Class Summary 
AreaMoments 
This class implements 2D statistical methods of moments as implemented by
Fujinaga (1997). 
AreaMomentsConstantQMFCC 
This class implements 2D statistical methods of moments as implemented by
Fujinaga (1997). 
AreaMomentsLogConstantQ 
This class implements 2D statistical methods of moments as implemented by
Fujinaga (1997). 
AreaMomentsMFCC 
This class implements 2D statistical methods of moments as implemented by
Fujinaga (1997). 
AreaPolynomialApproximation 
2D Polynomial Approximation Feature
Creates a set of polynomial factors for a 2D polynomial of order k*l where k is the number of
terms in the x direction and l is the number of terms in the y direction. 
AreaPolynomialApproximationConstantQMFCC 
2D Polynomial Approximation Feature
Creates a set of polynomial factors for a 2D polynomial of order k*l where k is the number of
terms in the x direction and l is the number of terms in the y direction. 
AreaPolynomialApproximationLogConstantQ 
2D Polynomial Approximation Feature
Creates a set of polynomial factors for a 2D polynomial of order k*l where k is the number of
terms in the x direction and l is the number of terms in the y direction. 
BeatHistogram 
A feature extractor that extracts the Beat Histogram from a signal. 
BeatHistogramLabels 
A "feature extractor" that calculates the bin labels, in beats per minute, of
a beat histogram. 
BeatSum 
A feature extractor that extracts the Beat Sum from a signal. 
Compactness 
A feature extractor that extracts the Compactness. 
ConstantQ 

ConstantQMFCC 

Derivative 
Class implementing the most basic discrete derivative of a dependant feature. 
FeatureExtractor 
The prototype for feature extractors. 
FFTBinFrequencies 
A "feature extractor" that calculates the bin labels, in Hz, of power spectrum
or magnitude spectrum bins that would be produced by the FFT of a window of the
size of that provided to the feature extractor. 
FractionOfLowEnergyWindows 
A feature extractor that extracts the Fraction Of Low Energy Windows from
window to window. 
HarmonicSpectralCentroid 
A variation on spectral centroid that is based upon peaks instead of bins. 
HarmonicSpectralFlux 
A variation on spectral flux that is based upon peaks instead of bins. 
HarmonicSpectralSmoothness 
A peak based calculation of smoothness. 
LogConstantQ 
Log Constant Q
Performs the log linear transform of the bins of the constant q transform to produce a representation whose linear
content better represents how the human ear hears differences in amplitude. 
LPC 
Calculates linear predictive coeffecients of an signal. 
LPCRemoved 
Linear Prediction Coeffecients calculated according to 'Numerical Recipes in C' (Press at al. 
MagnitudeSpectrum 
A feature extractor that extracts the FFT magnitude spectrum from a set of
samples. 
Mean 
Calculates the running mean of an underlying feature. 
MetaFeatureFactory 
This class combines 2 sets of functionality:
Factory for generating instances of a particular metafeature
Instance of this particular metafeature
This class is used in the following manner:
Create an instance of the class
Set dependant MetaFeatureFactory class
Cycle over features, generating 1 instance per feature

MFCC 
Utilizes the MFCC code from the OrangeCow Volume project. 
Moments 
Class that calculates the first 5 stastical method of moments (Fujinaga
1997). 
PeakFinder 
Implements a very basic peak detection algorithm. 
PowerSpectrum 
A feature extractor that extracts the FFT power spectrum from a set of
samples. 
RelativeDifferenceFunction 
This feature calculates the log of the derivative of the RMS. 
RMS 
A feature extractor that extracts the Root Mean Square (RMS) from a set of
samples. 
SpectralCentroid 
A feature extractor that extracts the Spectral Cecntroid. 
SpectralFlux 
A feature extractor that extracts the Spectral Flux from a window of samples and
the preceeding window. 
SpectralRolloffPoint 
A feature extractor that extracts the Spectral Rolloff Point. 
SpectralVariability 
A feature extractor that extracts the SpectralVariance. 
StandardDeviation 
Calculates the Standard Deviation of a feature over a large running window. 
StrengthOfStrongestBeat 
A feature extractor that extracts the Strength of Strongest Beat from
a signal. 
StrongestBeat 
A feature extractor that finds the strongest beat in a signal, int
beats per minute. 
StrongestFrequencyVariability 
A feature extractor that extracts the Strongest Frequency Variability
from window to window. 
StrongestFrequencyViaFFTMax 
A feature extractor that finds the strongest frequency component of a
signal, in Hz. 
StrongestFrequencyViaSpectralCentroid 
A feature extractor that finds the strongest frequency in Hz in a signal
by looking at the spectral centroid. 
StrongestFrequencyViaZeroCrossings 
A feature extractor that finds the strongest frequency in Hz in a signal
by looking at the zero crossings. 
ZeroCrossings 
A feature extractor that extracts the Zero Crossings from a set of
samples. 