lamejs
Version:
Pure JavaScript MP3 Encoder
553 lines (503 loc) • 22.9 kB
JavaScript
/*
* ReplayGainAnalysis - analyzes input samples and give the recommended dB change
* Copyright (C) 2001 David Robinson and Glen Sawyer
* Improvements and optimizations added by Frank Klemm, and by Marcel Muller
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* concept and filter values by David Robinson (David@Robinson.org)
* -- blame him if you think the idea is flawed
* original coding by Glen Sawyer (mp3gain@hotmail.com)
* -- blame him if you think this runs too slowly, or the coding is otherwise flawed
*
* lots of code improvements by Frank Klemm ( http://www.uni-jena.de/~pfk/mpp/ )
* -- credit him for all the _good_ programming ;)
*
*
* For an explanation of the concepts and the basic algorithms involved, go to:
* http://www.replaygain.org/
*/
/*
* Here's the deal. Call
*
* InitGainAnalysis ( long samplefreq );
*
* to initialize everything. Call
*
* AnalyzeSamples ( var Float_t* left_samples,
* var Float_t* right_samples,
* size_t num_samples,
* int num_channels );
*
* as many times as you want, with as many or as few samples as you want.
* If mono, pass the sample buffer in through left_samples, leave
* right_samples NULL, and make sure num_channels = 1.
*
* GetTitleGain()
*
* will return the recommended dB level change for all samples analyzed
* SINCE THE LAST TIME you called GetTitleGain() OR InitGainAnalysis().
*
* GetAlbumGain()
*
* will return the recommended dB level change for all samples analyzed
* since InitGainAnalysis() was called and finalized with GetTitleGain().
*
* Pseudo-code to process an album:
*
* Float_t l_samples [4096];
* Float_t r_samples [4096];
* size_t num_samples;
* unsigned int num_songs;
* unsigned int i;
*
* InitGainAnalysis ( 44100 );
* for ( i = 1; i <= num_songs; i++ ) {
* while ( ( num_samples = getSongSamples ( song[i], left_samples, right_samples ) ) > 0 )
* AnalyzeSamples ( left_samples, right_samples, num_samples, 2 );
* fprintf ("Recommended dB change for song %2d: %+6.2 dB\n", i, GetTitleGain() );
* }
* fprintf ("Recommended dB change for whole album: %+6.2 dB\n", GetAlbumGain() );
*/
/*
* So here's the main source of potential code confusion:
*
* The filters applied to the incoming samples are IIR filters,
* meaning they rely on up to <filter order> number of previous samples
* AND up to <filter order> number of previous filtered samples.
*
* I set up the AnalyzeSamples routine to minimize memory usage and interface
* complexity. The speed isn't compromised too much (I don't think), but the
* internal complexity is higher than it should be for such a relatively
* simple routine.
*
* Optimization/clarity suggestions are welcome.
*/
var common = require('./common.js');
var System = common.System;
var VbrMode = common.VbrMode;
var Float = common.Float;
var ShortBlock = common.ShortBlock;
var Util = common.Util;
var Arrays = common.Arrays;
var new_array_n = common.new_array_n;
var new_byte = common.new_byte;
var new_double = common.new_double;
var new_float = common.new_float;
var new_float_n = common.new_float_n;
var new_int = common.new_int;
var new_int_n = common.new_int_n;
var assert = common.assert;
/**
* Table entries per dB
*/
GainAnalysis.STEPS_per_dB = 100.;
/**
* Table entries for 0...MAX_dB (normal max. values are 70...80 dB)
*/
GainAnalysis.MAX_dB = 120.;
GainAnalysis.GAIN_NOT_ENOUGH_SAMPLES = -24601;
GainAnalysis.GAIN_ANALYSIS_ERROR = 0;
GainAnalysis.GAIN_ANALYSIS_OK = 1;
GainAnalysis.INIT_GAIN_ANALYSIS_ERROR = 0;
GainAnalysis.INIT_GAIN_ANALYSIS_OK = 1;
GainAnalysis.YULE_ORDER = 10;
GainAnalysis.MAX_ORDER = GainAnalysis.YULE_ORDER;
GainAnalysis.MAX_SAMP_FREQ = 48000;
GainAnalysis.RMS_WINDOW_TIME_NUMERATOR = 1;
GainAnalysis.RMS_WINDOW_TIME_DENOMINATOR = 20;
GainAnalysis.MAX_SAMPLES_PER_WINDOW = ((GainAnalysis.MAX_SAMP_FREQ * GainAnalysis.RMS_WINDOW_TIME_NUMERATOR) / GainAnalysis.RMS_WINDOW_TIME_DENOMINATOR + 1);
function GainAnalysis() {
/**
* calibration value for 89dB
*/
var PINK_REF = 64.82;
var YULE_ORDER = GainAnalysis.YULE_ORDER;
/**
* percentile which is louder than the proposed level
*/
var RMS_PERCENTILE = 0.95;
/**
* maximum allowed sample frequency [Hz]
*/
var MAX_SAMP_FREQ = GainAnalysis.MAX_SAMP_FREQ;
var RMS_WINDOW_TIME_NUMERATOR = GainAnalysis.RMS_WINDOW_TIME_NUMERATOR;
/**
* numerator / denominator = time slice size [s]
*/
var RMS_WINDOW_TIME_DENOMINATOR = GainAnalysis.RMS_WINDOW_TIME_DENOMINATOR;
/**
* max. Samples per Time slice
*/
var MAX_SAMPLES_PER_WINDOW = GainAnalysis.MAX_SAMPLES_PER_WINDOW;
var ABYule = [
[0.03857599435200, -3.84664617118067, -0.02160367184185,
7.81501653005538, -0.00123395316851, -11.34170355132042,
-0.00009291677959, 13.05504219327545, -0.01655260341619,
-12.28759895145294, 0.02161526843274, 9.48293806319790,
-0.02074045215285, -5.87257861775999, 0.00594298065125,
2.75465861874613, 0.00306428023191, -0.86984376593551,
0.00012025322027, 0.13919314567432, 0.00288463683916],
[0.05418656406430, -3.47845948550071, -0.02911007808948,
6.36317777566148, -0.00848709379851, -8.54751527471874,
-0.00851165645469, 9.47693607801280, -0.00834990904936,
-8.81498681370155, 0.02245293253339, 6.85401540936998,
-0.02596338512915, -4.39470996079559, 0.01624864962975,
2.19611684890774, -0.00240879051584, -0.75104302451432,
0.00674613682247, 0.13149317958808, -0.00187763777362],
[0.15457299681924, -2.37898834973084, -0.09331049056315,
2.84868151156327, -0.06247880153653, -2.64577170229825,
0.02163541888798, 2.23697657451713, -0.05588393329856,
-1.67148153367602, 0.04781476674921, 1.00595954808547,
0.00222312597743, -0.45953458054983, 0.03174092540049,
0.16378164858596, -0.01390589421898, -0.05032077717131,
0.00651420667831, 0.02347897407020, -0.00881362733839],
[0.30296907319327, -1.61273165137247, -0.22613988682123,
1.07977492259970, -0.08587323730772, -0.25656257754070,
0.03282930172664, -0.16276719120440, -0.00915702933434,
-0.22638893773906, -0.02364141202522, 0.39120800788284,
-0.00584456039913, -0.22138138954925, 0.06276101321749,
0.04500235387352, -0.00000828086748, 0.02005851806501,
0.00205861885564, 0.00302439095741, -0.02950134983287],
[0.33642304856132, -1.49858979367799, -0.25572241425570,
0.87350271418188, -0.11828570177555, 0.12205022308084,
0.11921148675203, -0.80774944671438, -0.07834489609479,
0.47854794562326, -0.00469977914380, -0.12453458140019,
-0.00589500224440, -0.04067510197014, 0.05724228140351,
0.08333755284107, 0.00832043980773, -0.04237348025746,
-0.01635381384540, 0.02977207319925, -0.01760176568150],
[0.44915256608450, -0.62820619233671, -0.14351757464547,
0.29661783706366, -0.22784394429749, -0.37256372942400,
-0.01419140100551, 0.00213767857124, 0.04078262797139,
-0.42029820170918, -0.12398163381748, 0.22199650564824,
0.04097565135648, 0.00613424350682, 0.10478503600251,
0.06747620744683, -0.01863887810927, 0.05784820375801,
-0.03193428438915, 0.03222754072173, 0.00541907748707],
[0.56619470757641, -1.04800335126349, -0.75464456939302,
0.29156311971249, 0.16242137742230, -0.26806001042947,
0.16744243493672, 0.00819999645858, -0.18901604199609,
0.45054734505008, 0.30931782841830, -0.33032403314006,
-0.27562961986224, 0.06739368333110, 0.00647310677246,
-0.04784254229033, 0.08647503780351, 0.01639907836189,
-0.03788984554840, 0.01807364323573, -0.00588215443421],
[0.58100494960553, -0.51035327095184, -0.53174909058578,
-0.31863563325245, -0.14289799034253, -0.20256413484477,
0.17520704835522, 0.14728154134330, 0.02377945217615,
0.38952639978999, 0.15558449135573, -0.23313271880868,
-0.25344790059353, -0.05246019024463, 0.01628462406333,
-0.02505961724053, 0.06920467763959, 0.02442357316099,
-0.03721611395801, 0.01818801111503, -0.00749618797172],
[0.53648789255105, -0.25049871956020, -0.42163034350696,
-0.43193942311114, -0.00275953611929, -0.03424681017675,
0.04267842219415, -0.04678328784242, -0.10214864179676,
0.26408300200955, 0.14590772289388, 0.15113130533216,
-0.02459864859345, -0.17556493366449, -0.11202315195388,
-0.18823009262115, -0.04060034127000, 0.05477720428674,
0.04788665548180, 0.04704409688120, -0.02217936801134]];
var ABButter = [
[0.98621192462708, -1.97223372919527, -1.97242384925416,
0.97261396931306, 0.98621192462708],
[0.98500175787242, -1.96977855582618, -1.97000351574484,
0.97022847566350, 0.98500175787242],
[0.97938932735214, -1.95835380975398, -1.95877865470428,
0.95920349965459, 0.97938932735214],
[0.97531843204928, -1.95002759149878, -1.95063686409857,
0.95124613669835, 0.97531843204928],
[0.97316523498161, -1.94561023566527, -1.94633046996323,
0.94705070426118, 0.97316523498161],
[0.96454515552826, -1.92783286977036, -1.92909031105652,
0.93034775234268, 0.96454515552826],
[0.96009142950541, -1.91858953033784, -1.92018285901082,
0.92177618768381, 0.96009142950541],
[0.95856916599601, -1.91542108074780, -1.91713833199203,
0.91885558323625, 0.95856916599601],
[0.94597685600279, -1.88903307939452, -1.89195371200558,
0.89487434461664, 0.94597685600279]];
/**
* When calling this procedure, make sure that ip[-order] and op[-order]
* point to real data
*/
//private void filterYule(final float[] input, int inputPos, float[] output,
//int outputPos, int nSamples, final float[] kernel) {
function filterYule(input, inputPos, output, outputPos, nSamples, kernel) {
while ((nSamples--) != 0) {
/* 1e-10 is a hack to avoid slowdown because of denormals */
output[outputPos] = 1e-10 + input[inputPos + 0] * kernel[0]
- output[outputPos - 1] * kernel[1] + input[inputPos - 1]
* kernel[2] - output[outputPos - 2] * kernel[3]
+ input[inputPos - 2] * kernel[4] - output[outputPos - 3]
* kernel[5] + input[inputPos - 3] * kernel[6]
- output[outputPos - 4] * kernel[7] + input[inputPos - 4]
* kernel[8] - output[outputPos - 5] * kernel[9]
+ input[inputPos - 5] * kernel[10] - output[outputPos - 6]
* kernel[11] + input[inputPos - 6] * kernel[12]
- output[outputPos - 7] * kernel[13] + input[inputPos - 7]
* kernel[14] - output[outputPos - 8] * kernel[15]
+ input[inputPos - 8] * kernel[16] - output[outputPos - 9]
* kernel[17] + input[inputPos - 9] * kernel[18]
- output[outputPos - 10] * kernel[19]
+ input[inputPos - 10] * kernel[20];
++outputPos;
++inputPos;
}
}
//private void filterButter(final float[] input, int inputPos,
// float[] output, int outputPos, int nSamples, final float[] kernel) {
function filterButter(input, inputPos, output, outputPos, nSamples, kernel) {
while ((nSamples--) != 0) {
output[outputPos] = input[inputPos + 0] * kernel[0]
- output[outputPos - 1] * kernel[1] + input[inputPos - 1]
* kernel[2] - output[outputPos - 2] * kernel[3]
+ input[inputPos - 2] * kernel[4];
++outputPos;
++inputPos;
}
}
/**
* @return INIT_GAIN_ANALYSIS_OK if successful, INIT_GAIN_ANALYSIS_ERROR if
* not
*/
function ResetSampleFrequency(rgData, samplefreq) {
/* zero out initial values */
for (var i = 0; i < MAX_ORDER; i++)
rgData.linprebuf[i] = rgData.lstepbuf[i] = rgData.loutbuf[i] = rgData.rinprebuf[i] = rgData.rstepbuf[i] = rgData.routbuf[i] = 0.;
switch (0 | (samplefreq)) {
case 48000:
rgData.reqindex = 0;
break;
case 44100:
rgData.reqindex = 1;
break;
case 32000:
rgData.reqindex = 2;
break;
case 24000:
rgData.reqindex = 3;
break;
case 22050:
rgData.reqindex = 4;
break;
case 16000:
rgData.reqindex = 5;
break;
case 12000:
rgData.reqindex = 6;
break;
case 11025:
rgData.reqindex = 7;
break;
case 8000:
rgData.reqindex = 8;
break;
default:
return INIT_GAIN_ANALYSIS_ERROR;
}
rgData.sampleWindow = 0 | ((samplefreq * RMS_WINDOW_TIME_NUMERATOR
+ RMS_WINDOW_TIME_DENOMINATOR - 1) / RMS_WINDOW_TIME_DENOMINATOR);
rgData.lsum = 0.;
rgData.rsum = 0.;
rgData.totsamp = 0;
Arrays.ill(rgData.A, 0);
return INIT_GAIN_ANALYSIS_OK;
}
this.InitGainAnalysis = function (rgData, samplefreq) {
if (ResetSampleFrequency(rgData, samplefreq) != INIT_GAIN_ANALYSIS_OK) {
return INIT_GAIN_ANALYSIS_ERROR;
}
rgData.linpre = MAX_ORDER;
rgData.rinpre = MAX_ORDER;
rgData.lstep = MAX_ORDER;
rgData.rstep = MAX_ORDER;
rgData.lout = MAX_ORDER;
rgData.rout = MAX_ORDER;
Arrays.fill(rgData.B, 0);
return INIT_GAIN_ANALYSIS_OK;
};
/**
* square
*/
function fsqr(d) {
return d * d;
}
this.AnalyzeSamples = function (rgData, left_samples, left_samplesPos, right_samples, right_samplesPos, num_samples,
num_channels) {
var curleft;
var curleftBase;
var curright;
var currightBase;
var batchsamples;
var cursamples;
var cursamplepos;
if (num_samples == 0)
return GAIN_ANALYSIS_OK;
cursamplepos = 0;
batchsamples = num_samples;
switch (num_channels) {
case 1:
right_samples = left_samples;
right_samplesPos = left_samplesPos;
break;
case 2:
break;
default:
return GAIN_ANALYSIS_ERROR;
}
if (num_samples < MAX_ORDER) {
System.arraycopy(left_samples, left_samplesPos, rgData.linprebuf,
MAX_ORDER, num_samples);
System.arraycopy(right_samples, right_samplesPos, rgData.rinprebuf,
MAX_ORDER, num_samples);
} else {
System.arraycopy(left_samples, left_samplesPos, rgData.linprebuf,
MAX_ORDER, MAX_ORDER);
System.arraycopy(right_samples, right_samplesPos, rgData.rinprebuf,
MAX_ORDER, MAX_ORDER);
}
while (batchsamples > 0) {
cursamples = batchsamples > rgData.sampleWindow - rgData.totsamp ? rgData.sampleWindow
- rgData.totsamp
: batchsamples;
if (cursamplepos < MAX_ORDER) {
curleft = rgData.linpre + cursamplepos;
curleftBase = rgData.linprebuf;
curright = rgData.rinpre + cursamplepos;
currightBase = rgData.rinprebuf;
if (cursamples > MAX_ORDER - cursamplepos)
cursamples = MAX_ORDER - cursamplepos;
} else {
curleft = left_samplesPos + cursamplepos;
curleftBase = left_samples;
curright = right_samplesPos + cursamplepos;
currightBase = right_samples;
}
filterYule(curleftBase, curleft, rgData.lstepbuf, rgData.lstep
+ rgData.totsamp, cursamples, ABYule[rgData.reqindex]);
filterYule(currightBase, curright, rgData.rstepbuf, rgData.rstep
+ rgData.totsamp, cursamples, ABYule[rgData.reqindex]);
filterButter(rgData.lstepbuf, rgData.lstep + rgData.totsamp,
rgData.loutbuf, rgData.lout + rgData.totsamp, cursamples,
ABButter[rgData.reqindex]);
filterButter(rgData.rstepbuf, rgData.rstep + rgData.totsamp,
rgData.routbuf, rgData.rout + rgData.totsamp, cursamples,
ABButter[rgData.reqindex]);
curleft = rgData.lout + rgData.totsamp;
/* Get the squared values */
curleftBase = rgData.loutbuf;
curright = rgData.rout + rgData.totsamp;
currightBase = rgData.routbuf;
var i = cursamples % 8;
while ((i--) != 0) {
rgData.lsum += fsqr(curleftBase[curleft++]);
rgData.rsum += fsqr(currightBase[curright++]);
}
i = cursamples / 8;
while ((i--) != 0) {
rgData.lsum += fsqr(curleftBase[curleft + 0])
+ fsqr(curleftBase[curleft + 1])
+ fsqr(curleftBase[curleft + 2])
+ fsqr(curleftBase[curleft + 3])
+ fsqr(curleftBase[curleft + 4])
+ fsqr(curleftBase[curleft + 5])
+ fsqr(curleftBase[curleft + 6])
+ fsqr(curleftBase[curleft + 7]);
curleft += 8;
rgData.rsum += fsqr(currightBase[curright + 0])
+ fsqr(currightBase[curright + 1])
+ fsqr(currightBase[curright + 2])
+ fsqr(currightBase[curright + 3])
+ fsqr(currightBase[curright + 4])
+ fsqr(currightBase[curright + 5])
+ fsqr(currightBase[curright + 6])
+ fsqr(currightBase[curright + 7]);
curright += 8;
}
batchsamples -= cursamples;
cursamplepos += cursamples;
rgData.totsamp += cursamples;
if (rgData.totsamp == rgData.sampleWindow) {
/* Get the Root Mean Square (RMS) for this set of samples */
var val = GainAnalysis.STEPS_per_dB
* 10.
* Math.log10((rgData.lsum + rgData.rsum)
/ rgData.totsamp * 0.5 + 1.e-37);
var ival = (val <= 0) ? 0 : 0 | val;
if (ival >= rgData.A.length)
ival = rgData.A.length - 1;
rgData.A[ival]++;
rgData.lsum = rgData.rsum = 0.;
System.arraycopy(rgData.loutbuf, rgData.totsamp,
rgData.loutbuf, 0, MAX_ORDER);
System.arraycopy(rgData.routbuf, rgData.totsamp,
rgData.routbuf, 0, MAX_ORDER);
System.arraycopy(rgData.lstepbuf, rgData.totsamp,
rgData.lstepbuf, 0, MAX_ORDER);
System.arraycopy(rgData.rstepbuf, rgData.totsamp,
rgData.rstepbuf, 0, MAX_ORDER);
rgData.totsamp = 0;
}
if (rgData.totsamp > rgData.sampleWindow) {
/*
* somehow I really screwed up: Error in programming! Contact
* author about totsamp > sampleWindow
*/
return GAIN_ANALYSIS_ERROR;
}
}
if (num_samples < MAX_ORDER) {
System.arraycopy(rgData.linprebuf, num_samples, rgData.linprebuf,
0, MAX_ORDER - num_samples);
System.arraycopy(rgData.rinprebuf, num_samples, rgData.rinprebuf,
0, MAX_ORDER - num_samples);
System.arraycopy(left_samples, left_samplesPos, rgData.linprebuf,
MAX_ORDER - num_samples, num_samples);
System.arraycopy(right_samples, right_samplesPos, rgData.rinprebuf,
MAX_ORDER - num_samples, num_samples);
} else {
System.arraycopy(left_samples, left_samplesPos + num_samples
- MAX_ORDER, rgData.linprebuf, 0, MAX_ORDER);
System.arraycopy(right_samples, right_samplesPos + num_samples
- MAX_ORDER, rgData.rinprebuf, 0, MAX_ORDER);
}
return GAIN_ANALYSIS_OK;
};
function analyzeResult(Array, len) {
var i;
var elems = 0;
for (i = 0; i < len; i++)
elems += Array[i];
if (elems == 0)
return GAIN_NOT_ENOUGH_SAMPLES;
var upper = 0 | Math.ceil(elems * (1. - RMS_PERCENTILE));
for (i = len; i-- > 0;) {
if ((upper -= Array[i]) <= 0)
break;
}
//return (float) ((float) PINK_REF - (float) i / (float) STEPS_per_dB);
return (PINK_REF - i / GainAnalysis.STEPS_per_dB);
}
this.GetTitleGain = function (rgData) {
var retval = analyzeResult(rgData.A, rgData.A.length);
for (var i = 0; i < rgData.A.length; i++) {
rgData.B[i] += rgData.A[i];
rgData.A[i] = 0;
}
for (var i = 0; i < MAX_ORDER; i++)
rgData.linprebuf[i] = rgData.lstepbuf[i] = rgData.loutbuf[i] = rgData.rinprebuf[i] = rgData.rstepbuf[i] = rgData.routbuf[i] = 0.;
rgData.totsamp = 0;
rgData.lsum = rgData.rsum = 0.;
return retval;
}
}
module.exports = GainAnalysis;