FFmpeg
af_replaygain.c
Go to the documentation of this file.
1 /*
2  * Copyright (c) 1998 - 2009 Conifer Software
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 /**
22  * @file
23  * ReplayGain scanner
24  */
25 
26 #include "libavutil/avassert.h"
28 #include "audio.h"
29 #include "avfilter.h"
30 #include "internal.h"
31 
32 #define HISTOGRAM_SLOTS 12000
33 #define BUTTER_ORDER 2
34 #define YULE_ORDER 10
35 
36 typedef struct ReplayGainFreqInfo {
38  double BYule[YULE_ORDER + 1];
39  double AYule[YULE_ORDER + 1];
40  double BButter[BUTTER_ORDER + 1];
41  double AButter[BUTTER_ORDER + 1];
43 
44 static const ReplayGainFreqInfo freqinfos[] =
45 {
46  {
47  192000,
48  { 0.01184742123123, -0.04631092400086, 0.06584226961238,
49  -0.02165588522478, -0.05656260778952, 0.08607493592760,
50  -0.03375544339786, -0.04216579932754, 0.06416711490648,
51  -0.03444708260844, 0.00697275872241 },
52  { 1.00000000000000, -5.24727318348167, 10.60821585192244,
53  -8.74127665810413, -1.33906071371683, 8.07972882096606,
54  -5.46179918950847, 0.54318070652536, 0.87450969224280,
55  -0.34656083539754, 0.03034796843589 },
56  { 0.99653501465135, -1.99307002930271, 0.99653501465135 },
57  { 1.00000000000000, -1.99305802314321, 0.99308203546221 },
58  },
59  {
60  176400,
61  { 0.00268568524529, -0.00852379426080, 0.00852704191347,
62  0.00146116310295, -0.00950855828762, 0.00625449515499,
63  0.00116183868722, -0.00362461417136, 0.00203961000134,
64  -0.00050664587933, 0.00004327455427 },
65  { 1.00000000000000, -5.57512782763045, 12.44291056065794,
66  -12.87462799681221, 3.08554846961576, 6.62493459880692,
67  -7.07662766313248, 2.51175542736441, 0.06731510802735,
68  -0.24567753819213, 0.03961404162376 },
69  { 0.99622916581118, -1.99245833162236, 0.99622916581118 },
70  { 1.00000000000000, -1.99244411238133, 0.99247255086339 },
71  },
72  {
73  144000,
74  { 0.00639682359450, -0.02556437970955, 0.04230854400938,
75  -0.03722462201267, 0.01718514827295, 0.00610592243009,
76  -0.03065965747365, 0.04345745003539, -0.03298592681309,
77  0.01320937236809, -0.00220304127757 },
78  { 1.00000000000000, -6.14814623523425, 15.80002457141566,
79  -20.78487587686937, 11.98848552310315, 3.36462015062606,
80  -10.22419868359470, 6.65599702146473, -1.67141861110485,
81  -0.05417956536718, 0.07374767867406 },
82  { 0.99538268958706, -1.99076537917413, 0.99538268958706 },
83  { 1.00000000000000, -1.99074405950505, 0.99078669884321 },
84  },
85  {
86  128000,
87  { 0.00553120584305, -0.02112620545016, 0.03549076243117,
88  -0.03362498312306, 0.01425867248183, 0.01344686928787,
89  -0.03392770787836, 0.03464136459530, -0.02039116051549,
90  0.00667420794705, -0.00093763762995 },
91  { 1.00000000000000, -6.14581710839925, 16.04785903675838,
92  -22.19089131407749, 15.24756471580286, -0.52001440400238,
93  -8.00488641699940, 6.60916094768855, -2.37856022810923,
94  0.33106947986101, 0.00459820832036 },
95  { 0.99480702681278, -1.98961405362557, 0.99480702681278 },
96  { 1.00000000000000, -1.98958708647324, 0.98964102077790 },
97  },
98  {
99  112000,
100  { 0.00528778718259, -0.01893240907245, 0.03185982561867,
101  -0.02926260297838, 0.00715743034072, 0.01985743355827,
102  -0.03222614850941, 0.02565681978192, -0.01210662313473,
103  0.00325436284541, -0.00044173593001 },
104  { 1.00000000000000, -6.24932108456288, 17.42344320538476,
105  -27.86819709054896, 26.79087344681326,-13.43711081485123,
106  -0.66023612948173, 6.03658091814935, -4.24926577030310,
107  1.40829268709186, -0.19480852628112 },
108  { 0.99406737810867, -1.98813475621734, 0.99406737810867 },
109  { 1.00000000000000, -1.98809955990514, 0.98816995252954 },
110  },
111  {
112  96000,
113  { 0.00588138296683, -0.01613559730421, 0.02184798954216,
114  -0.01742490405317, 0.00464635643780, 0.01117772513205,
115  -0.02123865824368, 0.01959354413350, -0.01079720643523,
116  0.00352183686289, -0.00063124341421 },
117  { 1.00000000000000, -5.97808823642008, 16.21362507964068,
118  -25.72923730652599, 25.40470663139513,-14.66166287771134,
119  2.81597484359752, 2.51447125969733, -2.23575306985286,
120  0.75788151036791, -0.10078025199029 },
121  { 0.99308203517541, -1.98616407035082, 0.99308203517541 },
122  { 1.00000000000000, -1.98611621154089, 0.98621192916075 },
123  },
124  {
125  88200,
126  { 0.02667482047416, -0.11377479336097, 0.23063167910965,
127  -0.30726477945593, 0.33188520686529, -0.33862680249063,
128  0.31807161531340, -0.23730796929880, 0.12273894790371,
129  -0.03840017967282, 0.00549673387936 },
130  { 1.00000000000000, -6.31836451657302, 18.31351310801799,
131  -31.88210014815921, 36.53792146976740,-28.23393036467559,
132  14.24725258227189, -4.04670980012854, 0.18865757280515,
133  0.25420333563908, -0.06012333531065 },
134  { 0.99247255046129, -1.98494510092259, 0.99247255046129 },
135  { 1.00000000000000, -1.98488843762335, 0.98500176422183 },
136  },
137  {
138  64000,
139  { 0.02613056568174, -0.08128786488109, 0.14937282347325,
140  -0.21695711675126, 0.25010286673402, -0.23162283619278,
141  0.17424041833052, -0.10299599216680, 0.04258696481981,
142  -0.00977952936493, 0.00105325558889 },
143  { 1.00000000000000, -5.73625477092119, 16.15249794355035,
144  -29.68654912464508, 39.55706155674083,-39.82524556246253,
145  30.50605345013009,-17.43051772821245, 7.05154573908017,
146  -1.80783839720514, 0.22127840210813 },
147  { 0.98964101933472, -1.97928203866944, 0.98964101933472 },
148  { 1.00000000000000, -1.97917472731009, 0.97938935002880 },
149  },
150  {
151  56000,
152  { 0.03144914734085, -0.06151729206963, 0.08066788708145,
153  -0.09737939921516, 0.08943210803999, -0.06989984672010,
154  0.04926972841044, -0.03161257848451, 0.01456837493506,
155  -0.00316015108496, 0.00132807215875 },
156  { 1.00000000000000, -4.87377313090032, 12.03922160140209,
157  -20.10151118381395, 25.10388534415171,-24.29065560815903,
158  18.27158469090663,-10.45249552560593, 4.30319491872003,
159  -1.13716992070185, 0.14510733527035 },
160  { 0.98816995007392, -1.97633990014784, 0.98816995007392 },
161  { 1.00000000000000, -1.97619994516973, 0.97647985512594 },
162  },
163  {
164  48000,
165  { 0.03857599435200, -0.02160367184185, -0.00123395316851,
166  -0.00009291677959, -0.01655260341619, 0.02161526843274,
167  -0.02074045215285, 0.00594298065125, 0.00306428023191,
168  0.00012025322027, 0.00288463683916 },
169  { 1.00000000000000, -3.84664617118067, 7.81501653005538,
170  -11.34170355132042, 13.05504219327545,-12.28759895145294,
171  9.48293806319790, -5.87257861775999, 2.75465861874613,
172  -0.86984376593551, 0.13919314567432 },
173  { 0.98621192462708, -1.97242384925416, 0.98621192462708 },
174  { 1.00000000000000, -1.97223372919527, 0.97261396931306 },
175  },
176  {
177  44100,
178  { 0.05418656406430, -0.02911007808948, -0.00848709379851,
179  -0.00851165645469, -0.00834990904936, 0.02245293253339,
180  -0.02596338512915, 0.01624864962975, -0.00240879051584,
181  0.00674613682247, -0.00187763777362 },
182  { 1.00000000000000, -3.47845948550071, 6.36317777566148,
183  -8.54751527471874, 9.47693607801280, -8.81498681370155,
184  6.85401540936998, -4.39470996079559, 2.19611684890774,
185  -0.75104302451432, 0.13149317958808 },
186  { 0.98500175787242, -1.97000351574484, 0.98500175787242 },
187  { 1.00000000000000, -1.96977855582618, 0.97022847566350 },
188  },
189  {
190  37800,
191  { 0.08717879977844, -0.01000374016172, -0.06265852122368,
192  -0.01119328800950, -0.00114279372960, 0.02081333954769,
193  -0.01603261863207, 0.01936763028546, 0.00760044736442,
194  -0.00303979112271, -0.00075088605788 },
195  { 1.00000000000000, -2.62816311472146, 3.53734535817992,
196  -3.81003448678921, 3.91291636730132, -3.53518605896288,
197  2.71356866157873, -1.86723311846592, 1.12075382367659,
198  -0.48574086886890, 0.11330544663849 },
199  { 0.98252400815195, -1.96504801630391, 0.98252400815195 },
200  { 1.00000000000000, -1.96474258269041, 0.96535344991740 },
201  },
202  {
203  32000,
204  { 0.15457299681924, -0.09331049056315, -0.06247880153653,
205  0.02163541888798, -0.05588393329856, 0.04781476674921,
206  0.00222312597743, 0.03174092540049, -0.01390589421898,
207  0.00651420667831, -0.00881362733839 },
208  { 1.00000000000000, -2.37898834973084, 2.84868151156327,
209  -2.64577170229825, 2.23697657451713, -1.67148153367602,
210  1.00595954808547, -0.45953458054983, 0.16378164858596,
211  -0.05032077717131, 0.02347897407020 },
212  { 0.97938932735214, -1.95877865470428, 0.97938932735214 },
213  { 1.00000000000000, -1.95835380975398, 0.95920349965459 },
214  },
215  {
216  24000,
217  { 0.30296907319327, -0.22613988682123, -0.08587323730772,
218  0.03282930172664, -0.00915702933434, -0.02364141202522,
219  -0.00584456039913, 0.06276101321749, -0.00000828086748,
220  0.00205861885564, -0.02950134983287 },
221  { 1.00000000000000, -1.61273165137247, 1.07977492259970,
222  -0.25656257754070, -0.16276719120440, -0.22638893773906,
223  0.39120800788284, -0.22138138954925, 0.04500235387352,
224  0.02005851806501, 0.00302439095741 },
225  { 0.97531843204928, -1.95063686409857, 0.97531843204928 },
226  { 1.00000000000000, -1.95002759149878, 0.95124613669835 },
227  },
228  {
229  22050,
230  { 0.33642304856132, -0.25572241425570, -0.11828570177555,
231  0.11921148675203, -0.07834489609479, -0.00469977914380,
232  -0.00589500224440, 0.05724228140351, 0.00832043980773,
233  -0.01635381384540, -0.01760176568150 },
234  { 1.00000000000000, -1.49858979367799, 0.87350271418188,
235  0.12205022308084, -0.80774944671438, 0.47854794562326,
236  -0.12453458140019, -0.04067510197014, 0.08333755284107,
237  -0.04237348025746, 0.02977207319925 },
238  { 0.97316523498161, -1.94633046996323, 0.97316523498161 },
239  { 1.00000000000000, -1.94561023566527, 0.94705070426118 },
240  },
241  {
242  18900,
243  { 0.38524531015142, -0.27682212062067, -0.09980181488805,
244  0.09951486755646, -0.08934020156622, -0.00322369330199,
245  -0.00110329090689, 0.03784509844682, 0.01683906213303,
246  -0.01147039862572, -0.01941767987192 },
247  { 1.00000000000000, -1.29708918404534, 0.90399339674203,
248  -0.29613799017877, -0.42326645916207, 0.37934887402200,
249  -0.37919795944938, 0.23410283284785, -0.03892971758879,
250  0.00403009552351, 0.03640166626278 },
251  { 0.96535326815829, -1.93070653631658, 0.96535326815829 },
252  { 1.00000000000000, -1.92950577983524, 0.93190729279793 },
253  },
254  {
255  16000,
256  { 0.44915256608450, -0.14351757464547, -0.22784394429749,
257  -0.01419140100551, 0.04078262797139, -0.12398163381748,
258  0.04097565135648, 0.10478503600251, -0.01863887810927,
259  -0.03193428438915, 0.00541907748707 },
260  { 1.00000000000000, -0.62820619233671, 0.29661783706366,
261  -0.37256372942400, 0.00213767857124, -0.42029820170918,
262  0.22199650564824, 0.00613424350682, 0.06747620744683,
263  0.05784820375801, 0.03222754072173 },
264  { 0.96454515552826, -1.92909031105652, 0.96454515552826 },
265  { 1.00000000000000, -1.92783286977036, 0.93034775234268 },
266  },
267  {
268  12000,
269  { 0.56619470757641, -0.75464456939302, 0.16242137742230,
270  0.16744243493672, -0.18901604199609, 0.30931782841830,
271  -0.27562961986224, 0.00647310677246, 0.08647503780351,
272  -0.03788984554840, -0.00588215443421 },
273  { 1.00000000000000, -1.04800335126349, 0.29156311971249,
274  -0.26806001042947, 0.00819999645858, 0.45054734505008,
275  -0.33032403314006, 0.06739368333110, -0.04784254229033,
276  0.01639907836189, 0.01807364323573 },
277  { 0.96009142950541, -1.92018285901082, 0.96009142950541 },
278  { 1.00000000000000, -1.91858953033784, 0.92177618768381 },
279  },
280  {
281  11025,
282  { 0.58100494960553, -0.53174909058578, -0.14289799034253,
283  0.17520704835522, 0.02377945217615, 0.15558449135573,
284  -0.25344790059353, 0.01628462406333, 0.06920467763959,
285  -0.03721611395801, -0.00749618797172 },
286  { 1.00000000000000, -0.51035327095184, -0.31863563325245,
287  -0.20256413484477, 0.14728154134330, 0.38952639978999,
288  -0.23313271880868, -0.05246019024463, -0.02505961724053,
289  0.02442357316099, 0.01818801111503 },
290  { 0.95856916599601, -1.91713833199203, 0.95856916599601 },
291  { 1.00000000000000, -1.91542108074780, 0.91885558323625 },
292  },
293  {
294  8000,
295  { 0.53648789255105, -0.42163034350696, -0.00275953611929,
296  0.04267842219415, -0.10214864179676, 0.14590772289388,
297  -0.02459864859345, -0.11202315195388, -0.04060034127000,
298  0.04788665548180, -0.02217936801134 },
299  { 1.00000000000000, -0.25049871956020, -0.43193942311114,
300  -0.03424681017675, -0.04678328784242, 0.26408300200955,
301  0.15113130533216, -0.17556493366449, -0.18823009262115,
302  0.05477720428674, 0.04704409688120 },
303  { 0.94597685600279, -1.89195371200558, 0.94597685600279 },
304  { 1.00000000000000, -1.88903307939452, 0.89487434461664 },
305  },
306 };
307 
308 typedef struct ReplayGainContext {
310  float peak;
312  const double *yule_coeff_a;
313  const double *yule_coeff_b;
314  const double *butter_coeff_a;
315  const double *butter_coeff_b;
316  float yule_hist_a[256];
317  float yule_hist_b[256];
318  float butter_hist_a[256];
319  float butter_hist_b[256];
321 
323 {
326  int i, ret;
327 
328  if ((ret = ff_add_format (&formats, AV_SAMPLE_FMT_FLT )) < 0 ||
329  (ret = ff_set_common_formats (ctx , formats )) < 0 ||
332  return ret;
333 
334  formats = NULL;
335  for (i = 0; i < FF_ARRAY_ELEMS(freqinfos); i++) {
337  return ret;
338  }
339 
341 }
342 
344 {
345  AVFilterContext *ctx = inlink->dst;
346  ReplayGainContext *s = ctx->priv;
347  int i;
348 
349  for (i = 0; i < FF_ARRAY_ELEMS(freqinfos); i++) {
350  if (freqinfos[i].sample_rate == inlink->sample_rate)
351  break;
352  }
354 
355  s->yule_coeff_a = freqinfos[i].AYule;
356  s->yule_coeff_b = freqinfos[i].BYule;
357  s->butter_coeff_a = freqinfos[i].AButter;
358  s->butter_coeff_b = freqinfos[i].BButter;
359 
360  s->yule_hist_i = 20;
361  s->butter_hist_i = 4;
362  inlink->partial_buf_size =
363  inlink->min_samples =
364  inlink->max_samples = inlink->sample_rate / 20;
365 
366  return 0;
367 }
368 
369 /*
370  * Update largest absolute sample value.
371  */
372 static void calc_stereo_peak(const float *samples, int nb_samples,
373  float *peak_p)
374 {
375  float peak = 0.0;
376 
377  while (nb_samples--) {
378  if (samples[0] > peak)
379  peak = samples[0];
380  else if (-samples[0] > peak)
381  peak = -samples[0];
382 
383  if (samples[1] > peak)
384  peak = samples[1];
385  else if (-samples[1] > peak)
386  peak = -samples[1];
387 
388  samples += 2;
389  }
390 
391  *peak_p = FFMAX(peak, *peak_p);
392 }
393 
394 /*
395  * Calculate stereo RMS level. Minimum value is about -100 dB for
396  * digital silence. The 90 dB offset is to compensate for the
397  * normalized float range and 3 dB is for stereo samples.
398  */
399 static double calc_stereo_rms(const float *samples, int nb_samples)
400 {
401  int count = nb_samples;
402  double sum = 1e-16;
403 
404  while (count--) {
405  sum += samples[0] * samples[0] + samples[1] * samples[1];
406  samples += 2;
407  }
408 
409  return 10 * log10 (sum / nb_samples) + 90.0 - 3.0;
410 }
411 
412 /*
413  * Optimized implementation of 2nd-order IIR stereo filter.
414  */
416  float *samples, int nb_samples)
417 {
418  const double *coeff_a = s->butter_coeff_a;
419  const double *coeff_b = s->butter_coeff_b;
420  float *hist_a = s->butter_hist_a;
421  float *hist_b = s->butter_hist_b;
422  double left, right;
423  int i, j;
424 
425  i = s->butter_hist_i;
426 
427  // If filter history is very small magnitude, clear it completely
428  // to prevent denormals from rattling around in there forever
429  // (slowing us down).
430 
431  for (j = -4; j < 0; ++j)
432  if (fabs(hist_a[i + j]) > 1e-10 || fabs(hist_b[i + j]) > 1e-10)
433  break;
434 
435  if (!j) {
436  memset(s->butter_hist_a, 0, sizeof(s->butter_hist_a));
437  memset(s->butter_hist_b, 0, sizeof(s->butter_hist_b));
438  }
439 
440  while (nb_samples--) {
441  left = (hist_b[i ] = samples[0]) * coeff_b[0];
442  right = (hist_b[i + 1] = samples[1]) * coeff_b[0];
443  left += hist_b[i - 2] * coeff_b[1] - hist_a[i - 2] * coeff_a[1];
444  right += hist_b[i - 1] * coeff_b[1] - hist_a[i - 1] * coeff_a[1];
445  left += hist_b[i - 4] * coeff_b[2] - hist_a[i - 4] * coeff_a[2];
446  right += hist_b[i - 3] * coeff_b[2] - hist_a[i - 3] * coeff_a[2];
447  samples[0] = hist_a[i ] = (float) left;
448  samples[1] = hist_a[i + 1] = (float) right;
449  samples += 2;
450 
451  if ((i += 2) == 256) {
452  memcpy(hist_a, hist_a + 252, sizeof(*hist_a) * 4);
453  memcpy(hist_b, hist_b + 252, sizeof(*hist_b) * 4);
454  i = 4;
455  }
456  }
457 
458  s->butter_hist_i = i;
459 }
460 
461 /*
462  * Optimized implementation of 10th-order IIR stereo filter.
463  */
465  float *dst, int nb_samples)
466 {
467  const double *coeff_a = s->yule_coeff_a;
468  const double *coeff_b = s->yule_coeff_b;
469  float *hist_a = s->yule_hist_a;
470  float *hist_b = s->yule_hist_b;
471  double left, right;
472  int i, j;
473 
474  i = s->yule_hist_i;
475 
476  // If filter history is very small magnitude, clear it completely to
477  // prevent denormals from rattling around in there forever
478  // (slowing us down).
479 
480  for (j = -20; j < 0; ++j)
481  if (fabs(hist_a[i + j]) > 1e-10 || fabs(hist_b[i + j]) > 1e-10)
482  break;
483 
484  if (!j) {
485  memset(s->yule_hist_a, 0, sizeof(s->yule_hist_a));
486  memset(s->yule_hist_b, 0, sizeof(s->yule_hist_b));
487  }
488 
489  while (nb_samples--) {
490  left = (hist_b[i] = src[0]) * coeff_b[0];
491  right = (hist_b[i + 1] = src[1]) * coeff_b[0];
492  left += hist_b[i - 2] * coeff_b[ 1] - hist_a[i - 2] * coeff_a[1 ];
493  right += hist_b[i - 1] * coeff_b[ 1] - hist_a[i - 1] * coeff_a[1 ];
494  left += hist_b[i - 4] * coeff_b[ 2] - hist_a[i - 4] * coeff_a[2 ];
495  right += hist_b[i - 3] * coeff_b[ 2] - hist_a[i - 3] * coeff_a[2 ];
496  left += hist_b[i - 6] * coeff_b[ 3] - hist_a[i - 6] * coeff_a[3 ];
497  right += hist_b[i - 5] * coeff_b[ 3] - hist_a[i - 5] * coeff_a[3 ];
498  left += hist_b[i - 8] * coeff_b[ 4] - hist_a[i - 8] * coeff_a[4 ];
499  right += hist_b[i - 7] * coeff_b[ 4] - hist_a[i - 7] * coeff_a[4 ];
500  left += hist_b[i - 10] * coeff_b[ 5] - hist_a[i - 10] * coeff_a[5 ];
501  right += hist_b[i - 9] * coeff_b[ 5] - hist_a[i - 9] * coeff_a[5 ];
502  left += hist_b[i - 12] * coeff_b[ 6] - hist_a[i - 12] * coeff_a[6 ];
503  right += hist_b[i - 11] * coeff_b[ 6] - hist_a[i - 11] * coeff_a[6 ];
504  left += hist_b[i - 14] * coeff_b[ 7] - hist_a[i - 14] * coeff_a[7 ];
505  right += hist_b[i - 13] * coeff_b[ 7] - hist_a[i - 13] * coeff_a[7 ];
506  left += hist_b[i - 16] * coeff_b[ 8] - hist_a[i - 16] * coeff_a[8 ];
507  right += hist_b[i - 15] * coeff_b[ 8] - hist_a[i - 15] * coeff_a[8 ];
508  left += hist_b[i - 18] * coeff_b[ 9] - hist_a[i - 18] * coeff_a[9 ];
509  right += hist_b[i - 17] * coeff_b[ 9] - hist_a[i - 17] * coeff_a[9 ];
510  left += hist_b[i - 20] * coeff_b[10] - hist_a[i - 20] * coeff_a[10];
511  right += hist_b[i - 19] * coeff_b[10] - hist_a[i - 19] * coeff_a[10];
512  dst[0] = hist_a[i ] = (float)left;
513  dst[1] = hist_a[i + 1] = (float)right;
514  src += 2;
515  dst += 2;
516 
517  if ((i += 2) == 256) {
518  memcpy(hist_a, hist_a + 236, sizeof(*hist_a) * 20);
519  memcpy(hist_b, hist_b + 236, sizeof(*hist_b) * 20);
520  i = 20;
521  }
522  }
523 
524  s->yule_hist_i = i;
525 }
526 
527 /*
528  * Calculate the ReplayGain value from the specified loudness histogram;
529  * clip to -24 / +64 dB.
530  */
531 static float calc_replaygain(uint32_t *histogram)
532 {
533  uint32_t loud_count = 0, total_windows = 0;
534  float gain;
535  int i;
536 
537  for (i = 0; i < HISTOGRAM_SLOTS; i++)
538  total_windows += histogram [i];
539 
540  while (i--)
541  if ((loud_count += histogram [i]) * 20 >= total_windows)
542  break;
543 
544  gain = (float)(64.54 - i / 100.0);
545 
546  return av_clipf(gain, -24.0, 64.0);
547 }
548 
550 {
551  AVFilterContext *ctx = inlink->dst;
552  AVFilterLink *outlink = ctx->outputs[0];
553  ReplayGainContext *s = ctx->priv;
554  uint32_t level;
555  AVFrame *out;
556 
557  out = ff_get_audio_buffer(outlink, in->nb_samples);
558  if (!out) {
559  av_frame_free(&in);
560  return AVERROR(ENOMEM);
561  }
562 
563  calc_stereo_peak((float *)in->data[0],
564  in->nb_samples, &s->peak);
565  yule_filter_stereo_samples(s, (const float *)in->data[0],
566  (float *)out->data[0],
567  out->nb_samples);
568  butter_filter_stereo_samples(s, (float *)out->data[0],
569  out->nb_samples);
570  level = (uint32_t)floor(100 * calc_stereo_rms((float *)out->data[0],
571  out->nb_samples));
572  level = av_clip(level, 0, HISTOGRAM_SLOTS - 1);
573 
574  s->histogram[level]++;
575 
576  av_frame_free(&out);
577  return ff_filter_frame(outlink, in);
578 }
579 
581 {
582  ReplayGainContext *s = ctx->priv;
583  float gain = calc_replaygain(s->histogram);
584 
585  av_log(ctx, AV_LOG_INFO, "track_gain = %+.2f dB\n", gain);
586  av_log(ctx, AV_LOG_INFO, "track_peak = %.6f\n", s->peak);
587 }
588 
589 static const AVFilterPad replaygain_inputs[] = {
590  {
591  .name = "default",
592  .type = AVMEDIA_TYPE_AUDIO,
593  .filter_frame = filter_frame,
594  .config_props = config_input,
595  },
596  { NULL }
597 };
598 
599 static const AVFilterPad replaygain_outputs[] = {
600  {
601  .name = "default",
602  .type = AVMEDIA_TYPE_AUDIO,
603  },
604  { NULL }
605 };
606 
608  .name = "replaygain",
609  .description = NULL_IF_CONFIG_SMALL("ReplayGain scanner."),
610  .query_formats = query_formats,
611  .uninit = uninit,
612  .priv_size = sizeof(ReplayGainContext),
615 };
formats
formats
Definition: signature.h:48
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:86
AVFilterChannelLayouts
A list of supported channel layouts.
Definition: formats.h:85
level
uint8_t level
Definition: svq3.c:207
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
out
FILE * out
Definition: movenc.c:54
YULE_ORDER
#define YULE_ORDER
Definition: af_replaygain.c:34
ff_set_common_channel_layouts
int ff_set_common_channel_layouts(AVFilterContext *ctx, AVFilterChannelLayouts *layouts)
A helper for query_formats() which sets all links to the same list of channel layouts/sample rates.
Definition: formats.c:549
ff_filter_frame
int ff_filter_frame(AVFilterLink *link, AVFrame *frame)
Send a frame of data to the next filter.
Definition: avfilter.c:1080
HISTOGRAM_SLOTS
#define HISTOGRAM_SLOTS
Definition: af_replaygain.c:32
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
count
void INT64 INT64 count
Definition: avisynth_c.h:767
av_frame_free
void av_frame_free(AVFrame **frame)
Free the frame and any dynamically allocated objects in it, e.g.
Definition: frame.c:202
AVFrame
This structure describes decoded (raw) audio or video data.
Definition: frame.h:295
query_formats
static int query_formats(AVFilterContext *ctx)
Definition: af_replaygain.c:322
ReplayGainContext::butter_hist_i
int butter_hist_i
Definition: af_replaygain.c:311
AVFilter::name
const char * name
Filter name.
Definition: avfilter.h:148
sample_rate
sample_rate
Definition: ffmpeg_filter.c:191
ReplayGainContext
Definition: af_replaygain.c:308
ReplayGainContext::butter_coeff_a
const double * butter_coeff_a
Definition: af_replaygain.c:314
AVFilterFormats
A list of supported formats for one end of a filter link.
Definition: formats.h:64
calc_replaygain
static float calc_replaygain(uint32_t *histogram)
Definition: af_replaygain.c:531
butter_filter_stereo_samples
static void butter_filter_stereo_samples(ReplayGainContext *s, float *samples, int nb_samples)
Definition: af_replaygain.c:415
src
#define src
Definition: vp8dsp.c:254
AV_CH_LAYOUT_STEREO
#define AV_CH_LAYOUT_STEREO
Definition: channel_layout.h:86
AVFilterPad
A filter pad used for either input or output.
Definition: internal.h:54
avassert.h
av_cold
#define av_cold
Definition: attributes.h:84
ff_set_common_formats
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:568
ReplayGainContext::yule_coeff_a
const double * yule_coeff_a
Definition: af_replaygain.c:312
ff_add_channel_layout
int ff_add_channel_layout(AVFilterChannelLayouts **l, uint64_t channel_layout)
Definition: formats.c:343
s
#define s(width, name)
Definition: cbs_vp9.c:257
AVMEDIA_TYPE_AUDIO
@ AVMEDIA_TYPE_AUDIO
Definition: avutil.h:202
av_assert0
#define av_assert0(cond)
assert() equivalent, that is always enabled.
Definition: avassert.h:37
outputs
static const AVFilterPad outputs[]
Definition: af_acontrast.c:203
ctx
AVFormatContext * ctx
Definition: movenc.c:48
ReplayGainFreqInfo::AButter
double AButter[BUTTER_ORDER+1]
Definition: af_replaygain.c:41
ReplayGainContext::yule_coeff_b
const double * yule_coeff_b
Definition: af_replaygain.c:313
ReplayGainContext::yule_hist_i
int yule_hist_i
Definition: af_replaygain.c:311
ReplayGainContext::butter_coeff_b
const double * butter_coeff_b
Definition: af_replaygain.c:315
ReplayGainContext::butter_hist_a
float butter_hist_a[256]
Definition: af_replaygain.c:318
filter_frame
static int filter_frame(AVFilterLink *inlink, AVFrame *in)
Definition: af_replaygain.c:549
ReplayGainContext::peak
float peak
Definition: af_replaygain.c:310
NULL
#define NULL
Definition: coverity.c:32
replaygain_inputs
static const AVFilterPad replaygain_inputs[]
Definition: af_replaygain.c:589
ReplayGainFreqInfo::sample_rate
int sample_rate
Definition: af_replaygain.c:37
ff_add_format
int ff_add_format(AVFilterFormats **avff, int64_t fmt)
Add fmt to the list of media formats contained in *avff.
Definition: formats.c:337
config_input
static int config_input(AVFilterLink *inlink)
Definition: af_replaygain.c:343
inputs
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 inputs
Definition: filter_design.txt:243
ff_af_replaygain
AVFilter ff_af_replaygain
Definition: af_replaygain.c:607
uninit
static av_cold void uninit(AVFilterContext *ctx)
Definition: af_replaygain.c:580
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:188
FFMAX
#define FFMAX(a, b)
Definition: common.h:94
ReplayGainContext::histogram
uint32_t histogram[HISTOGRAM_SLOTS]
Definition: af_replaygain.c:309
AV_LOG_INFO
#define AV_LOG_INFO
Standard information.
Definition: log.h:187
yule_filter_stereo_samples
static void yule_filter_stereo_samples(ReplayGainContext *s, const float *src, float *dst, int nb_samples)
Definition: af_replaygain.c:464
internal.h
layout
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 layout
Definition: filter_design.txt:18
in
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
Definition: audio_convert.c:326
i
#define i(width, name, range_min, range_max)
Definition: cbs_h2645.c:259
ReplayGainContext::yule_hist_b
float yule_hist_b[256]
Definition: af_replaygain.c:317
replaygain_outputs
static const AVFilterPad replaygain_outputs[]
Definition: af_replaygain.c:599
BUTTER_ORDER
#define BUTTER_ORDER
Definition: af_replaygain.c:33
AVFilterPad::name
const char * name
Pad name.
Definition: internal.h:60
calc_stereo_rms
static double calc_stereo_rms(const float *samples, int nb_samples)
Definition: af_replaygain.c:399
AVFilter
Filter definition.
Definition: avfilter.h:144
ret
ret
Definition: filter_design.txt:187
freqinfos
static const ReplayGainFreqInfo freqinfos[]
Definition: af_replaygain.c:44
ReplayGainFreqInfo::BButter
double BButter[BUTTER_ORDER+1]
Definition: af_replaygain.c:40
ReplayGainFreqInfo
Definition: af_replaygain.c:36
left
Tag MUST be and< 10hcoeff half pel interpolation filter coefficients, hcoeff[0] are the 2 middle coefficients[1] are the next outer ones and so on, resulting in a filter like:...eff[2], hcoeff[1], hcoeff[0], hcoeff[0], hcoeff[1], hcoeff[2] ... the sign of the coefficients is not explicitly stored but alternates after each coeff and coeff[0] is positive, so ...,+,-,+,-,+,+,-,+,-,+,... hcoeff[0] is not explicitly stored but found by subtracting the sum of all stored coefficients with signs from 32 hcoeff[0]=32 - hcoeff[1] - hcoeff[2] - ... a good choice for hcoeff and htaps is htaps=6 hcoeff={40,-10, 2} an alternative which requires more computations at both encoder and decoder side and may or may not be better is htaps=8 hcoeff={42,-14, 6,-2}ref_frames minimum of the number of available reference frames and max_ref_frames for example the first frame after a key frame always has ref_frames=1spatial_decomposition_type wavelet type 0 is a 9/7 symmetric compact integer wavelet 1 is a 5/3 symmetric compact integer wavelet others are reserved stored as delta from last, last is reset to 0 if always_reset||keyframeqlog quality(logarithmic quantizer scale) stored as delta from last, last is reset to 0 if always_reset||keyframemv_scale stored as delta from last, last is reset to 0 if always_reset||keyframe FIXME check that everything works fine if this changes between framesqbias dequantization bias stored as delta from last, last is reset to 0 if always_reset||keyframeblock_max_depth maximum depth of the block tree stored as delta from last, last is reset to 0 if always_reset||keyframequant_table quantization tableHighlevel bitstream structure:==============================--------------------------------------------|Header|--------------------------------------------|------------------------------------|||Block0||||split?||||yes no||||......... intra?||||:Block01 :yes no||||:Block02 :....... ..........||||:Block03 ::y DC ::ref index:||||:Block04 ::cb DC ::motion x :||||......... :cr DC ::motion y :||||....... ..........|||------------------------------------||------------------------------------|||Block1|||...|--------------------------------------------|------------ ------------ ------------|||Y subbands||Cb subbands||Cr subbands||||--- ---||--- ---||--- ---|||||LL0||HL0||||LL0||HL0||||LL0||HL0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||LH0||HH0||||LH0||HH0||||LH0||HH0|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HL1||LH1||||HL1||LH1||||HL1||LH1|||||--- ---||--- ---||--- ---||||--- ---||--- ---||--- ---|||||HH1||HL2||||HH1||HL2||||HH1||HL2|||||...||...||...|||------------ ------------ ------------|--------------------------------------------Decoding process:=================------------|||Subbands|------------||||------------|Intra DC||||LL0 subband prediction ------------|\ Dequantization ------------------- \||Reference frames|\ IDWT|------- -------|Motion \|||Frame 0||Frame 1||Compensation . OBMC v -------|------- -------|--------------. \------> Frame n output Frame Frame<----------------------------------/|...|------------------- Range Coder:============Binary Range Coder:------------------- The implemented range coder is an adapted version based upon "Range encoding: an algorithm for removing redundancy from a digitised message." by G. N. N. Martin. The symbols encoded by the Snow range coder are bits(0|1). The associated probabilities are not fix but change depending on the symbol mix seen so far. bit seen|new state ---------+----------------------------------------------- 0|256 - state_transition_table[256 - old_state];1|state_transition_table[old_state];state_transition_table={ 0, 0, 0, 0, 0, 0, 0, 0, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 190, 191, 192, 194, 194, 195, 196, 197, 198, 199, 200, 201, 202, 202, 204, 205, 206, 207, 208, 209, 209, 210, 211, 212, 213, 215, 215, 216, 217, 218, 219, 220, 220, 222, 223, 224, 225, 226, 227, 227, 229, 229, 230, 231, 232, 234, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 248, 0, 0, 0, 0, 0, 0, 0};FIXME Range Coding of integers:------------------------- FIXME Neighboring Blocks:===================left and top are set to the respective blocks unless they are outside of the image in which case they are set to the Null block top-left is set to the top left block unless it is outside of the image in which case it is set to the left block if this block has no larger parent block or it is at the left side of its parent block and the top right block is not outside of the image then the top right block is used for top-right else the top-left block is used Null block y, cb, cr are 128 level, ref, mx and my are 0 Motion Vector Prediction:=========================1. the motion vectors of all the neighboring blocks are scaled to compensate for the difference of reference frames scaled_mv=(mv *(256 *(current_reference+1)/(mv.reference+1))+128)> the median of the scaled left
Definition: snow.txt:386
FF_ARRAY_ELEMS
#define FF_ARRAY_ELEMS(a)
Definition: sinewin_tablegen_template.c:38
channel_layout.h
ReplayGainContext::yule_hist_a
float yule_hist_a[256]
Definition: af_replaygain.c:316
avfilter.h
calc_stereo_peak
static void calc_stereo_peak(const float *samples, int nb_samples, float *peak_p)
Definition: af_replaygain.c:372
samples
Filter the word “frame” indicates either a video frame or a group of audio samples
Definition: filter_design.txt:8
AVFilterContext
An instance of a filter.
Definition: avfilter.h:338
audio.h
ReplayGainFreqInfo::AYule
double AYule[YULE_ORDER+1]
Definition: af_replaygain.c:39
ReplayGainContext::butter_hist_b
float butter_hist_b[256]
Definition: af_replaygain.c:319
av_log
#define av_log(a,...)
Definition: tableprint_vlc.h:28
ff_set_common_samplerates
int ff_set_common_samplerates(AVFilterContext *ctx, AVFilterFormats *samplerates)
Definition: formats.c:556
AV_SAMPLE_FMT_FLT
@ AV_SAMPLE_FMT_FLT
float
Definition: samplefmt.h:63
ReplayGainFreqInfo::BYule
double BYule[YULE_ORDER+1]
Definition: af_replaygain.c:38