes;
38  uint8_t **delayptrs;
40  int *samples;
41  int eof;
42  int64_t next_pts;
43 
44  void (*echo_samples)(struct AudioEchoContext *ctx, uint8_t **delayptrs,
45  uint8_t * const *src, uint8_t **dst,
46  int nb_samples, int channels);
48 
49 #define OFFSET(x) offsetof(AudioEchoContext, x)
50 #define A AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
51 
52 static const AVOption aecho_options[] = {
53  { "in_gain", "set signal input gain", OFFSET(in_gain), AV_OPT_TYPE_FLOAT, {.dbl=0.6}, 0, 1, A },
54  { "out_gain", "set signal output gain", OFFSET(out_gain), AV_OPT_TYPE_FLOAT, {.dbl=0.3}, 0, 1, A },
55  { "delays", "set list of signal delays", OFFSET(delays), AV_OPT_TYPE_STRING, {.str="1000"}, 0, 0, A },
56  { "decays", "set list of signal decays", OFFSET(decays), AV_OPT_TYPE_STRING, {.str="0.5"}, 0, 0, A },
57  { NULL }
58 };
59 
61 
62 static void count_items(char *item_str, int *nb_items)
63 {
64  char *p;
65 
66  *nb_items = 1;
67  for (p = item_str; *p; p++) {
68  if (*p == '|')
69  (*nb_items)++;
70  }
71 
72 }
73 
74 static void fill_items(char *item_str, int *nb_items, float *items)
75 {
76  char *p, *saveptr = NULL;
77  int i, new_nb_items = 0;
78 
79  p = item_str;
80  for (i = 0; i < *nb_items; i++) {
81  char *tstr = av_strtok(p, "|", &saveptr);
82  p = NULL;
83  if (tstr)
84  new_nb_items += av_sscanf(tstr, "%f", &items[new_nb_items]) == 1;
85  }
86 
87  *nb_items = new_nb_items;
88 }
89 
91 {
92  AudioEchoContext *s = ctx->priv;
93 
94  av_freep(&s->delay);
95  av_freep(&s->decay);
96  av_freep(&s->samples);
97 
98  if (s->delayptrs)
99  av_freep(&s->delayptrs[0]);
100  av_freep(&s->delayptrs);
101 }
102 
104 {
105  AudioEchoContext *s = ctx->priv;
106  int nb_delays, nb_decays, i;
107 
108  if (!s->delays || !s->decays) {
109  av_log(ctx, AV_LOG_ERROR, "Missing delays and/or decays.\n");
110  return AVERROR(EINVAL);
111  }
112 
113  count_items(s->delays, &nb_delays);
114  count_items(s->decays, &nb_decays);
115 
116  s->delay = av_realloc_f(s->delay, nb_delays, sizeof(*s->delay));
117  s->decay = av_realloc_f(s->decay, nb_decays, sizeof(*s->decay));
118  if (!s->delay || !s->decay)
119  return AVERROR(ENOMEM);
120 
121  fill_items(s->delays, &nb_delays, s->delay);
122  fill_items(s->decays, &nb_decays, s->decay);
123 
124  if (nb_delays != nb_decays) {
125  av_log(ctx, AV_LOG_ERROR, "Number of delays %d differs from number of decays %d.\n", nb_delays, nb_decays);
126  return AVERROR(EINVAL);
127  }
128 
129  s->nb_echoes = nb_delays;
130  if (!s->nb_echoes) {
131  av_log(ctx, AV_LOG_ERROR, "At least one decay & delay must be set.\n");
132  return AVERROR(EINVAL);
133  }
134 
135  s->samples = av_realloc_f(s->samples, nb_delays, sizeof(*s->samples));
136  if (!s->samples)
137  return AVERROR(ENOMEM);
138 
139  for (i = 0; i < nb_delays; i++) {
140  if (s->delay[i] <= 0 || s->delay[i] > 90000) {
141  av_log(ctx, AV_LOG_ERROR, "delay[%d]: %f is out of allowed range: (0, 90000]\n", i, s->delay[i]);
142  return AVERROR(EINVAL);
143  }
144  if (s->decay[i] <= 0 || s->decay[i] > 1) {
145  av_log(ctx, AV_LOG_ERROR, "decay[%d]: %f is out of allowed range: (0, 1]\n", i, s->decay[i]);
146  return AVERROR(EINVAL);
147  }
148  }
149 
150  s->next_pts = AV_NOPTS_VALUE;
151 
152  av_log(ctx, AV_LOG_DEBUG, "nb_echoes:%d\n", s->nb_echoes);
153  return 0;
154 }
155 
156 #define MOD(a, b) (((a) >= (b)) ? (a) - (b) : (a))
157 
158 #define ECHO(name, type, min, max) \
159 static void echo_samples_## name ##p(AudioEchoContext *ctx, \
160  uint8_t **delayptrs, \
161  uint8_t * const *src, uint8_t **dst, \
162  int nb_samples, int channels) \
163 { \
164  const double out_gain = ctx->out_gain; \
165  const double in_gain = ctx->in_gain; \
166  const int nb_echoes = ctx->nb_echoes; \
167  const int max_samples = ctx->max_samples; \
168  int i, j, chan, av_uninit(index); \
169  \
170  av_assert1(channels > 0); /* would corrupt delay_index */ \
171  \
172  for (chan = 0; chan < channels; chan++) { \
173  const type *s = (type *)src[chan]; \
174  type *d = (type *)dst[chan]; \
175  type *dbuf = (type *)delayptrs[chan]; \
176  \
177  index = ctx->delay_index; \
178  for (i = 0; i < nb_samples; i++, s++, d++) { \
179  double out, in; \
180  \
181  in = *s; \
182  out = in * in_gain; \
183  for (j = 0; j < nb_echoes; j++) { \
184  int ix = index + max_samples - ctx->samples[j]; \
185  ix = MOD(ix, max_samples); \
186  out += dbuf[ix] * ctx->decay[j]; \
187  } \
188  out *= out_gain; \
189  \
190  *d = av_clipd(out, min, max); \
191  dbuf[index] = in; \
192  \
193  index = MOD(index + 1, max_samples); \
194  } \
195  } \
196  ctx->delay_index = index; \
197 }
198 
199 ECHO(dbl, double, -1.0, 1.0 )
200 ECHO(flt, float, -1.0, 1.0 )
201 ECHO(s16, int16_t, INT16_MIN, INT16_MAX)
202 ECHO(s32, int32_t, INT32_MIN, INT32_MAX)
203 
204 static int config_output(AVFilterLink *outlink)
205 {
206  AVFilterContext *ctx = outlink->src;
207  AudioEchoContext *s = ctx->priv;
208  float volume = 1.0;
209  int i;
210 
211  for (i = 0; i < s->nb_echoes; i++) {
212  s->samples[i] = s->delay[i] * outlink->sample_rate / 1000.0;
213  s->max_samples = FFMAX(s->max_samples, s->samples[i]);
214  volume += s->decay[i];
215  }
216 
217  if (s->max_samples <= 0) {
218  av_log(ctx, AV_LOG_ERROR, "Nothing to echo - missing delay samples.\n");
219  return AVERROR(EINVAL);
220  }
221  s->fade_out = s->max_samples;
222 
223  if (volume * s->in_gain * s->out_gain > 1.0)
225  "out_gain %f can cause saturation of output\n", s->out_gain);
226 
227  switch (outlink->format) {
228  case AV_SAMPLE_FMT_DBLP: s->echo_samples = echo_samples_dblp; break;
229  case AV_SAMPLE_FMT_FLTP: s->echo_samples = echo_samples_fltp; break;
230  case AV_SAMPLE_FMT_S16P: s->echo_samples = echo_samples_s16p; break;
231  case AV_SAMPLE_FMT_S32P: s->echo_samples = echo_samples_s32p; break;
232  }
233 
234 
235  if (s->delayptrs)
236  av_freep(&s->delayptrs[0]);
237  av_freep(&s->delayptrs);
238 
239  return av_samples_alloc_array_and_samples(&s->delayptrs, NULL,
240  outlink->ch_layout.nb_channels,
241  s->max_samples,
242  outlink->format, 0);
243 }
244 
246 {
247  AVFilterContext *ctx = inlink->dst;
248  AudioEchoContext *s = ctx->priv;
249  AVFrame *out_frame;
250 
252  out_frame = frame;
253  } else {
254  out_frame = ff_get_audio_buffer(ctx->outputs[0], frame->nb_samples);
255  if (!out_frame) {
257  return AVERROR(ENOMEM);
258  }
259  av_frame_copy_props(out_frame, frame);
260  }
261 
262  s->echo_samples(s, s->delayptrs, frame->extended_data, out_frame->extended_data,
263  frame->nb_samples, inlink->ch_layout.nb_channels);
264 
265  s->next_pts = frame->pts + av_rescale_q(frame->nb_samples, (AVRational){1, inlink->sample_rate}, inlink->time_base);
266 
267  if (frame != out_frame)
269 
270  return ff_filter_frame(ctx->outputs[0], out_frame);
271 }
272 
273 static int request_frame(AVFilterLink *outlink)
274 {
275  AVFilterContext *ctx = outlink->src;
276  AudioEchoContext *s = ctx->priv;
277  int nb_samples = FFMIN(s->fade_out, 2048);
278  AVFrame *frame = ff_get_audio_buffer(outlink, nb_samples);
279 
280  if (!frame)
281  return AVERROR(ENOMEM);
282  s->fade_out -= nb_samples;
283 
284  av_samples_set_silence(frame->extended_data, 0,
285  frame->nb_samples,
286  outlink->ch_layout.nb_channels,
287  frame->format);
288 
289  s->echo_samples(s, s->delayptrs, frame->extended_data, frame->extended_data,
290  frame->nb_samples, outlink->ch_layout.nb_channels);
291 
292  frame->pts = s->next_pts;
293  if (s->next_pts != AV_NOPTS_VALUE)
294  s->next_pts += av_rescale_q(nb_samples, (AVRational){1, outlink->sample_rate}, outlink->time_base);
295 
296  return ff_filter_frame(outlink, frame);
297 }
298 
300 {
301  AVFilterLink *inlink = ctx->inputs[0];
302  AVFilterLink *outlink = ctx->outputs[0];
303  AudioEchoContext *s = ctx->priv;
304  AVFrame *in;
305  int ret, status;
306  int64_t pts;
307 
309 
311  if (ret < 0)
312  return ret;
313  if (ret > 0)
314  return filter_frame(inlink, in);
315 
316  if (!s->eof && ff_inlink_acknowledge_status(inlink, &status, &pts)) {
317  if (status == AVERROR_EOF)
318  s->eof = 1;
319  }
320 
321  if (s->eof && s->fade_out <= 0) {
322  ff_outlink_set_status(outlink, AVERROR_EOF, s->next_pts);
323  return 0;
324  }
325 
326  if (!s->eof)
328 
329  return request_frame(outlink);
330 }
331 
332 static const AVFilterPad aecho_outputs[] = {
333  {
334  .name = "default",
335  .config_props = config_output,
336  .type = AVMEDIA_TYPE_AUDIO,
337  },
338 };
339 
341  .name = "aecho",
342  .description = NULL_IF_CONFIG_SMALL("Add echoing to the audio."),
343  .priv_size = sizeof(AudioEchoContext),
344  .priv_class = &aecho_class,
345  .init = init,
346  .activate = activate,
347  .uninit = uninit,
352 };
aecho_outputs
static const AVFilterPad aecho_outputs[]
Definition: af_aecho.c:332
AudioEchoContext::max_samples
int max_samples
Definition: af_aecho.c:39
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:97
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:186
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:1015
AVERROR_EOF
#define AVERROR_EOF
End of file.
Definition: error.h:57
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:160
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:374
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:357
OFFSET
#define OFFSET(x)
Definition: af_aecho.c:49
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:36
AudioEchoContext::out_gain
float out_gain
Definition: af_aecho.c:33
FFMAX
#define FFMAX(a, b)
Definition: macros.h:47
AVFilter::name
const char * name
Filter name.
Definition: avfilter.h:170
AudioEchoContext::delayptrs
uint8_t ** delayptrs
Definition: af_aecho.c:38
AVChannelLayout::nb_channels
int nb_channels
Number of channels in this layout.
Definition: channel_layout.h:313
AudioEchoContext::eof
int eof
Definition: af_aecho.c:41
ff_af_aecho
const AVFilter ff_af_aecho
Definition: af_aecho.c:340
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:199
ECHO
#define ECHO(name, type, min, max)
Definition: af_aecho.c:158
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:1442
AudioEchoContext::decays
char * decays
Definition: af_aecho.c:34
samplefmt.h
init
static av_cold int init(AVFilterContext *ctx)
Definition: af_aecho.c:103
pts
static int64_t pts
Definition: transcode_aac.c:644
config_output
static int config_output(AVFilterLink *outlink)
Definition: af_aecho.c:204
AVFilterPad
A filter pad used for either input or output.
Definition: internal.h:33
avassert.h
AV_LOG_ERROR
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:180
av_cold
#define av_cold
Definition: attributes.h:90
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:189
A
#define A
Definition: af_aecho.c:50
uninit
static av_cold void uninit(AVFilterContext *ctx)
Definition: af_aecho.c:90
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:44
filters.h
AV_LOG_DEBUG
#define AV_LOG_DEBUG
Stuff which is only useful for libav* developers.
Definition: log.h:201
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_INPUTS
#define FILTER_INPUTS(array)
Definition: internal.h:182
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:962
AVClass
Describe the class of an AVClass context structure.
Definition: log.h:66
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:709
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:33
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:1389
fill_items
static void fill_items(char *item_str, int *nb_items, float *items)
Definition: af_aecho.c:74
aecho_options
static const AVOption aecho_options[]
Definition: af_aecho.c:52
AudioEchoContext
Definition: af_aecho.c:31
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
AudioEchoContext::samples
int * samples
Definition: af_aecho.c:40
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:645
AudioEchoContext::delays
char * delays
Definition: af_aecho.c:34
FF_FILTER_FORWARD_WANTED
FF_FILTER_FORWARD_WANTED(outlink, inlink)
AudioEchoContext::fade_out
int fade_out
Definition: af_aecho.c:39
AV_SAMPLE_FMT_S16P
@ AV_SAMPLE_FMT_S16P
signed 16 bits, planar
Definition: samplefmt.h:64
internal.h
AV_OPT_TYPE_FLOAT
@ AV_OPT_TYPE_FLOAT
Definition: opt.h:248
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:42
AVFrame::extended_data
uint8_t ** extended_data
pointers to the data planes/channels.
Definition: frame.h:435
FFMIN
#define FFMIN(a, b)
Definition: macros.h:49
AVFilterPad::name
const char * name
Pad name.
Definition: internal.h:39
AVFilter
Filter definition.
Definition: avfilter.h:166
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:273
count_items
static void count_items(char *item_str, int *nb_items)
Definition: af_aecho.c:62
status
ov_status_e status
Definition: dnn_backend_openvino.c:101
avfilter.h
filter_frame
static int filter_frame(AVFilterLink *inlink, AVFrame *frame)
Definition: af_aecho.c:245
AV_SAMPLE_FMT_DBLP
@ AV_SAMPLE_FMT_DBLP
double, planar
Definition: samplefmt.h:67
AudioEchoContext::decay
float * decay
Definition: af_aecho.c:35
AVFilterContext
An instance of a filter.
Definition: avfilter.h:407
AudioEchoContext::delay
float * delay
Definition: af_aecho.c:35
activate
static int activate(AVFilterContext *ctx)
Definition: af_aecho.c:299
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:37
FILTER_OUTPUTS
#define FILTER_OUTPUTS(array)
Definition: internal.h:183
av_freep
#define av_freep(p)
Definition: tableprint_vlc.h:34
src
INIT_CLIP pixel * src
Definition: h264pred_template.c:418
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
Definition: opt.h:249
AudioEchoContext::in_gain
float in_gain
Definition: af_aecho.c:33
FILTER_SAMPLEFMTS
#define FILTER_SAMPLEFMTS(...)
Definition: internal.h:170