public static void main(String[] args) throws UnsupportedAudioFileException, IOException {
AudioInputStream inputAudio = AudioSystem.getAudioInputStream(new File(args[0]));
int samplingRate = (int) inputAudio.getFormat().getSampleRate();
AudioDoubleDataSource signal = new AudioDoubleDataSource(inputAudio);
double[] x = signal.getAllData();
int i;
int numBands = 4;
double overlapAround1000Hz = 100.0;
FIRBandPassFilterBankAnalyser analyser =
new FIRBandPassFilterBankAnalyser(numBands, samplingRate, overlapAround1000Hz);
Subband[] subbands = analyser.apply(x);
DDSAudioInputStream outputAudio;
AudioFormat outputFormat;
String outFileName;
// Write highpass components 0 to numLevels-1
for (i = 0; i < subbands.length; i++) {
outputFormat =
new AudioFormat(
(int) (subbands[i].samplingRate),
inputAudio.getFormat().getSampleSizeInBits(),
inputAudio.getFormat().getChannels(),
true,
true);
outputAudio =
new DDSAudioInputStream(new BufferedDoubleDataSource(subbands[i].waveform), outputFormat);
outFileName =
args[0].substring(0, args[0].length() - 4) + "_band" + String.valueOf(i + 1) + ".wav";
AudioSystem.write(outputAudio, AudioFileFormat.Type.WAVE, new File(outFileName));
}
}
/**
* Command line interface to the vocal tract linear scaler effect.
*
* @param args the command line arguments. Exactly two arguments are expected: (1) the factor by
* which to scale the vocal tract (between 0.25 = very long and 4.0 = very short vocal tract);
* (2) the filename of the wav file to modify. Will produce a file basename_factor.wav, where
* basename is the filename without the extension.
* @throws Exception if processing fails for some reason.
*/
public static void main(String[] args) throws Exception {
if (args.length != 2) {
System.err.println(
"Usage: java " + VocalTractLinearScalerEffect.class.getName() + " <factor> <filename>");
System.exit(1);
}
float factor = Float.parseFloat(args[0]);
String filename = args[1];
AudioDoubleDataSource input =
new AudioDoubleDataSource(AudioSystem.getAudioInputStream(new File(filename)));
AudioFormat format = input.getAudioFormat();
VocalTractLinearScalerEffect effect =
new VocalTractLinearScalerEffect((int) format.getSampleRate());
DoubleDataSource output = effect.apply(input, "amount:" + factor);
DDSAudioInputStream audioOut = new DDSAudioInputStream(output, format);
String outFilename = FilenameUtils.removeExtension(filename) + "_" + factor + ".wav";
AudioSystem.write(audioOut, AudioFileFormat.Type.WAVE, new File(outFilename));
System.out.println("Created file " + outFilename);
}
public static void main(String[] args) throws UnsupportedAudioFileException, IOException {
// File input
AudioInputStream inputAudio = AudioSystem.getAudioInputStream(new File(args[0]));
int samplingRate = (int) inputAudio.getFormat().getSampleRate();
AudioDoubleDataSource signal = new AudioDoubleDataSource(inputAudio);
double[] x = signal.getAllData();
double maxOrig = MathUtils.getAbsMax(x);
SinusoidalAnalyzer sa = null;
SinusoidalTracks st = null;
PitchSynchronousSinusoidalAnalyzer pa = null;
//
// Analysis
float deltaInHz = SinusoidalAnalysisParams.DEFAULT_DELTA_IN_HZ;
float numPeriods = PitchSynchronousSinusoidalAnalyzer.DEFAULT_ANALYSIS_PERIODS;
boolean isSilentSynthesis = false;
int windowType = Window.HANNING;
boolean bRefinePeakEstimatesParabola = false;
boolean bRefinePeakEstimatesBias = false;
boolean bSpectralReassignment = false;
boolean bAdjustNeighFreqDependent = false;
// int spectralEnvelopeType = SinusoidalAnalysisParams.LP_SPEC;
int spectralEnvelopeType = SinusoidalAnalysisParams.SEEVOC_SPEC;
float[] initialPeakLocationsInHz = null;
initialPeakLocationsInHz = new float[1];
for (int i = 0; i < 1; i++) initialPeakLocationsInHz[i] = (i + 1) * 350.0f;
boolean isFixedRateAnalysis = false;
boolean isRealSpeech = true;
double startFreqInHz = 0.0;
double endFreqInHz = 0.5 * samplingRate;
SinusoidalAnalysisParams params =
new SinusoidalAnalysisParams(
samplingRate,
startFreqInHz,
endFreqInHz,
windowType,
bRefinePeakEstimatesParabola,
bRefinePeakEstimatesBias,
bSpectralReassignment,
bAdjustNeighFreqDependent);
if (isFixedRateAnalysis) {
// Fixed window size and skip rate analysis
double[] f0s = null;
float ws_f0 = -1.0f;
float ss_f0 = -1.0f;
sa = new SinusoidalAnalyzer(params);
if (spectralEnvelopeType == SinusoidalAnalysisParams.SEEVOC_SPEC) // Pitch info needed
{
String strPitchFile = args[0].substring(0, args[0].length() - 4) + ".ptc";
PitchReaderWriter f0 = new PitchReaderWriter(strPitchFile);
f0s = f0.contour;
ws_f0 = (float) f0.header.windowSizeInSeconds;
ss_f0 = (float) f0.header.skipSizeInSeconds;
}
st =
sa.analyzeFixedRate(
x, 0.020f, 0.010f, deltaInHz, spectralEnvelopeType, f0s, ws_f0, ss_f0);
//
} else {
// Pitch synchronous analysis
String strPitchFile = args[0].substring(0, args[0].length() - 4) + ".ptc";
PitchReaderWriter f0 = new PitchReaderWriter(strPitchFile);
int pitchMarkOffset = 0;
PitchMarks pm =
SignalProcUtils.pitchContour2pitchMarks(
f0.contour,
samplingRate,
x.length,
f0.header.windowSizeInSeconds,
f0.header.skipSizeInSeconds,
true,
pitchMarkOffset);
pa = new PitchSynchronousSinusoidalAnalyzer(params);
st =
pa.analyzePitchSynchronous(
x, pm, numPeriods, -1.0f, deltaInHz, spectralEnvelopeType, initialPeakLocationsInHz);
isSilentSynthesis = false;
}
//
// Resynthesis
PeakMatchedSinusoidalSynthesizer ss = new PeakMatchedSinusoidalSynthesizer(samplingRate);
x = ss.synthesize(st, isSilentSynthesis);
//
// File output
DDSAudioInputStream outputAudio =
new DDSAudioInputStream(new BufferedDoubleDataSource(x), inputAudio.getFormat());
String outFileName =
args[0].substring(0, args[0].length() - 4) + "_sinResynthFullbandPitchSynch.wav";
AudioSystem.write(outputAudio, AudioFileFormat.Type.WAVE, new File(outFileName));
//
}