Package marytts.signalproc.analysis

A collection of analysis algorithms for signal processing.
Important classes are as follows: LpcAnalyser: Linear prediction analysis using autocorrelation appraoch and Durbin recursion LsfAnalyser: Computation of line spectral frequencies (LSFs, or line spectral pairs - LSPs) based on LpcAnalyser EnergyAnalyser: Energy contour estimation with voice activity detection support F0TrackerAutocorrelationHeuristic: An autocorrelation based f0 analysis algorithm extended with heuristic post-processing to reduce voiced/unvoiced errors and f0 doubling/halving problems.

See: Description

Class Summary

Class	Description
AlignedLabels
AlignedLabels.AlignedTimeStretch
AlignmentData
AudioFileDuration
CepstrumLPCAnalyser
CepstrumSpeechAnalyser
ComplexLinearPredictor	A class for nesting results of complex valued linear prediction
EnergyAnalyser
EnergyAnalyser_dB
EnergyContourRms	A class that extracts frame based root-mean-square (RMS) energy contour
EnergyFileHeader	Implements a structured header with file I/O functionality for binary energy contour files
F0ReaderWriter	File I/O for binary pitch contour files
F0Tracker
F0TrackerAutocorrelationDP
F0TrackerAutocorrelationHeuristic	Autocorrelation based F0 tracker with heuristic rules based on statistics for smoothing and halving/doubling prevention
FeatureFileHeader	Generic file header for binary acoustic feature files
FestivalUtt	A wrapper class to read fields in Festival UTT files
FrameBasedAnalyser<T>
FrameBasedAnalyser.FrameAnalysisResult<T>
Label	A class to keep all information on a single EST format label
Labels	A collection of EST formatted labels with ascii text file input/output functionality
Lpc2Lpcc	Internally does the conversion between LPCs and LPCCs.
Lpc2Lprefc	Internally does the conversion between LPCs and reflection coefficients.
LpcAnalyser
LpcAnalyser.LpCoeffs
Lpcc2Lpc	Internally does the conversion between LPCCs and LPCs.
LpcPitchSynchronousAnalyser	A Pitch-synchronous LPC analyser.
Lprefc2Lpc	Internally performs conversion between reflection coefficients and LPCs.
LsfAnalyser	Demonstration program to accompany the subroutines described in the articles by J.
LsfFileHeader	Implements a structured header with file I/O functionality for binary files that store frame based line spectral frequency vectors
Lsfs	A wrapper class for line spectral frequencies.
MfccFileHeader	Implements a structured header with file I/O functionality for binary files that store frame based mel frequency cepstral coefficient vectors
MfccRaw2MfccConverter	Converts binary MFCC files in raw SPTK format into Mary MFCC file format with header
Mfccs	A wrapper class for frame based mel frequency cepstral coefficient vectors.
PitchFileHeader	Implements a structured header with file I/O functionality for binary files that store frame based f0 values
PitchFrameAnalyser
PitchMarks	A wrapper class to store pitch marks as integer sample indices
PitchReaderWriter	File I/O for binary pitch contour files
PraatScriptGenerator	A simple class to generate a Praat script.
ReflectionCoefficients
RegularizedCepstrumEstimator
RegularizedPostWarpedCepstrumEstimator	Implements a cepstral envelope estimation algorithm for fitting an envelope to discrete frequency amplitudes.
RegularizedPreWarpedCepstrumEstimator	Implements the regularized cepstral envelope estimation in Cappe, O., Laroche, J., and Moulines E., 1995, "Regularized estimation of cepstrum envelope from discrete frequency points", in IEEE ASSP Workshop on app.
SeevocAnalyser	Spectral Enveope Estimation Vocoder (SEEVOC) - a simple implementation Reference: Paul, D., 1981, "The Spectral Envelope Estimation Vocoder", IEEE Trans.
ShortTermAutocorrelationAnalyser
ShortTermCepstrumAnalyser
ShortTermLogSpectrumAnalyser
ShortTermPhaseSpectrumAnalyser
ShortTermSpectrumAnalyser
Signal2EnergyConverter
SpectrumWithPeakIndices
SPTKPitchReaderWriter	File I/O for binary SPTK pitch contour files
VoiceQuality	A wrapper class for frame based voice quality parameters
VoiceQualityFileHeader

Package marytts.signalproc.analysis Description

A collection of analysis algorithms for signal processing.
Important classes are as follows:

• LpcAnalyser: Linear prediction analysis using autocorrelation appraoch and Durbin recursion
• LsfAnalyser: Computation of line spectral frequencies (LSFs, or line spectral pairs - LSPs) based on LpcAnalyser
• EnergyAnalyser: Energy contour estimation with voice activity detection support
• F0TrackerAutocorrelationHeuristic: An autocorrelation based f0 analysis algorithm extended with heuristic post-processing to reduce voiced/unvoiced errors and f0 doubling/halving problems. This tracker works better as compared to Praat but worse as compared to Snack/Wavesurfer. Snack employs RAPT (Robust Algorithm for Pitch Tracking [Talkin, 1995])
• SeevocAnalyser: A basic implementation of the Spectral Envelope Estimation Vocoder which fits a spectral envelope to spectral peaks using peak detection and linear interpolation. This is used by sinusoidal models.
• RegularizedCepstrumEstimator (and classes derived from it): Estimation of a spectral envelope by cepstrum method using spectral amplitudes measured in frequency domain. These methods are used in sinusoidal and harmonics plus noise models

[Talkin, 1995] D. Talkin, "A robust algorithm for pitch tracking (RAPT)," in Speech Coding and Synthesis (W. B. Kleijn and K. K. Paliwal, eds.), ch. 14, Elsevier Science, 1995, pp. 495-518.