FFmpeg  4.4.4
af_aiir.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2018 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 <float.h>
22 
23 #include "libavutil/avassert.h"
24 #include "libavutil/avstring.h"
25 #include "libavutil/intreadwrite.h"
26 #include "libavutil/opt.h"
28 #include "audio.h"
29 #include "avfilter.h"
30 #include "internal.h"
31 
32 typedef struct ThreadData {
33  AVFrame *in, *out;
34 } ThreadData;
35 
36 typedef struct Pair {
37  int a, b;
38 } Pair;
39 
40 typedef struct BiquadContext {
41  double a[3];
42  double b[3];
43  double w1, w2;
45 
46 typedef struct IIRChannel {
47  int nb_ab[2];
48  double *ab[2];
49  double g;
50  double *cache[2];
51  double fir;
53  int clippings;
54 } IIRChannel;
55 
56 typedef struct AudioIIRContext {
57  const AVClass *class;
58  char *a_str, *b_str, *g_str;
59  double dry_gain, wet_gain;
60  double mix;
61  int normalize;
62  int format;
63  int process;
64  int precision;
65  int response;
66  int w, h;
69 
71 
73  int channels;
75 
76  int (*iir_channel)(AVFilterContext *ctx, void *arg, int ch, int nb_jobs);
78 
80 {
81  AudioIIRContext *s = ctx->priv;
84  enum AVSampleFormat sample_fmts[] = {
87  };
88  static const enum AVPixelFormat pix_fmts[] = {
91  };
92  int ret;
93 
94  if (s->response) {
95  AVFilterLink *videolink = ctx->outputs[1];
96 
98  if ((ret = ff_formats_ref(formats, &videolink->incfg.formats)) < 0)
99  return ret;
100  }
101 
103  if (!layouts)
104  return AVERROR(ENOMEM);
106  if (ret < 0)
107  return ret;
108 
109  sample_fmts[0] = s->sample_format;
111  if (!formats)
112  return AVERROR(ENOMEM);
114  if (ret < 0)
115  return ret;
116 
118  if (!formats)
119  return AVERROR(ENOMEM);
121 }
122 
123 #define IIR_CH(name, type, min, max, need_clipping) \
124 static int iir_ch_## name(AVFilterContext *ctx, void *arg, int ch, int nb_jobs) \
125 { \
126  AudioIIRContext *s = ctx->priv; \
127  const double ig = s->dry_gain; \
128  const double og = s->wet_gain; \
129  const double mix = s->mix; \
130  ThreadData *td = arg; \
131  AVFrame *in = td->in, *out = td->out; \
132  const type *src = (const type *)in->extended_data[ch]; \
133  double *oc = (double *)s->iir[ch].cache[0]; \
134  double *ic = (double *)s->iir[ch].cache[1]; \
135  const int nb_a = s->iir[ch].nb_ab[0]; \
136  const int nb_b = s->iir[ch].nb_ab[1]; \
137  const double *a = s->iir[ch].ab[0]; \
138  const double *b = s->iir[ch].ab[1]; \
139  const double g = s->iir[ch].g; \
140  int *clippings = &s->iir[ch].clippings; \
141  type *dst = (type *)out->extended_data[ch]; \
142  int n; \
143  \
144  for (n = 0; n < in->nb_samples; n++) { \
145  double sample = 0.; \
146  int x; \
147  \
148  memmove(&ic[1], &ic[0], (nb_b - 1) * sizeof(*ic)); \
149  memmove(&oc[1], &oc[0], (nb_a - 1) * sizeof(*oc)); \
150  ic[0] = src[n] * ig; \
151  for (x = 0; x < nb_b; x++) \
152  sample += b[x] * ic[x]; \
153  \
154  for (x = 1; x < nb_a; x++) \
155  sample -= a[x] * oc[x]; \
156  \
157  oc[0] = sample; \
158  sample *= og * g; \
159  sample = sample * mix + ic[0] * (1. - mix); \
160  if (need_clipping && sample < min) { \
161  (*clippings)++; \
162  dst[n] = min; \
163  } else if (need_clipping && sample > max) { \
164  (*clippings)++; \
165  dst[n] = max; \
166  } else { \
167  dst[n] = sample; \
168  } \
169  } \
170  \
171  return 0; \
172 }
173 
174 IIR_CH(s16p, int16_t, INT16_MIN, INT16_MAX, 1)
175 IIR_CH(s32p, int32_t, INT32_MIN, INT32_MAX, 1)
176 IIR_CH(fltp, float, -1., 1., 0)
177 IIR_CH(dblp, double, -1., 1., 0)
178 
179 #define SERIAL_IIR_CH(name, type, min, max, need_clipping) \
180 static int iir_ch_serial_## name(AVFilterContext *ctx, void *arg, \
181  int ch, int nb_jobs) \
182 { \
183  AudioIIRContext *s = ctx->priv; \
184  const double ig = s->dry_gain; \
185  const double og = s->wet_gain; \
186  const double mix = s->mix; \
187  const double imix = 1. - mix; \
188  ThreadData *td = arg; \
189  AVFrame *in = td->in, *out = td->out; \
190  const type *src = (const type *)in->extended_data[ch]; \
191  type *dst = (type *)out->extended_data[ch]; \
192  IIRChannel *iir = &s->iir[ch]; \
193  const double g = iir->g; \
194  int *clippings = &iir->clippings; \
195  int nb_biquads = (FFMAX(iir->nb_ab[0], iir->nb_ab[1]) + 1) / 2; \
196  int n, i; \
197  \
198  for (i = nb_biquads - 1; i >= 0; i--) { \
199  const double a1 = -iir->biquads[i].a[1]; \
200  const double a2 = -iir->biquads[i].a[2]; \
201  const double b0 = iir->biquads[i].b[0]; \
202  const double b1 = iir->biquads[i].b[1]; \
203  const double b2 = iir->biquads[i].b[2]; \
204  double w1 = iir->biquads[i].w1; \
205  double w2 = iir->biquads[i].w2; \
206  \
207  for (n = 0; n < in->nb_samples; n++) { \
208  double i0 = ig * (i ? dst[n] : src[n]); \
209  double o0 = i0 * b0 + w1; \
210  \
211  w1 = b1 * i0 + w2 + a1 * o0; \
212  w2 = b2 * i0 + a2 * o0; \
213  o0 *= og * g; \
214  \
215  o0 = o0 * mix + imix * i0; \
216  if (need_clipping && o0 < min) { \
217  (*clippings)++; \
218  dst[n] = min; \
219  } else if (need_clipping && o0 > max) { \
220  (*clippings)++; \
221  dst[n] = max; \
222  } else { \
223  dst[n] = o0; \
224  } \
225  } \
226  iir->biquads[i].w1 = w1; \
227  iir->biquads[i].w2 = w2; \
228  } \
229  \
230  return 0; \
231 }
232 
233 SERIAL_IIR_CH(s16p, int16_t, INT16_MIN, INT16_MAX, 1)
234 SERIAL_IIR_CH(s32p, int32_t, INT32_MIN, INT32_MAX, 1)
235 SERIAL_IIR_CH(fltp, float, -1., 1., 0)
236 SERIAL_IIR_CH(dblp, double, -1., 1., 0)
237 
238 #define PARALLEL_IIR_CH(name, type, min, max, need_clipping) \
239 static int iir_ch_parallel_## name(AVFilterContext *ctx, void *arg, \
240  int ch, int nb_jobs) \
241 { \
242  AudioIIRContext *s = ctx->priv; \
243  const double ig = s->dry_gain; \
244  const double og = s->wet_gain; \
245  const double mix = s->mix; \
246  const double imix = 1. - mix; \
247  ThreadData *td = arg; \
248  AVFrame *in = td->in, *out = td->out; \
249  const type *src = (const type *)in->extended_data[ch]; \
250  type *dst = (type *)out->extended_data[ch]; \
251  IIRChannel *iir = &s->iir[ch]; \
252  const double g = iir->g; \
253  const double fir = iir->fir; \
254  int *clippings = &iir->clippings; \
255  int nb_biquads = (FFMAX(iir->nb_ab[0], iir->nb_ab[1]) + 1) / 2; \
256  int n, i; \
257  \
258  for (i = 0; i < nb_biquads; i++) { \
259  const double a1 = -iir->biquads[i].a[1]; \
260  const double a2 = -iir->biquads[i].a[2]; \
261  const double b1 = iir->biquads[i].b[1]; \
262  const double b2 = iir->biquads[i].b[2]; \
263  double w1 = iir->biquads[i].w1; \
264  double w2 = iir->biquads[i].w2; \
265  \
266  for (n = 0; n < in->nb_samples; n++) { \
267  double i0 = ig * src[n]; \
268  double o0 = w1; \
269  \
270  w1 = b1 * i0 + w2 + a1 * o0; \
271  w2 = b2 * i0 + a2 * o0; \
272  o0 *= og * g; \
273  o0 += dst[n]; \
274  \
275  if (need_clipping && o0 < min) { \
276  (*clippings)++; \
277  dst[n] = min; \
278  } else if (need_clipping && o0 > max) { \
279  (*clippings)++; \
280  dst[n] = max; \
281  } else { \
282  dst[n] = o0; \
283  } \
284  } \
285  iir->biquads[i].w1 = w1; \
286  iir->biquads[i].w2 = w2; \
287  } \
288  \
289  for (n = 0; n < in->nb_samples; n++) { \
290  dst[n] += fir * src[n]; \
291  dst[n] = dst[n] * mix + imix * src[n]; \
292  } \
293  \
294  return 0; \
295 }
296 
297 PARALLEL_IIR_CH(s16p, int16_t, INT16_MIN, INT16_MAX, 1)
298 PARALLEL_IIR_CH(s32p, int32_t, INT32_MIN, INT32_MAX, 1)
299 PARALLEL_IIR_CH(fltp, float, -1., 1., 0)
300 PARALLEL_IIR_CH(dblp, double, -1., 1., 0)
301 
302 #define LATTICE_IIR_CH(name, type, min, max, need_clipping) \
303 static int iir_ch_lattice_## name(AVFilterContext *ctx, void *arg, \
304  int ch, int nb_jobs) \
305 { \
306  AudioIIRContext *s = ctx->priv; \
307  const double ig = s->dry_gain; \
308  const double og = s->wet_gain; \
309  const double mix = s->mix; \
310  ThreadData *td = arg; \
311  AVFrame *in = td->in, *out = td->out; \
312  const type *src = (const type *)in->extended_data[ch]; \
313  double n0, n1, p0, *x = (double *)s->iir[ch].cache[0]; \
314  const int nb_stages = s->iir[ch].nb_ab[1]; \
315  const double *v = s->iir[ch].ab[0]; \
316  const double *k = s->iir[ch].ab[1]; \
317  const double g = s->iir[ch].g; \
318  int *clippings = &s->iir[ch].clippings; \
319  type *dst = (type *)out->extended_data[ch]; \
320  int n; \
321  \
322  for (n = 0; n < in->nb_samples; n++) { \
323  const double in = src[n] * ig; \
324  double out = 0.; \
325  \
326  n1 = in; \
327  for (int i = nb_stages - 1; i >= 0; i--) { \
328  n0 = n1 - k[i] * x[i]; \
329  p0 = n0 * k[i] + x[i]; \
330  out += p0 * v[i+1]; \
331  x[i] = p0; \
332  n1 = n0; \
333  } \
334  \
335  out += n1 * v[0]; \
336  memmove(&x[1], &x[0], nb_stages * sizeof(*x)); \
337  x[0] = n1; \
338  out *= og * g; \
339  out = out * mix + in * (1. - mix); \
340  if (need_clipping && out < min) { \
341  (*clippings)++; \
342  dst[n] = min; \
343  } else if (need_clipping && out > max) { \
344  (*clippings)++; \
345  dst[n] = max; \
346  } else { \
347  dst[n] = out; \
348  } \
349  } \
350  \
351  return 0; \
352 }
353 
354 LATTICE_IIR_CH(s16p, int16_t, INT16_MIN, INT16_MAX, 1)
355 LATTICE_IIR_CH(s32p, int32_t, INT32_MIN, INT32_MAX, 1)
356 LATTICE_IIR_CH(fltp, float, -1., 1., 0)
357 LATTICE_IIR_CH(dblp, double, -1., 1., 0)
358 
359 static void count_coefficients(char *item_str, int *nb_items)
360 {
361  char *p;
362 
363  if (!item_str)
364  return;
365 
366  *nb_items = 1;
367  for (p = item_str; *p && *p != '|'; p++) {
368  if (*p == ' ')
369  (*nb_items)++;
370  }
371 }
372 
373 static int read_gains(AVFilterContext *ctx, char *item_str, int nb_items)
374 {
375  AudioIIRContext *s = ctx->priv;
376  char *p, *arg, *old_str, *prev_arg = NULL, *saveptr = NULL;
377  int i;
378 
379  p = old_str = av_strdup(item_str);
380  if (!p)
381  return AVERROR(ENOMEM);
382  for (i = 0; i < nb_items; i++) {
383  if (!(arg = av_strtok(p, "|", &saveptr)))
384  arg = prev_arg;
385 
386  if (!arg) {
387  av_freep(&old_str);
388  return AVERROR(EINVAL);
389  }
390 
391  p = NULL;
392  if (av_sscanf(arg, "%lf", &s->iir[i].g) != 1) {
393  av_log(ctx, AV_LOG_ERROR, "Invalid gains supplied: %s\n", arg);
394  av_freep(&old_str);
395  return AVERROR(EINVAL);
396  }
397 
398  prev_arg = arg;
399  }
400 
401  av_freep(&old_str);
402 
403  return 0;
404 }
405 
406 static int read_tf_coefficients(AVFilterContext *ctx, char *item_str, int nb_items, double *dst)
407 {
408  char *p, *arg, *old_str, *saveptr = NULL;
409  int i;
410 
411  p = old_str = av_strdup(item_str);
412  if (!p)
413  return AVERROR(ENOMEM);
414  for (i = 0; i < nb_items; i++) {
415  if (!(arg = av_strtok(p, " ", &saveptr)))
416  break;
417 
418  p = NULL;
419  if (av_sscanf(arg, "%lf", &dst[i]) != 1) {
420  av_log(ctx, AV_LOG_ERROR, "Invalid coefficients supplied: %s\n", arg);
421  av_freep(&old_str);
422  return AVERROR(EINVAL);
423  }
424  }
425 
426  av_freep(&old_str);
427 
428  return 0;
429 }
430 
431 static int read_zp_coefficients(AVFilterContext *ctx, char *item_str, int nb_items, double *dst, const char *format)
432 {
433  char *p, *arg, *old_str, *saveptr = NULL;
434  int i;
435 
436  p = old_str = av_strdup(item_str);
437  if (!p)
438  return AVERROR(ENOMEM);
439  for (i = 0; i < nb_items; i++) {
440  if (!(arg = av_strtok(p, " ", &saveptr)))
441  break;
442 
443  p = NULL;
444  if (av_sscanf(arg, format, &dst[i*2], &dst[i*2+1]) != 2) {
445  av_log(ctx, AV_LOG_ERROR, "Invalid coefficients supplied: %s\n", arg);
446  av_freep(&old_str);
447  return AVERROR(EINVAL);
448  }
449  }
450 
451  av_freep(&old_str);
452 
453  return 0;
454 }
455 
456 static const char *const format[] = { "%lf", "%lf %lfi", "%lf %lfr", "%lf %lfd", "%lf %lfi" };
457 
458 static int read_channels(AVFilterContext *ctx, int channels, uint8_t *item_str, int ab)
459 {
460  AudioIIRContext *s = ctx->priv;
461  char *p, *arg, *old_str, *prev_arg = NULL, *saveptr = NULL;
462  int i, ret;
463 
464  p = old_str = av_strdup(item_str);
465  if (!p)
466  return AVERROR(ENOMEM);
467  for (i = 0; i < channels; i++) {
468  IIRChannel *iir = &s->iir[i];
469 
470  if (!(arg = av_strtok(p, "|", &saveptr)))
471  arg = prev_arg;
472 
473  if (!arg) {
474  av_freep(&old_str);
475  return AVERROR(EINVAL);
476  }
477 
478  count_coefficients(arg, &iir->nb_ab[ab]);
479 
480  p = NULL;
481  iir->cache[ab] = av_calloc(iir->nb_ab[ab] + 1, sizeof(double));
482  iir->ab[ab] = av_calloc(iir->nb_ab[ab] * (!!s->format + 1), sizeof(double));
483  if (!iir->ab[ab] || !iir->cache[ab]) {
484  av_freep(&old_str);
485  return AVERROR(ENOMEM);
486  }
487 
488  if (s->format > 0) {
489  ret = read_zp_coefficients(ctx, arg, iir->nb_ab[ab], iir->ab[ab], format[s->format]);
490  } else {
491  ret = read_tf_coefficients(ctx, arg, iir->nb_ab[ab], iir->ab[ab]);
492  }
493  if (ret < 0) {
494  av_freep(&old_str);
495  return ret;
496  }
497  prev_arg = arg;
498  }
499 
500  av_freep(&old_str);
501 
502  return 0;
503 }
504 
505 static void cmul(double re, double im, double re2, double im2, double *RE, double *IM)
506 {
507  *RE = re * re2 - im * im2;
508  *IM = re * im2 + re2 * im;
509 }
510 
511 static int expand(AVFilterContext *ctx, double *pz, int n, double *coefs)
512 {
513  coefs[2 * n] = 1.0;
514 
515  for (int i = 1; i <= n; i++) {
516  for (int j = n - i; j < n; j++) {
517  double re, im;
518 
519  cmul(coefs[2 * (j + 1)], coefs[2 * (j + 1) + 1],
520  pz[2 * (i - 1)], pz[2 * (i - 1) + 1], &re, &im);
521 
522  coefs[2 * j] -= re;
523  coefs[2 * j + 1] -= im;
524  }
525  }
526 
527  for (int i = 0; i < n + 1; i++) {
528  if (fabs(coefs[2 * i + 1]) > FLT_EPSILON) {
529  av_log(ctx, AV_LOG_ERROR, "coefs: %f of z^%d is not real; poles/zeros are not complex conjugates.\n",
530  coefs[2 * i + 1], i);
531  return AVERROR(EINVAL);
532  }
533  }
534 
535  return 0;
536 }
537 
538 static void normalize_coeffs(AVFilterContext *ctx, int ch)
539 {
540  AudioIIRContext *s = ctx->priv;
541  IIRChannel *iir = &s->iir[ch];
542  double sum_den = 0.;
543 
544  if (!s->normalize)
545  return;
546 
547  for (int i = 0; i < iir->nb_ab[1]; i++) {
548  sum_den += iir->ab[1][i];
549  }
550 
551  if (sum_den > 1e-6) {
552  double factor, sum_num = 0.;
553 
554  for (int i = 0; i < iir->nb_ab[0]; i++) {
555  sum_num += iir->ab[0][i];
556  }
557 
558  factor = sum_num / sum_den;
559 
560  for (int i = 0; i < iir->nb_ab[1]; i++) {
561  iir->ab[1][i] *= factor;
562  }
563  }
564 }
565 
567 {
568  AudioIIRContext *s = ctx->priv;
569  int ch, i, j, ret = 0;
570 
571  for (ch = 0; ch < channels; ch++) {
572  IIRChannel *iir = &s->iir[ch];
573  double *topc, *botc;
574 
575  topc = av_calloc((iir->nb_ab[1] + 1) * 2, sizeof(*topc));
576  botc = av_calloc((iir->nb_ab[0] + 1) * 2, sizeof(*botc));
577  if (!topc || !botc) {
578  ret = AVERROR(ENOMEM);
579  goto fail;
580  }
581 
582  ret = expand(ctx, iir->ab[0], iir->nb_ab[0], botc);
583  if (ret < 0) {
584  goto fail;
585  }
586 
587  ret = expand(ctx, iir->ab[1], iir->nb_ab[1], topc);
588  if (ret < 0) {
589  goto fail;
590  }
591 
592  for (j = 0, i = iir->nb_ab[1]; i >= 0; j++, i--) {
593  iir->ab[1][j] = topc[2 * i];
594  }
595  iir->nb_ab[1]++;
596 
597  for (j = 0, i = iir->nb_ab[0]; i >= 0; j++, i--) {
598  iir->ab[0][j] = botc[2 * i];
599  }
600  iir->nb_ab[0]++;
601 
602  normalize_coeffs(ctx, ch);
603 
604 fail:
605  av_free(topc);
606  av_free(botc);
607  if (ret < 0)
608  break;
609  }
610 
611  return ret;
612 }
613 
615 {
616  AudioIIRContext *s = ctx->priv;
617  int ch, ret;
618 
619  for (ch = 0; ch < channels; ch++) {
620  IIRChannel *iir = &s->iir[ch];
621  int nb_biquads = (FFMAX(iir->nb_ab[0], iir->nb_ab[1]) + 1) / 2;
622  int current_biquad = 0;
623 
624  iir->biquads = av_calloc(nb_biquads, sizeof(BiquadContext));
625  if (!iir->biquads)
626  return AVERROR(ENOMEM);
627 
628  while (nb_biquads--) {
629  Pair outmost_pole = { -1, -1 };
630  Pair nearest_zero = { -1, -1 };
631  double zeros[4] = { 0 };
632  double poles[4] = { 0 };
633  double b[6] = { 0 };
634  double a[6] = { 0 };
635  double min_distance = DBL_MAX;
636  double max_mag = 0;
637  double factor;
638  int i;
639 
640  for (i = 0; i < iir->nb_ab[0]; i++) {
641  double mag;
642 
643  if (isnan(iir->ab[0][2 * i]) || isnan(iir->ab[0][2 * i + 1]))
644  continue;
645  mag = hypot(iir->ab[0][2 * i], iir->ab[0][2 * i + 1]);
646 
647  if (mag > max_mag) {
648  max_mag = mag;
649  outmost_pole.a = i;
650  }
651  }
652 
653  for (i = 0; i < iir->nb_ab[0]; i++) {
654  if (isnan(iir->ab[0][2 * i]) || isnan(iir->ab[0][2 * i + 1]))
655  continue;
656 
657  if (iir->ab[0][2 * i ] == iir->ab[0][2 * outmost_pole.a ] &&
658  iir->ab[0][2 * i + 1] == -iir->ab[0][2 * outmost_pole.a + 1]) {
659  outmost_pole.b = i;
660  break;
661  }
662  }
663 
664  av_log(ctx, AV_LOG_VERBOSE, "outmost_pole is %d.%d\n", outmost_pole.a, outmost_pole.b);
665 
666  if (outmost_pole.a < 0 || outmost_pole.b < 0)
667  return AVERROR(EINVAL);
668 
669  for (i = 0; i < iir->nb_ab[1]; i++) {
670  double distance;
671 
672  if (isnan(iir->ab[1][2 * i]) || isnan(iir->ab[1][2 * i + 1]))
673  continue;
674  distance = hypot(iir->ab[0][2 * outmost_pole.a ] - iir->ab[1][2 * i ],
675  iir->ab[0][2 * outmost_pole.a + 1] - iir->ab[1][2 * i + 1]);
676 
677  if (distance < min_distance) {
678  min_distance = distance;
679  nearest_zero.a = i;
680  }
681  }
682 
683  for (i = 0; i < iir->nb_ab[1]; i++) {
684  if (isnan(iir->ab[1][2 * i]) || isnan(iir->ab[1][2 * i + 1]))
685  continue;
686 
687  if (iir->ab[1][2 * i ] == iir->ab[1][2 * nearest_zero.a ] &&
688  iir->ab[1][2 * i + 1] == -iir->ab[1][2 * nearest_zero.a + 1]) {
689  nearest_zero.b = i;
690  break;
691  }
692  }
693 
694  av_log(ctx, AV_LOG_VERBOSE, "nearest_zero is %d.%d\n", nearest_zero.a, nearest_zero.b);
695 
696  if (nearest_zero.a < 0 || nearest_zero.b < 0)
697  return AVERROR(EINVAL);
698 
699  poles[0] = iir->ab[0][2 * outmost_pole.a ];
700  poles[1] = iir->ab[0][2 * outmost_pole.a + 1];
701 
702  zeros[0] = iir->ab[1][2 * nearest_zero.a ];
703  zeros[1] = iir->ab[1][2 * nearest_zero.a + 1];
704 
705  if (nearest_zero.a == nearest_zero.b && outmost_pole.a == outmost_pole.b) {
706  zeros[2] = 0;
707  zeros[3] = 0;
708 
709  poles[2] = 0;
710  poles[3] = 0;
711  } else {
712  poles[2] = iir->ab[0][2 * outmost_pole.b ];
713  poles[3] = iir->ab[0][2 * outmost_pole.b + 1];
714 
715  zeros[2] = iir->ab[1][2 * nearest_zero.b ];
716  zeros[3] = iir->ab[1][2 * nearest_zero.b + 1];
717  }
718 
719  ret = expand(ctx, zeros, 2, b);
720  if (ret < 0)
721  return ret;
722 
723  ret = expand(ctx, poles, 2, a);
724  if (ret < 0)
725  return ret;
726 
727  iir->ab[0][2 * outmost_pole.a] = iir->ab[0][2 * outmost_pole.a + 1] = NAN;
728  iir->ab[0][2 * outmost_pole.b] = iir->ab[0][2 * outmost_pole.b + 1] = NAN;
729  iir->ab[1][2 * nearest_zero.a] = iir->ab[1][2 * nearest_zero.a + 1] = NAN;
730  iir->ab[1][2 * nearest_zero.b] = iir->ab[1][2 * nearest_zero.b + 1] = NAN;
731 
732  iir->biquads[current_biquad].a[0] = 1.;
733  iir->biquads[current_biquad].a[1] = a[2] / a[4];
734  iir->biquads[current_biquad].a[2] = a[0] / a[4];
735  iir->biquads[current_biquad].b[0] = b[4] / a[4];
736  iir->biquads[current_biquad].b[1] = b[2] / a[4];
737  iir->biquads[current_biquad].b[2] = b[0] / a[4];
738 
739  if (s->normalize &&
740  fabs(iir->biquads[current_biquad].b[0] +
741  iir->biquads[current_biquad].b[1] +
742  iir->biquads[current_biquad].b[2]) > 1e-6) {
743  factor = (iir->biquads[current_biquad].a[0] +
744  iir->biquads[current_biquad].a[1] +
745  iir->biquads[current_biquad].a[2]) /
746  (iir->biquads[current_biquad].b[0] +
747  iir->biquads[current_biquad].b[1] +
748  iir->biquads[current_biquad].b[2]);
749 
750  av_log(ctx, AV_LOG_VERBOSE, "factor=%f\n", factor);
751 
752  iir->biquads[current_biquad].b[0] *= factor;
753  iir->biquads[current_biquad].b[1] *= factor;
754  iir->biquads[current_biquad].b[2] *= factor;
755  }
756 
757  iir->biquads[current_biquad].b[0] *= (current_biquad ? 1.0 : iir->g);
758  iir->biquads[current_biquad].b[1] *= (current_biquad ? 1.0 : iir->g);
759  iir->biquads[current_biquad].b[2] *= (current_biquad ? 1.0 : iir->g);
760 
761  av_log(ctx, AV_LOG_VERBOSE, "a=%f %f %f:b=%f %f %f\n",
762  iir->biquads[current_biquad].a[0],
763  iir->biquads[current_biquad].a[1],
764  iir->biquads[current_biquad].a[2],
765  iir->biquads[current_biquad].b[0],
766  iir->biquads[current_biquad].b[1],
767  iir->biquads[current_biquad].b[2]);
768 
769  current_biquad++;
770  }
771  }
772 
773  return 0;
774 }
775 
776 static void biquad_process(double *x, double *y, int length,
777  double b0, double b1, double b2,
778  double a1, double a2)
779 {
780  double w1 = 0., w2 = 0.;
781 
782  a1 = -a1;
783  a2 = -a2;
784 
785  for (int n = 0; n < length; n++) {
786  double out, in = x[n];
787 
788  y[n] = out = in * b0 + w1;
789  w1 = b1 * in + w2 + a1 * out;
790  w2 = b2 * in + a2 * out;
791  }
792 }
793 
794 static void solve(double *matrix, double *vector, int n, double *y, double *x, double *lu)
795 {
796  double sum = 0.;
797 
798  for (int i = 0; i < n; i++) {
799  for (int j = i; j < n; j++) {
800  sum = 0.;
801  for (int k = 0; k < i; k++)
802  sum += lu[i * n + k] * lu[k * n + j];
803  lu[i * n + j] = matrix[j * n + i] - sum;
804  }
805  for (int j = i + 1; j < n; j++) {
806  sum = 0.;
807  for (int k = 0; k < i; k++)
808  sum += lu[j * n + k] * lu[k * n + i];
809  lu[j * n + i] = (1. / lu[i * n + i]) * (matrix[i * n + j] - sum);
810  }
811  }
812 
813  for (int i = 0; i < n; i++) {
814  sum = 0.;
815  for (int k = 0; k < i; k++)
816  sum += lu[i * n + k] * y[k];
817  y[i] = vector[i] - sum;
818  }
819 
820  for (int i = n - 1; i >= 0; i--) {
821  sum = 0.;
822  for (int k = i + 1; k < n; k++)
823  sum += lu[i * n + k] * x[k];
824  x[i] = (1 / lu[i * n + i]) * (y[i] - sum);
825  }
826 }
827 
829 {
830  AudioIIRContext *s = ctx->priv;
831  int ret = 0;
832 
833  for (int ch = 0; ch < channels; ch++) {
834  IIRChannel *iir = &s->iir[ch];
835  int nb_biquads = (FFMAX(iir->nb_ab[0], iir->nb_ab[1]) + 1) / 2;
836  int length = nb_biquads * 2 + 1;
837  double *impulse = av_calloc(length, sizeof(*impulse));
838  double *y = av_calloc(length, sizeof(*y));
839  double *resp = av_calloc(length, sizeof(*resp));
840  double *M = av_calloc((length - 1) * 2 * nb_biquads, sizeof(*M));
841  double *W = av_calloc((length - 1) * 2 * nb_biquads, sizeof(*W));
842 
843  if (!impulse || !y || !resp || !M) {
844  av_free(impulse);
845  av_free(y);
846  av_free(resp);
847  av_free(M);
848  av_free(W);
849  return AVERROR(ENOMEM);
850  }
851 
852  impulse[0] = 1.;
853 
854  for (int n = 0; n < nb_biquads; n++) {
855  BiquadContext *biquad = &iir->biquads[n];
856 
857  biquad_process(n ? y : impulse, y, length,
858  biquad->b[0], biquad->b[1], biquad->b[2],
859  biquad->a[1], biquad->a[2]);
860  }
861 
862  for (int n = 0; n < nb_biquads; n++) {
863  BiquadContext *biquad = &iir->biquads[n];
864 
865  biquad_process(impulse, resp, length - 1,
866  1., 0., 0., biquad->a[1], biquad->a[2]);
867 
868  memcpy(M + n * 2 * (length - 1), resp, sizeof(*resp) * (length - 1));
869  memcpy(M + n * 2 * (length - 1) + length, resp, sizeof(*resp) * (length - 2));
870  memset(resp, 0, length * sizeof(*resp));
871  }
872 
873  solve(M, &y[1], length - 1, &impulse[1], resp, W);
874 
875  iir->fir = y[0];
876 
877  for (int n = 0; n < nb_biquads; n++) {
878  BiquadContext *biquad = &iir->biquads[n];
879 
880  biquad->b[0] = 0.;
881  biquad->b[1] = resp[n * 2 + 0];
882  biquad->b[2] = resp[n * 2 + 1];
883  }
884 
885  av_free(impulse);
886  av_free(y);
887  av_free(resp);
888  av_free(M);
889  av_free(W);
890 
891  if (ret < 0)
892  return ret;
893  }
894 
895  return 0;
896 }
897 
899 {
900  AudioIIRContext *s = ctx->priv;
901  int ch;
902 
903  for (ch = 0; ch < channels; ch++) {
904  IIRChannel *iir = &s->iir[ch];
905  int n;
906 
907  for (n = 0; n < iir->nb_ab[0]; n++) {
908  double r = iir->ab[0][2*n];
909  double angle = iir->ab[0][2*n+1];
910 
911  iir->ab[0][2*n] = r * cos(angle);
912  iir->ab[0][2*n+1] = r * sin(angle);
913  }
914 
915  for (n = 0; n < iir->nb_ab[1]; n++) {
916  double r = iir->ab[1][2*n];
917  double angle = iir->ab[1][2*n+1];
918 
919  iir->ab[1][2*n] = r * cos(angle);
920  iir->ab[1][2*n+1] = r * sin(angle);
921  }
922  }
923 }
924 
926 {
927  AudioIIRContext *s = ctx->priv;
928  int ch;
929 
930  for (ch = 0; ch < channels; ch++) {
931  IIRChannel *iir = &s->iir[ch];
932  int n;
933 
934  for (n = 0; n < iir->nb_ab[0]; n++) {
935  double sr = iir->ab[0][2*n];
936  double si = iir->ab[0][2*n+1];
937 
938  iir->ab[0][2*n] = exp(sr) * cos(si);
939  iir->ab[0][2*n+1] = exp(sr) * sin(si);
940  }
941 
942  for (n = 0; n < iir->nb_ab[1]; n++) {
943  double sr = iir->ab[1][2*n];
944  double si = iir->ab[1][2*n+1];
945 
946  iir->ab[1][2*n] = exp(sr) * cos(si);
947  iir->ab[1][2*n+1] = exp(sr) * sin(si);
948  }
949  }
950 }
951 
952 static double fact(double i)
953 {
954  if (i <= 0.)
955  return 1.;
956  return i * fact(i - 1.);
957 }
958 
959 static double coef_sf2zf(double *a, int N, int n)
960 {
961  double z = 0.;
962 
963  for (int i = 0; i <= N; i++) {
964  double acc = 0.;
965 
966  for (int k = FFMAX(n - N + i, 0); k <= FFMIN(i, n); k++) {
967  acc += ((fact(i) * fact(N - i)) /
968  (fact(k) * fact(i - k) * fact(n - k) * fact(N - i - n + k))) *
969  ((k & 1) ? -1. : 1.);
970  }
971 
972  z += a[i] * pow(2., i) * acc;
973  }
974 
975  return z;
976 }
977 
979 {
980  AudioIIRContext *s = ctx->priv;
981  int ch;
982 
983  for (ch = 0; ch < channels; ch++) {
984  IIRChannel *iir = &s->iir[ch];
985  double *temp0 = av_calloc(iir->nb_ab[0], sizeof(*temp0));
986  double *temp1 = av_calloc(iir->nb_ab[1], sizeof(*temp1));
987 
988  if (!temp0 || !temp1)
989  goto next;
990 
991  memcpy(temp0, iir->ab[0], iir->nb_ab[0] * sizeof(*temp0));
992  memcpy(temp1, iir->ab[1], iir->nb_ab[1] * sizeof(*temp1));
993 
994  for (int n = 0; n < iir->nb_ab[0]; n++)
995  iir->ab[0][n] = coef_sf2zf(temp0, iir->nb_ab[0] - 1, n);
996 
997  for (int n = 0; n < iir->nb_ab[1]; n++)
998  iir->ab[1][n] = coef_sf2zf(temp1, iir->nb_ab[1] - 1, n);
999 
1000 next:
1001  av_free(temp0);
1002  av_free(temp1);
1003  }
1004 }
1005 
1007 {
1008  AudioIIRContext *s = ctx->priv;
1009  int ch;
1010 
1011  for (ch = 0; ch < channels; ch++) {
1012  IIRChannel *iir = &s->iir[ch];
1013  int n;
1014 
1015  for (n = 0; n < iir->nb_ab[0]; n++) {
1016  double r = iir->ab[0][2*n];
1017  double angle = M_PI*iir->ab[0][2*n+1]/180.;
1018 
1019  iir->ab[0][2*n] = r * cos(angle);
1020  iir->ab[0][2*n+1] = r * sin(angle);
1021  }
1022 
1023  for (n = 0; n < iir->nb_ab[1]; n++) {
1024  double r = iir->ab[1][2*n];
1025  double angle = M_PI*iir->ab[1][2*n+1]/180.;
1026 
1027  iir->ab[1][2*n] = r * cos(angle);
1028  iir->ab[1][2*n+1] = r * sin(angle);
1029  }
1030  }
1031 }
1032 
1034 {
1035  AudioIIRContext *s = ctx->priv;
1036  int ch;
1037 
1038  for (ch = 0; ch < channels; ch++) {
1039  IIRChannel *iir = &s->iir[ch];
1040 
1041  for (int n = 0; n < iir->nb_ab[0]; n++) {
1042  double pr = hypot(iir->ab[0][2*n], iir->ab[0][2*n+1]);
1043 
1044  if (pr >= 1.) {
1045  av_log(ctx, AV_LOG_WARNING, "pole %d at channel %d is unstable\n", n, ch);
1046  break;
1047  }
1048  }
1049  }
1050 }
1051 
1052 static void drawtext(AVFrame *pic, int x, int y, const char *txt, uint32_t color)
1053 {
1054  const uint8_t *font;
1055  int font_height;
1056  int i;
1057 
1058  font = avpriv_cga_font, font_height = 8;
1059 
1060  for (i = 0; txt[i]; i++) {
1061  int char_y, mask;
1062 
1063  uint8_t *p = pic->data[0] + y * pic->linesize[0] + (x + i * 8) * 4;
1064  for (char_y = 0; char_y < font_height; char_y++) {
1065  for (mask = 0x80; mask; mask >>= 1) {
1066  if (font[txt[i] * font_height + char_y] & mask)
1067  AV_WL32(p, color);
1068  p += 4;
1069  }
1070  p += pic->linesize[0] - 8 * 4;
1071  }
1072  }
1073 }
1074 
1075 static void draw_line(AVFrame *out, int x0, int y0, int x1, int y1, uint32_t color)
1076 {
1077  int dx = FFABS(x1-x0);
1078  int dy = FFABS(y1-y0), sy = y0 < y1 ? 1 : -1;
1079  int err = (dx>dy ? dx : -dy) / 2, e2;
1080 
1081  for (;;) {
1082  AV_WL32(out->data[0] + y0 * out->linesize[0] + x0 * 4, color);
1083 
1084  if (x0 == x1 && y0 == y1)
1085  break;
1086 
1087  e2 = err;
1088 
1089  if (e2 >-dx) {
1090  err -= dy;
1091  x0--;
1092  }
1093 
1094  if (e2 < dy) {
1095  err += dx;
1096  y0 += sy;
1097  }
1098  }
1099 }
1100 
1101 static double distance(double x0, double x1, double y0, double y1)
1102 {
1103  return hypot(x0 - x1, y0 - y1);
1104 }
1105 
1106 static void get_response(int channel, int format, double w,
1107  const double *b, const double *a,
1108  int nb_b, int nb_a, double *magnitude, double *phase)
1109 {
1110  double realz, realp;
1111  double imagz, imagp;
1112  double real, imag;
1113  double div;
1114 
1115  if (format == 0) {
1116  realz = 0., realp = 0.;
1117  imagz = 0., imagp = 0.;
1118  for (int x = 0; x < nb_a; x++) {
1119  realz += cos(-x * w) * a[x];
1120  imagz += sin(-x * w) * a[x];
1121  }
1122 
1123  for (int x = 0; x < nb_b; x++) {
1124  realp += cos(-x * w) * b[x];
1125  imagp += sin(-x * w) * b[x];
1126  }
1127 
1128  div = realp * realp + imagp * imagp;
1129  real = (realz * realp + imagz * imagp) / div;
1130  imag = (imagz * realp - imagp * realz) / div;
1131 
1132  *magnitude = hypot(real, imag);
1133  *phase = atan2(imag, real);
1134  } else {
1135  double p = 1., z = 1.;
1136  double acc = 0.;
1137 
1138  for (int x = 0; x < nb_a; x++) {
1139  z *= distance(cos(w), a[2 * x], sin(w), a[2 * x + 1]);
1140  acc += atan2(sin(w) - a[2 * x + 1], cos(w) - a[2 * x]);
1141  }
1142 
1143  for (int x = 0; x < nb_b; x++) {
1144  p *= distance(cos(w), b[2 * x], sin(w), b[2 * x + 1]);
1145  acc -= atan2(sin(w) - b[2 * x + 1], cos(w) - b[2 * x]);
1146  }
1147 
1148  *magnitude = z / p;
1149  *phase = acc;
1150  }
1151 }
1152 
1154 {
1155  AudioIIRContext *s = ctx->priv;
1156  double *mag, *phase, *temp, *delay, min = DBL_MAX, max = -DBL_MAX;
1157  double min_delay = DBL_MAX, max_delay = -DBL_MAX, min_phase, max_phase;
1158  int prev_ymag = -1, prev_yphase = -1, prev_ydelay = -1;
1159  char text[32];
1160  int ch, i;
1161 
1162  memset(out->data[0], 0, s->h * out->linesize[0]);
1163 
1164  phase = av_malloc_array(s->w, sizeof(*phase));
1165  temp = av_malloc_array(s->w, sizeof(*temp));
1166  mag = av_malloc_array(s->w, sizeof(*mag));
1167  delay = av_malloc_array(s->w, sizeof(*delay));
1168  if (!mag || !phase || !delay || !temp)
1169  goto end;
1170 
1171  ch = av_clip(s->ir_channel, 0, s->channels - 1);
1172  for (i = 0; i < s->w; i++) {
1173  const double *b = s->iir[ch].ab[0];
1174  const double *a = s->iir[ch].ab[1];
1175  const int nb_b = s->iir[ch].nb_ab[0];
1176  const int nb_a = s->iir[ch].nb_ab[1];
1177  double w = i * M_PI / (s->w - 1);
1178  double m, p;
1179 
1180  get_response(ch, s->format, w, b, a, nb_b, nb_a, &m, &p);
1181 
1182  mag[i] = s->iir[ch].g * m;
1183  phase[i] = p;
1184  min = fmin(min, mag[i]);
1185  max = fmax(max, mag[i]);
1186  }
1187 
1188  temp[0] = 0.;
1189  for (i = 0; i < s->w - 1; i++) {
1190  double d = phase[i] - phase[i + 1];
1191  temp[i + 1] = ceil(fabs(d) / (2. * M_PI)) * 2. * M_PI * ((d > M_PI) - (d < -M_PI));
1192  }
1193 
1194  min_phase = phase[0];
1195  max_phase = phase[0];
1196  for (i = 1; i < s->w; i++) {
1197  temp[i] += temp[i - 1];
1198  phase[i] += temp[i];
1199  min_phase = fmin(min_phase, phase[i]);
1200  max_phase = fmax(max_phase, phase[i]);
1201  }
1202 
1203  for (i = 0; i < s->w - 1; i++) {
1204  double div = s->w / (double)sample_rate;
1205 
1206  delay[i + 1] = -(phase[i] - phase[i + 1]) / div;
1207  min_delay = fmin(min_delay, delay[i + 1]);
1208  max_delay = fmax(max_delay, delay[i + 1]);
1209  }
1210  delay[0] = delay[1];
1211 
1212  for (i = 0; i < s->w; i++) {
1213  int ymag = mag[i] / max * (s->h - 1);
1214  int ydelay = (delay[i] - min_delay) / (max_delay - min_delay) * (s->h - 1);
1215  int yphase = (phase[i] - min_phase) / (max_phase - min_phase) * (s->h - 1);
1216 
1217  ymag = s->h - 1 - av_clip(ymag, 0, s->h - 1);
1218  yphase = s->h - 1 - av_clip(yphase, 0, s->h - 1);
1219  ydelay = s->h - 1 - av_clip(ydelay, 0, s->h - 1);
1220 
1221  if (prev_ymag < 0)
1222  prev_ymag = ymag;
1223  if (prev_yphase < 0)
1224  prev_yphase = yphase;
1225  if (prev_ydelay < 0)
1226  prev_ydelay = ydelay;
1227 
1228  draw_line(out, i, ymag, FFMAX(i - 1, 0), prev_ymag, 0xFFFF00FF);
1229  draw_line(out, i, yphase, FFMAX(i - 1, 0), prev_yphase, 0xFF00FF00);
1230  draw_line(out, i, ydelay, FFMAX(i - 1, 0), prev_ydelay, 0xFF00FFFF);
1231 
1232  prev_ymag = ymag;
1233  prev_yphase = yphase;
1234  prev_ydelay = ydelay;
1235  }
1236 
1237  if (s->w > 400 && s->h > 100) {
1238  drawtext(out, 2, 2, "Max Magnitude:", 0xDDDDDDDD);
1239  snprintf(text, sizeof(text), "%.2f", max);
1240  drawtext(out, 15 * 8 + 2, 2, text, 0xDDDDDDDD);
1241 
1242  drawtext(out, 2, 12, "Min Magnitude:", 0xDDDDDDDD);
1243  snprintf(text, sizeof(text), "%.2f", min);
1244  drawtext(out, 15 * 8 + 2, 12, text, 0xDDDDDDDD);
1245 
1246  drawtext(out, 2, 22, "Max Phase:", 0xDDDDDDDD);
1247  snprintf(text, sizeof(text), "%.2f", max_phase);
1248  drawtext(out, 15 * 8 + 2, 22, text, 0xDDDDDDDD);
1249 
1250  drawtext(out, 2, 32, "Min Phase:", 0xDDDDDDDD);
1251  snprintf(text, sizeof(text), "%.2f", min_phase);
1252  drawtext(out, 15 * 8 + 2, 32, text, 0xDDDDDDDD);
1253 
1254  drawtext(out, 2, 42, "Max Delay:", 0xDDDDDDDD);
1255  snprintf(text, sizeof(text), "%.2f", max_delay);
1256  drawtext(out, 11 * 8 + 2, 42, text, 0xDDDDDDDD);
1257 
1258  drawtext(out, 2, 52, "Min Delay:", 0xDDDDDDDD);
1259  snprintf(text, sizeof(text), "%.2f", min_delay);
1260  drawtext(out, 11 * 8 + 2, 52, text, 0xDDDDDDDD);
1261  }
1262 
1263 end:
1264  av_free(delay);
1265  av_free(temp);
1266  av_free(phase);
1267  av_free(mag);
1268 }
1269 
1270 static int config_output(AVFilterLink *outlink)
1271 {
1272  AVFilterContext *ctx = outlink->src;
1273  AudioIIRContext *s = ctx->priv;
1274  AVFilterLink *inlink = ctx->inputs[0];
1275  int ch, ret, i;
1276 
1277  s->channels = inlink->channels;
1278  s->iir = av_calloc(s->channels, sizeof(*s->iir));
1279  if (!s->iir)
1280  return AVERROR(ENOMEM);
1281 
1282  ret = read_gains(ctx, s->g_str, inlink->channels);
1283  if (ret < 0)
1284  return ret;
1285 
1286  ret = read_channels(ctx, inlink->channels, s->a_str, 0);
1287  if (ret < 0)
1288  return ret;
1289 
1290  ret = read_channels(ctx, inlink->channels, s->b_str, 1);
1291  if (ret < 0)
1292  return ret;
1293 
1294  if (s->format == -1) {
1295  convert_sf2tf(ctx, inlink->channels);
1296  s->format = 0;
1297  } else if (s->format == 2) {
1298  convert_pr2zp(ctx, inlink->channels);
1299  } else if (s->format == 3) {
1300  convert_pd2zp(ctx, inlink->channels);
1301  } else if (s->format == 4) {
1302  convert_sp2zp(ctx, inlink->channels);
1303  }
1304  if (s->format > 0) {
1305  check_stability(ctx, inlink->channels);
1306  }
1307 
1308  av_frame_free(&s->video);
1309  if (s->response) {
1310  s->video = ff_get_video_buffer(ctx->outputs[1], s->w, s->h);
1311  if (!s->video)
1312  return AVERROR(ENOMEM);
1313 
1314  draw_response(ctx, s->video, inlink->sample_rate);
1315  }
1316 
1317  if (s->format == 0)
1318  av_log(ctx, AV_LOG_WARNING, "transfer function coefficients format is not recommended for too high number of zeros/poles.\n");
1319 
1320  if (s->format > 0 && s->process == 0) {
1321  av_log(ctx, AV_LOG_WARNING, "Direct processsing is not recommended for zp coefficients format.\n");
1322 
1323  ret = convert_zp2tf(ctx, inlink->channels);
1324  if (ret < 0)
1325  return ret;
1326  } else if (s->format == -2 && s->process > 0) {
1327  av_log(ctx, AV_LOG_ERROR, "Only direct processing is implemented for lattice-ladder function.\n");
1328  return AVERROR_PATCHWELCOME;
1329  } else if (s->format <= 0 && s->process == 1) {
1330  av_log(ctx, AV_LOG_ERROR, "Serial processing is not implemented for transfer function.\n");
1331  return AVERROR_PATCHWELCOME;
1332  } else if (s->format <= 0 && s->process == 2) {
1333  av_log(ctx, AV_LOG_ERROR, "Parallel processing is not implemented for transfer function.\n");
1334  return AVERROR_PATCHWELCOME;
1335  } else if (s->format > 0 && s->process == 1) {
1336  ret = decompose_zp2biquads(ctx, inlink->channels);
1337  if (ret < 0)
1338  return ret;
1339  } else if (s->format > 0 && s->process == 2) {
1340  if (s->precision > 1)
1341  av_log(ctx, AV_LOG_WARNING, "Parallel processing is not recommended for fixed-point precisions.\n");
1342  ret = decompose_zp2biquads(ctx, inlink->channels);
1343  if (ret < 0)
1344  return ret;
1345  ret = convert_serial2parallel(ctx, inlink->channels);
1346  if (ret < 0)
1347  return ret;
1348  }
1349 
1350  for (ch = 0; s->format == -2 && ch < inlink->channels; ch++) {
1351  IIRChannel *iir = &s->iir[ch];
1352 
1353  if (iir->nb_ab[0] != iir->nb_ab[1] + 1) {
1354  av_log(ctx, AV_LOG_ERROR, "Number of ladder coefficients must be one more than number of reflection coefficients.\n");
1355  return AVERROR(EINVAL);
1356  }
1357  }
1358 
1359  for (ch = 0; s->format == 0 && ch < inlink->channels; ch++) {
1360  IIRChannel *iir = &s->iir[ch];
1361 
1362  for (i = 1; i < iir->nb_ab[0]; i++) {
1363  iir->ab[0][i] /= iir->ab[0][0];
1364  }
1365 
1366  iir->ab[0][0] = 1.0;
1367  for (i = 0; i < iir->nb_ab[1]; i++) {
1368  iir->ab[1][i] *= iir->g;
1369  }
1370 
1371  normalize_coeffs(ctx, ch);
1372  }
1373 
1374  switch (inlink->format) {
1375  case AV_SAMPLE_FMT_DBLP: s->iir_channel = s->process == 2 ? iir_ch_parallel_dblp : s->process == 1 ? iir_ch_serial_dblp : iir_ch_dblp; break;
1376  case AV_SAMPLE_FMT_FLTP: s->iir_channel = s->process == 2 ? iir_ch_parallel_fltp : s->process == 1 ? iir_ch_serial_fltp : iir_ch_fltp; break;
1377  case AV_SAMPLE_FMT_S32P: s->iir_channel = s->process == 2 ? iir_ch_parallel_s32p : s->process == 1 ? iir_ch_serial_s32p : iir_ch_s32p; break;
1378  case AV_SAMPLE_FMT_S16P: s->iir_channel = s->process == 2 ? iir_ch_parallel_s16p : s->process == 1 ? iir_ch_serial_s16p : iir_ch_s16p; break;
1379  }
1380 
1381  if (s->format == -2) {
1382  switch (inlink->format) {
1383  case AV_SAMPLE_FMT_DBLP: s->iir_channel = iir_ch_lattice_dblp; break;
1384  case AV_SAMPLE_FMT_FLTP: s->iir_channel = iir_ch_lattice_fltp; break;
1385  case AV_SAMPLE_FMT_S32P: s->iir_channel = iir_ch_lattice_s32p; break;
1386  case AV_SAMPLE_FMT_S16P: s->iir_channel = iir_ch_lattice_s16p; break;
1387  }
1388  }
1389 
1390  return 0;
1391 }
1392 
1393 static int filter_frame(AVFilterLink *inlink, AVFrame *in)
1394 {
1395  AVFilterContext *ctx = inlink->dst;
1396  AudioIIRContext *s = ctx->priv;
1397  AVFilterLink *outlink = ctx->outputs[0];
1398  ThreadData td;
1399  AVFrame *out;
1400  int ch, ret;
1401 
1402  if (av_frame_is_writable(in) && s->process != 2) {
1403  out = in;
1404  } else {
1405  out = ff_get_audio_buffer(outlink, in->nb_samples);
1406  if (!out) {
1407  av_frame_free(&in);
1408  return AVERROR(ENOMEM);
1409  }
1411  }
1412 
1413  td.in = in;
1414  td.out = out;
1415  ctx->internal->execute(ctx, s->iir_channel, &td, NULL, outlink->channels);
1416 
1417  for (ch = 0; ch < outlink->channels; ch++) {
1418  if (s->iir[ch].clippings > 0)
1419  av_log(ctx, AV_LOG_WARNING, "Channel %d clipping %d times. Please reduce gain.\n",
1420  ch, s->iir[ch].clippings);
1421  s->iir[ch].clippings = 0;
1422  }
1423 
1424  if (in != out)
1425  av_frame_free(&in);
1426 
1427  if (s->response) {
1428  AVFilterLink *outlink = ctx->outputs[1];
1429  int64_t old_pts = s->video->pts;
1430  int64_t new_pts = av_rescale_q(out->pts, ctx->inputs[0]->time_base, outlink->time_base);
1431 
1432  if (new_pts > old_pts) {
1433  AVFrame *clone;
1434 
1435  s->video->pts = new_pts;
1436  clone = av_frame_clone(s->video);
1437  if (!clone)
1438  return AVERROR(ENOMEM);
1439  ret = ff_filter_frame(outlink, clone);
1440  if (ret < 0)
1441  return ret;
1442  }
1443  }
1444 
1445  return ff_filter_frame(outlink, out);
1446 }
1447 
1448 static int config_video(AVFilterLink *outlink)
1449 {
1450  AVFilterContext *ctx = outlink->src;
1451  AudioIIRContext *s = ctx->priv;
1452 
1453  outlink->sample_aspect_ratio = (AVRational){1,1};
1454  outlink->w = s->w;
1455  outlink->h = s->h;
1456  outlink->frame_rate = s->rate;
1457  outlink->time_base = av_inv_q(outlink->frame_rate);
1458 
1459  return 0;
1460 }
1461 
1463 {
1464  AudioIIRContext *s = ctx->priv;
1465  AVFilterPad pad, vpad;
1466  int ret;
1467 
1468  if (!s->a_str || !s->b_str || !s->g_str) {
1469  av_log(ctx, AV_LOG_ERROR, "Valid coefficients are mandatory.\n");
1470  return AVERROR(EINVAL);
1471  }
1472 
1473  switch (s->precision) {
1474  case 0: s->sample_format = AV_SAMPLE_FMT_DBLP; break;
1475  case 1: s->sample_format = AV_SAMPLE_FMT_FLTP; break;
1476  case 2: s->sample_format = AV_SAMPLE_FMT_S32P; break;
1477  case 3: s->sample_format = AV_SAMPLE_FMT_S16P; break;
1478  default: return AVERROR_BUG;
1479  }
1480 
1481  pad = (AVFilterPad){
1482  .name = "default",
1483  .type = AVMEDIA_TYPE_AUDIO,
1484  .config_props = config_output,
1485  };
1486 
1487  ret = ff_insert_outpad(ctx, 0, &pad);
1488  if (ret < 0)
1489  return ret;
1490 
1491  if (s->response) {
1492  vpad = (AVFilterPad){
1493  .name = "filter_response",
1494  .type = AVMEDIA_TYPE_VIDEO,
1495  .config_props = config_video,
1496  };
1497 
1498  ret = ff_insert_outpad(ctx, 1, &vpad);
1499  if (ret < 0)
1500  return ret;
1501  }
1502 
1503  return 0;
1504 }
1505 
1507 {
1508  AudioIIRContext *s = ctx->priv;
1509  int ch;
1510 
1511  if (s->iir) {
1512  for (ch = 0; ch < s->channels; ch++) {
1513  IIRChannel *iir = &s->iir[ch];
1514  av_freep(&iir->ab[0]);
1515  av_freep(&iir->ab[1]);
1516  av_freep(&iir->cache[0]);
1517  av_freep(&iir->cache[1]);
1518  av_freep(&iir->biquads);
1519  }
1520  }
1521  av_freep(&s->iir);
1522 
1523  av_frame_free(&s->video);
1524 }
1525 
1526 static const AVFilterPad inputs[] = {
1527  {
1528  .name = "default",
1529  .type = AVMEDIA_TYPE_AUDIO,
1530  .filter_frame = filter_frame,
1531  },
1532  { NULL }
1533 };
1534 
1535 #define OFFSET(x) offsetof(AudioIIRContext, x)
1536 #define AF AV_OPT_FLAG_AUDIO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
1537 #define VF AV_OPT_FLAG_VIDEO_PARAM|AV_OPT_FLAG_FILTERING_PARAM
1538 
1539 static const AVOption aiir_options[] = {
1540  { "zeros", "set B/numerator/zeros/reflection coefficients", OFFSET(b_str), AV_OPT_TYPE_STRING, {.str="1+0i 1-0i"}, 0, 0, AF },
1541  { "z", "set B/numerator/zeros/reflection coefficients", OFFSET(b_str), AV_OPT_TYPE_STRING, {.str="1+0i 1-0i"}, 0, 0, AF },
1542  { "poles", "set A/denominator/poles/ladder coefficients", OFFSET(a_str), AV_OPT_TYPE_STRING, {.str="1+0i 1-0i"}, 0, 0, AF },
1543  { "p", "set A/denominator/poles/ladder coefficients", OFFSET(a_str), AV_OPT_TYPE_STRING, {.str="1+0i 1-0i"}, 0, 0, AF },
1544  { "gains", "set channels gains", OFFSET(g_str), AV_OPT_TYPE_STRING, {.str="1|1"}, 0, 0, AF },
1545  { "k", "set channels gains", OFFSET(g_str), AV_OPT_TYPE_STRING, {.str="1|1"}, 0, 0, AF },
1546  { "dry", "set dry gain", OFFSET(dry_gain), AV_OPT_TYPE_DOUBLE, {.dbl=1}, 0, 1, AF },
1547  { "wet", "set wet gain", OFFSET(wet_gain), AV_OPT_TYPE_DOUBLE, {.dbl=1}, 0, 1, AF },
1548  { "format", "set coefficients format", OFFSET(format), AV_OPT_TYPE_INT, {.i64=1}, -2, 4, AF, "format" },
1549  { "f", "set coefficients format", OFFSET(format), AV_OPT_TYPE_INT, {.i64=1}, -2, 4, AF, "format" },
1550  { "ll", "lattice-ladder function", 0, AV_OPT_TYPE_CONST, {.i64=-2}, 0, 0, AF, "format" },
1551  { "sf", "analog transfer function", 0, AV_OPT_TYPE_CONST, {.i64=-1}, 0, 0, AF, "format" },
1552  { "tf", "digital transfer function", 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, AF, "format" },
1553  { "zp", "Z-plane zeros/poles", 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, AF, "format" },
1554  { "pr", "Z-plane zeros/poles (polar radians)", 0, AV_OPT_TYPE_CONST, {.i64=2}, 0, 0, AF, "format" },
1555  { "pd", "Z-plane zeros/poles (polar degrees)", 0, AV_OPT_TYPE_CONST, {.i64=3}, 0, 0, AF, "format" },
1556  { "sp", "S-plane zeros/poles", 0, AV_OPT_TYPE_CONST, {.i64=4}, 0, 0, AF, "format" },
1557  { "process", "set kind of processing", OFFSET(process), AV_OPT_TYPE_INT, {.i64=1}, 0, 2, AF, "process" },
1558  { "r", "set kind of processing", OFFSET(process), AV_OPT_TYPE_INT, {.i64=1}, 0, 2, AF, "process" },
1559  { "d", "direct", 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, AF, "process" },
1560  { "s", "serial", 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, AF, "process" },
1561  { "p", "parallel", 0, AV_OPT_TYPE_CONST, {.i64=2}, 0, 0, AF, "process" },
1562  { "precision", "set filtering precision", OFFSET(precision),AV_OPT_TYPE_INT, {.i64=0}, 0, 3, AF, "precision" },
1563  { "e", "set precision", OFFSET(precision),AV_OPT_TYPE_INT, {.i64=0}, 0, 3, AF, "precision" },
1564  { "dbl", "double-precision floating-point", 0, AV_OPT_TYPE_CONST, {.i64=0}, 0, 0, AF, "precision" },
1565  { "flt", "single-precision floating-point", 0, AV_OPT_TYPE_CONST, {.i64=1}, 0, 0, AF, "precision" },
1566  { "i32", "32-bit integers", 0, AV_OPT_TYPE_CONST, {.i64=2}, 0, 0, AF, "precision" },
1567  { "i16", "16-bit integers", 0, AV_OPT_TYPE_CONST, {.i64=3}, 0, 0, AF, "precision" },
1568  { "normalize", "normalize coefficients", OFFSET(normalize),AV_OPT_TYPE_BOOL, {.i64=1}, 0, 1, AF },
1569  { "n", "normalize coefficients", OFFSET(normalize),AV_OPT_TYPE_BOOL, {.i64=1}, 0, 1, AF },
1570  { "mix", "set mix", OFFSET(mix), AV_OPT_TYPE_DOUBLE, {.dbl=1}, 0, 1, AF },
1571  { "response", "show IR frequency response", OFFSET(response), AV_OPT_TYPE_BOOL, {.i64=0}, 0, 1, VF },
1572  { "channel", "set IR channel to display frequency response", OFFSET(ir_channel), AV_OPT_TYPE_INT, {.i64=0}, 0, 1024, VF },
1573  { "size", "set video size", OFFSET(w), AV_OPT_TYPE_IMAGE_SIZE, {.str = "hd720"}, 0, 0, VF },
1574  { "rate", "set video rate", OFFSET(rate), AV_OPT_TYPE_VIDEO_RATE, {.str = "25"}, 0, INT32_MAX, VF },
1575  { NULL },
1576 };
1577 
1579 
1581  .name = "aiir",
1582  .description = NULL_IF_CONFIG_SMALL("Apply Infinite Impulse Response filter with supplied coefficients."),
1583  .priv_size = sizeof(AudioIIRContext),
1584  .priv_class = &aiir_class,
1585  .init = init,
1586  .uninit = uninit,
1588  .inputs = inputs,
1591 };
static enum AVSampleFormat sample_fmts[]
Definition: adpcmenc.c:925
#define LATTICE_IIR_CH(name, type, min, max, need_clipping)
Definition: af_aiir.c:302
#define PARALLEL_IIR_CH(name, type, min, max, need_clipping)
Definition: af_aiir.c:238
static void check_stability(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:1033
static void drawtext(AVFrame *pic, int x, int y, const char *txt, uint32_t color)
Definition: af_aiir.c:1052
static int read_zp_coefficients(AVFilterContext *ctx, char *item_str, int nb_items, double *dst, const char *format)
Definition: af_aiir.c:431
static void convert_sf2tf(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:978
static int config_video(AVFilterLink *outlink)
Definition: af_aiir.c:1448
static int read_tf_coefficients(AVFilterContext *ctx, char *item_str, int nb_items, double *dst)
Definition: af_aiir.c:406
static double distance(double x0, double x1, double y0, double y1)
Definition: af_aiir.c:1101
static void draw_response(AVFilterContext *ctx, AVFrame *out, int sample_rate)
Definition: af_aiir.c:1153
static void convert_pr2zp(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:898
static void convert_pd2zp(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:1006
static const AVOption aiir_options[]
Definition: af_aiir.c:1539
static int expand(AVFilterContext *ctx, double *pz, int n, double *coefs)
Definition: af_aiir.c:511
static int decompose_zp2biquads(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:614
static int query_formats(AVFilterContext *ctx)
Definition: af_aiir.c:79
static void draw_line(AVFrame *out, int x0, int y0, int x1, int y1, uint32_t color)
Definition: af_aiir.c:1075
static double fact(double i)
Definition: af_aiir.c:952
#define IIR_CH(name, type, min, max, need_clipping)
Definition: af_aiir.c:123
static int read_channels(AVFilterContext *ctx, int channels, uint8_t *item_str, int ab)
Definition: af_aiir.c:458
static void normalize_coeffs(AVFilterContext *ctx, int ch)
Definition: af_aiir.c:538
static const AVFilterPad inputs[]
Definition: af_aiir.c:1526
static int read_gains(AVFilterContext *ctx, char *item_str, int nb_items)
Definition: af_aiir.c:373
static const char *const format[]
Definition: af_aiir.c:456
#define AF
Definition: af_aiir.c:1536
#define SERIAL_IIR_CH(name, type, min, max, need_clipping)
Definition: af_aiir.c:179
#define VF
Definition: af_aiir.c:1537
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
Definition: af_aiir.c:1393
static void convert_sp2zp(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:925
AVFILTER_DEFINE_CLASS(aiir)
static int convert_serial2parallel(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:828
static int convert_zp2tf(AVFilterContext *ctx, int channels)
Definition: af_aiir.c:566
static void biquad_process(double *x, double *y, int length, double b0, double b1, double b2, double a1, double a2)
Definition: af_aiir.c:776
static void count_coefficients(char *item_str, int *nb_items)
Definition: af_aiir.c:359
static double coef_sf2zf(double *a, int N, int n)
Definition: af_aiir.c:959
static av_cold int init(AVFilterContext *ctx)
Definition: af_aiir.c:1462
static av_cold void uninit(AVFilterContext *ctx)
Definition: af_aiir.c:1506
static void get_response(int channel, int format, double w, const double *b, const double *a, int nb_b, int nb_a, double *magnitude, double *phase)
Definition: af_aiir.c:1106
static void cmul(double re, double im, double re2, double im2, double *RE, double *IM)
Definition: af_aiir.c:505
#define OFFSET(x)
Definition: af_aiir.c:1535
static int config_output(AVFilterLink *outlink)
Definition: af_aiir.c:1270
static void solve(double *matrix, double *vector, int n, double *y, double *x, double *lu)
Definition: af_aiir.c:794
AVFilter ff_af_aiir
Definition: af_aiir.c:1580
@ biquad
Definition: af_biquads.c:73
#define N
Definition: af_mcompand.c:54
channels
Definition: aptx.h:33
#define av_cold
Definition: attributes.h:88
AVFrame * ff_get_audio_buffer(AVFilterLink *link, int nb_samples)
Request an audio samples buffer with a specific set of permissions.
Definition: audio.c:86
uint8_t pi<< 24) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0f/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_U8, uint8_t,(*(const uint8_t *) pi - 0x80) *(1.0/(1<< 7))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S16, int16_t,(*(const int16_t *) pi >> 8)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0f/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S16, int16_t, *(const int16_t *) pi *(1.0/(1<< 15))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_S32, int32_t,(*(const int32_t *) pi >> 24)+0x80) CONV_FUNC_GROUP(AV_SAMPLE_FMT_FLT, float, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0f/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_DBL, double, AV_SAMPLE_FMT_S32, int32_t, *(const int32_t *) pi *(1.0/(1U<< 31))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_FLT, float, av_clip_uint8(lrintf(*(const float *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_FLT, float, av_clip_int16(lrintf(*(const float *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_FLT, float, av_clipl_int32(llrintf(*(const float *) pi *(1U<< 31)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_U8, uint8_t, AV_SAMPLE_FMT_DBL, double, av_clip_uint8(lrint(*(const double *) pi *(1<< 7))+0x80)) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S16, int16_t, AV_SAMPLE_FMT_DBL, double, av_clip_int16(lrint(*(const double *) pi *(1<< 15)))) CONV_FUNC_GROUP(AV_SAMPLE_FMT_S32, int32_t, AV_SAMPLE_FMT_DBL, double, av_clipl_int32(llrint(*(const double *) pi *(1U<< 31)))) #define SET_CONV_FUNC_GROUP(ofmt, ifmt) static void set_generic_function(AudioConvert *ac) { } void ff_audio_convert_free(AudioConvert **ac) { if(! *ac) return;ff_dither_free(&(*ac) ->dc);av_freep(ac);} AudioConvert *ff_audio_convert_alloc(AVAudioResampleContext *avr, enum AVSampleFormat out_fmt, enum AVSampleFormat in_fmt, int channels, int sample_rate, int apply_map) { AudioConvert *ac;int in_planar, out_planar;ac=av_mallocz(sizeof(*ac));if(!ac) return NULL;ac->avr=avr;ac->out_fmt=out_fmt;ac->in_fmt=in_fmt;ac->channels=channels;ac->apply_map=apply_map;if(avr->dither_method !=AV_RESAMPLE_DITHER_NONE &&av_get_packed_sample_fmt(out_fmt)==AV_SAMPLE_FMT_S16 &&av_get_bytes_per_sample(in_fmt) > 2) { ac->dc=ff_dither_alloc(avr, out_fmt, in_fmt, channels, sample_rate, apply_map);if(!ac->dc) { av_free(ac);return NULL;} return ac;} in_planar=ff_sample_fmt_is_planar(in_fmt, channels);out_planar=ff_sample_fmt_is_planar(out_fmt, channels);if(in_planar==out_planar) { ac->func_type=CONV_FUNC_TYPE_FLAT;ac->planes=in_planar ? ac->channels :1;} else if(in_planar) ac->func_type=CONV_FUNC_TYPE_INTERLEAVE;else ac->func_type=CONV_FUNC_TYPE_DEINTERLEAVE;set_generic_function(ac);if(ARCH_AARCH64) ff_audio_convert_init_aarch64(ac);if(ARCH_ARM) ff_audio_convert_init_arm(ac);if(ARCH_X86) ff_audio_convert_init_x86(ac);return ac;} int ff_audio_convert(AudioConvert *ac, AudioData *out, AudioData *in) { int use_generic=1;int len=in->nb_samples;int p;if(ac->dc) { av_log(ac->avr, AV_LOG_TRACE, "%d samples - audio_convert: %s to %s (dithered)\n", len, av_get_sample_fmt_name(ac->in_fmt), av_get_sample_fmt_name(ac->out_fmt));return ff_convert_dither(ac-> in
uint8_t
int32_t
simple assert() macros that are a bit more flexible than ISO C assert().
#define IM(x, ch)
#define RE(x, ch)
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:1096
Main libavfilter public API header.
#define flags(name, subs,...)
Definition: cbs_av1.c:561
#define s(width, name)
Definition: cbs_vp9.c:257
#define fail()
Definition: checkasm.h:133
#define FFMIN(a, b)
Definition: common.h:105
#define av_clip
Definition: common.h:122
#define FFMAX(a, b)
Definition: common.h:103
#define FFABS(a)
Absolute value, Note, INT_MIN / INT64_MIN result in undefined behavior as they are not representable ...
Definition: common.h:72
#define NULL
Definition: coverity.c:32
static __device__ float ceil(float a)
Definition: cuda_runtime.h:176
static __device__ float fabs(float a)
Definition: cuda_runtime.h:182
#define max(a, b)
Definition: cuda_runtime.h:33
double fmin(double, double)
double fmax(double, double)
channel
Use these values when setting the channel map with ebur128_set_channel().
Definition: ebur128.h:39
int8_t exp
Definition: eval.c:72
int
sample_rate
float im
Definition: fft.c:82
float re
Definition: fft.c:82
AVFilterChannelLayouts * ff_all_channel_counts(void)
Construct an AVFilterChannelLayouts coding for any channel layout, with known or unknown disposition.
Definition: formats.c:436
int ff_formats_ref(AVFilterFormats *f, AVFilterFormats **ref)
Add ref as a new reference to formats.
Definition: formats.c:466
int ff_set_common_formats(AVFilterContext *ctx, AVFilterFormats *formats)
A helper for query_formats() which sets all links to the same list of formats.
Definition: formats.c:587
AVFilterFormats * ff_make_format_list(const int *fmts)
Create a list of supported formats.
Definition: formats.c:286
int ff_set_common_samplerates(AVFilterContext *ctx, AVFilterFormats *samplerates)
Definition: formats.c:575
int ff_set_common_channel_layouts(AVFilterContext *ctx, AVFilterChannelLayouts *channel_layouts)
A helper for query_formats() which sets all links to the same list of channel layouts/sample rates.
Definition: formats.c:568
AVFilterFormats * ff_all_samplerates(void)
Definition: formats.c:421
@ AV_OPT_TYPE_IMAGE_SIZE
offset must point to two consecutive integers
Definition: opt.h:235
@ AV_OPT_TYPE_CONST
Definition: opt.h:234
@ AV_OPT_TYPE_VIDEO_RATE
offset must point to AVRational
Definition: opt.h:238
@ AV_OPT_TYPE_INT
Definition: opt.h:225
@ AV_OPT_TYPE_DOUBLE
Definition: opt.h:227
@ AV_OPT_TYPE_BOOL
Definition: opt.h:242
@ AV_OPT_TYPE_STRING
Definition: opt.h:229
#define AVFILTER_FLAG_DYNAMIC_OUTPUTS
The number of the filter outputs is not determined just by AVFilter.outputs.
Definition: avfilter.h:112
#define AVFILTER_FLAG_SLICE_THREADS
The filter supports multithreading by splitting frames into multiple parts and processing them concur...
Definition: avfilter.h:117
#define AVERROR_PATCHWELCOME
Not yet implemented in FFmpeg, patches welcome.
Definition: error.h:62
#define AVERROR_BUG
Internal bug, also see AVERROR_BUG2.
Definition: error.h:50
#define AVERROR(e)
Definition: error.h:43
int av_frame_is_writable(AVFrame *frame)
Check if the frame data is writable.
Definition: frame.c:594
AVFrame * av_frame_clone(const AVFrame *src)
Create a new frame that references the same data as src.
Definition: frame.c:540
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:203
int av_frame_copy_props(AVFrame *dst, const AVFrame *src)
Copy only "metadata" fields from src to dst.
Definition: frame.c:658
#define AV_LOG_WARNING
Something somehow does not look correct.
Definition: log.h:200
#define AV_LOG_VERBOSE
Detailed information.
Definition: log.h:210
#define AV_LOG_ERROR
Something went wrong and cannot losslessly be recovered.
Definition: log.h:194
static av_always_inline AVRational av_inv_q(AVRational q)
Invert a rational.
Definition: rational.h:159
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
char * av_strdup(const char *s)
Duplicate a string.
Definition: mem.c:253
void * av_calloc(size_t nmemb, size_t size)
Non-inlined equivalent of av_mallocz_array().
Definition: mem.c:245
@ AVMEDIA_TYPE_AUDIO
Definition: avutil.h:202
@ AVMEDIA_TYPE_VIDEO
Definition: avutil.h:201
AVSampleFormat
Audio sample formats.
Definition: samplefmt.h:58
@ AV_SAMPLE_FMT_FLTP
float, planar
Definition: samplefmt.h:69
@ AV_SAMPLE_FMT_S16P
signed 16 bits, planar
Definition: samplefmt.h:67
@ AV_SAMPLE_FMT_S32P
signed 32 bits, planar
Definition: samplefmt.h:68
@ AV_SAMPLE_FMT_NONE
Definition: samplefmt.h:59
@ AV_SAMPLE_FMT_DBLP
double, planar
Definition: samplefmt.h:70
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:186
int av_sscanf(const char *string, const char *format,...)
See libc sscanf manual for more information.
Definition: avsscanf.c:962
int i
Definition: input.c:407
#define AV_WL32(p, v)
Definition: intreadwrite.h:426
static int mix(int c0, int c1)
Definition: 4xm.c:715
const char * arg
Definition: jacosubdec.c:66
static int ff_insert_outpad(AVFilterContext *f, unsigned index, AVFilterPad *p)
Insert a new output pad for the filter.
Definition: internal.h:248
common internal API header
#define NULL_IF_CONFIG_SMALL(x)
Return NULL if CONFIG_SMALL is true, otherwise the argument without modification.
Definition: internal.h:117
static enum AVPixelFormat pix_fmts[]
Definition: libkvazaar.c:309
#define isnan(x)
Definition: libm.h:340
static av_const double hypot(double x, double y)
Definition: libm.h:366
uint8_t w
Definition: llviddspenc.c:39
static const uint16_t mask[17]
Definition: lzw.c:38
#define NAN
Definition: mathematics.h:64
#define M_PI
Definition: mathematics.h:52
enum MovChannelLayoutTag * layouts
Definition: mov_chan.c:434
AVOptions.
AVPixelFormat
Pixel format.
Definition: pixfmt.h:64
@ AV_PIX_FMT_NONE
Definition: pixfmt.h:65
@ AV_PIX_FMT_RGB0
packed RGB 8:8:8, 32bpp, RGBXRGBX... X=unused/undefined
Definition: pixfmt.h:238
#define a2
Definition: regdef.h:48
#define td
Definition: regdef.h:70
#define a1
Definition: regdef.h:47
formats
Definition: signature.h:48
#define snprintf
Definition: snprintf.h:34
Describe the class of an AVClass context structure.
Definition: log.h:67
A list of supported channel layouts.
Definition: formats.h:86
An instance of a filter.
Definition: avfilter.h:341
AVFilterFormats * formats
List of supported formats (pixel or sample).
Definition: avfilter.h:445
A list of supported formats for one end of a filter link.
Definition: formats.h:65
A filter pad used for either input or output.
Definition: internal.h:54
const char * name
Pad name.
Definition: internal.h:60
Filter definition.
Definition: avfilter.h:145
const char * name
Filter name.
Definition: avfilter.h:149
AVFormatInternal * internal
An opaque field for libavformat internal usage.
Definition: avformat.h:1699
This structure describes decoded (raw) audio or video data.
Definition: frame.h:318
uint8_t * data[AV_NUM_DATA_POINTERS]
pointer to the picture/channel planes.
Definition: frame.h:332
int linesize[AV_NUM_DATA_POINTERS]
For video, size in bytes of each picture line.
Definition: frame.h:349
AVOption.
Definition: opt.h:248
Rational number (pair of numerator and denominator).
Definition: rational.h:58
char * a_str
Definition: af_aiir.c:58
double wet_gain
Definition: af_aiir.c:59
int ir_channel
Definition: af_aiir.c:67
int(* iir_channel)(AVFilterContext *ctx, void *arg, int ch, int nb_jobs)
Definition: af_aiir.c:76
char * b_str
Definition: af_aiir.c:58
IIRChannel * iir
Definition: af_aiir.c:72
AVRational rate
Definition: af_aiir.c:68
double mix
Definition: af_aiir.c:60
double dry_gain
Definition: af_aiir.c:59
AVFrame * video
Definition: af_aiir.c:70
char * g_str
Definition: af_aiir.c:58
enum AVSampleFormat sample_format
Definition: af_aiir.c:74
double a[3]
Definition: af_aiir.c:41
double w1
Definition: af_aiir.c:43
double b[3]
Definition: af_aiir.c:42
double w2
Definition: af_aiir.c:43
int nb_ab[2]
Definition: af_aiir.c:47
BiquadContext * biquads
Definition: af_aiir.c:52
double * cache[2]
Definition: af_aiir.c:50
double fir
Definition: af_aiir.c:51
double * ab[2]
Definition: af_aiir.c:48
int clippings
Definition: af_aiir.c:53
double g
Definition: af_aiir.c:49
Definition: af_aiir.c:36
int b
Definition: af_aiir.c:37
int a
Definition: af_aiir.c:37
Used for passing data between threads.
Definition: dsddec.c:67
AVFrame * out
Definition: af_adeclick.c:502
AVFrame * in
Definition: af_adenorm.c:223
#define av_free(p)
#define av_malloc_array(a, b)
#define av_freep(p)
#define av_log(a,...)
FILE * out
Definition: movenc.c:54
AVFormatContext * ctx
Definition: movenc.c:48
@ W
Definition: vf_addroi.c:26
const char * b
Definition: vf_curves.c:118
const char * r
Definition: vf_curves.c:116
else temp
Definition: vf_mcdeint.c:259
static void process(NormalizeContext *s, AVFrame *in, AVFrame *out)
Definition: vf_normalize.c:156
static const int factor[16]
Definition: vf_pp7.c:77
static double b1(void *priv, double x, double y)
Definition: vf_xfade.c:1665
static double b2(void *priv, double x, double y)
Definition: vf_xfade.c:1666
static double b0(void *priv, double x, double y)
Definition: vf_xfade.c:1664
AVFrame * ff_get_video_buffer(AVFilterLink *link, int w, int h)
Request a picture buffer with a specific set of permissions.
Definition: video.c:104
float min
#define M(a, b)
Definition: vp3dsp.c:45
const uint8_t avpriv_cga_font[2048]
Definition: xga_font_data.c:29
CGA/EGA/VGA ROM font data.
int acc
Definition: yuv2rgb.c:555