mgcep

Documentation

Overview

Package mgcep provides support for Mel-Generalized Cepstrum Analysis.

Index

• func C2IR(ceps []float64, length int) []float64
• func CalcMCepAlpha(sampleRate int) float64
• func Cepstrum2LogAmplitude(ceps []float64) []float64
• func FreqT(ceps []float64, order int, alpha float64) []float64
• func FrequencyWarping(ceps []float64, order int, alpha float64) []float64
• func GC2GC(c1 []float64, gamma1 float64, m2 int, gamma2 float64) []float64
• func GNorm(ceps []float64, gamma float64) []float64
• func IGNorm(normalizedCeps []float64, gamma float64) []float64
• func LogAmp2MCep(logAmp []float64, order int, alpha float64) []float64
• func LogAmplitude2Cepsturm(logAmp []float64) []float64
• func MCep(audioBuffer []float64, order int, alpha float64) []float64
• func MCep2Energy(mc []float64, alpha float64, length int) float64
• func MCep2LogAmp(melCeps []float64, numFreqBins int, alpha float64) []float64
• func MGC2MGC(c1 []float64, alpha1, gamma1 float64, m2 int, alpha2, gamma2 float64) []float64
• func MGCep(audioBuffer []float64, order int, alpha, gamma float64) []float64
• func UELS(audioBuffer []float64, order int) []float64

Functions

func C2IR

func C2IR(ceps []float64, length int) []float64

C2IR performs conversion from minimum phase cepstral coefficients to impule response. Note: a bit differ in original function c2ir. m+1 = nc

func CalcMCepAlpha

func CalcMCepAlpha(sampleRate int) float64

Calculate all-pass constant (alpha) for a given sample frequency.

func Cepstrum2LogAmplitude

func Cepstrum2LogAmplitude(ceps []float64) []float64

Cepstrum2Amplitude return log amplitude spectrum from cepstrum. Note: This function requires symmetric cepstrum as the argument.

func FreqT

func FreqT(ceps []float64, order int, alpha float64) []float64

Alias

func FrequencyWarping

func FrequencyWarping(ceps []float64, order int, alpha float64) []float64

FrequencyWarping returns frequency-warped (mel) cesptrum given a cepstrum. The order is a desired order of the cepstrum and alhpa is an all-pass constant.

func GC2GC

func GC2GC(c1 []float64, gamma1 float64, m2 int, gamma2 float64) []float64

GC2GC peforms Generalized cepstrum transformation.

func GNorm

func GNorm(ceps []float64, gamma float64) []float64

GNorm performs Gamma Normalization to Cepstrum.

func IGNorm

func IGNorm(normalizedCeps []float64, gamma float64) []float64

IGNorm performs inverse Gamma Normalization to Cepstrum.

func LogAmp2MCep

func LogAmp2MCep(logAmp []float64, order int, alpha float64) []float64

LogAmp2MCep returns mel-cepstrum from a given log amplitude spectrum. Note: This function requires symmetric log-amplitude that length correspond to the number of frequency bins in fourier analysis.

func LogAmplitude2Cepsturm

func LogAmplitude2Cepsturm(logAmp []float64) []float64

Amplitude2Cepstrum returns (real) cepstrum from an amplitude spectrum.

func MCep

func MCep(audioBuffer []float64, order int, alpha float64) []float64

func MCep2Energy

func MCep2Energy(mc []float64, alpha float64, length int) float64

func MCep2LogAmp

func MCep2LogAmp(melCeps []float64, numFreqBins int, alpha float64) []float64

MCep2LogAmp returns log amplitude spectrum from a given mel-cepstrum.

func MGC2MGC

func MGC2MGC(c1 []float64, alpha1, gamma1 float64,
	m2 int, alpha2, gamma2 float64) []float64

MGC2MGC performs Mel generalized cepstrum transformation.

func MGCep

func MGCep(audioBuffer []float64,
	order int, alpha, gamma float64) []float64

MGCep performs Mel-Generalized Cepstrum analysis and returns mgcep coefficients. Parameters:

order: order of mel-generalized cepstrum
alpha: frequency warping parameter
gamma: parameter of generalized log function

Return:

mel-generalized cepstrum coefficients (length of order+1)

func UELS

func UELS(audioBuffer []float64, order int) []float64

TODO(ryuichi) replace with pure Go.