acorr

Functions

int main(int argc, char *argv[])

acorr [ option ] [ infile ]

-l int
- frame length \((1 \le L)\)
-m int
- order of autocorrelation coefficients \((0 \le M < L)\)
-q int
- input format
  - 0 amplitude spectrum in dB
  - 1 log amplitude spectrum
  - 2 amplitude spectrum
  - 3 power spectrum
  - 4 windowed waveform
-o int
- output format
  - 0 autocorrelation
  - 1 normalized autocorrelation
  - 2 biased autocorrelation
  - 3 unbiased autocorrelation
infile str
- double-type data sequence
stdout
- double-type autocorrelation sequence.

If -o 1, output the normalized autocorrelation:

\[ \begin{array}{cccc} 1, & r(1)/r(0), & \ldots, & r(M)/r(0), \end{array} \]

where \(r(m)\) is the \(m\)-th autocorrelation coefficient. If -o 2, output the biased autocorrelation function:

\[ \begin{array}{cccc} r(0)/L, & r(1)/L, & \ldots, & r(M)/L. \end{array} \]

If -o 3, output the unbiased autocorrelation function:

\[ \begin{array}{cccc} r(0)/L, & r(1)/(L-1), & \ldots, & r(M)/(L-M). \end{array} \]

The below example extracts 10-th order autocorrelation coefficients from windowed waveform.

x2x +sd data.short | frame | window | acorr -m 10 > data.acr

Parameters:

argc – [in] Number of arguments.
argv – [in] Argument vector.

Returns:

0 on success, 1 on failure.

See also

c2acr levdur lpc

class AutocorrelationAnalysis

Calculate autocorrelation from waveform or spectrum.

Public Functions

AutocorrelationAnalysis(int frame_length, int num_order, bool waveform_input)

Parameters:

frame_length – [in] Frame length or number of FFT bins, \(L\).
num_order – [in] Order of autocorrelation, \(M\).
waveform_input – [in] If true, assume waveform input.

inline bool IsValid() const

Returns:: True if this object is valid.

bool Run(const std::vector<double> &input, std::vector<double> *autocorrelation, AutocorrelationAnalysis::Buffer *buffer) const

Parameters:

input – [in] \(L\)-length waveform or \((L/2+1)\)-length power spectrum.
autocorrelation – [out] \(M\)-th order autocorrelation coefficients.
buffer – [out] Buffer.

Returns:

True on success, false on failure.

class Buffer: Buffer for AutocorrelationAnalysis class.

class WaveformToAutocorrelation

Calculate autocorrelation.

The input is the framed waveform signal:

\[ \begin{array}{cccc} x(0), & x(1), & \ldots, & x(L-1), \end{array} \]

where \(L\) is the frame length. The output is the \(M\)-th order autocorrelation coefficients:

\[ \begin{array}{cccc} r(0), & r(1), & \ldots, & r(M). \end{array} \]

The autocorrelation is given by

\[ r(m) = \sum_{l=0}^{L-1-m} x(l)x(l+m), \]

where \(m\) is the lag.

Public Functions

WaveformToAutocorrelation(int frame_length, int num_order)

Parameters:

frame_length – [in] Frame length, \(L\).
num_order – [in] Order of autocorrelation, \(M\).

inline int GetFrameLength() const

Returns:: Frame length.

inline int GetNumOrder() const

Returns:: Order of autocorrelation.

inline bool IsValid() const

Returns:: True if this object is valid.

bool Run(const std::vector<double> &waveform, std::vector<double> *autocorrelation) const

Parameters:

waveform – [in] \(L\)-length framed waveform.
autocorrelation – [out] \(M\)-th order autocorrelation coefficients.

class SpectrumToAutocorrelation

Calculate autocorrelation from power spectrum.

The input is the half of power spectrum:

\[ \begin{array}{cccc} |X(0)|^2, & |X(1)|^2, & \ldots, & |X(L/2)|^2, \end{array} \]

where \(L\) is the FFT length. The output is the \(M\)-th order autocorrelation coefficients:

\[ \begin{array}{cccc} r(0), & r(1), & \ldots, & r(M). \end{array} \]

The autocorrelation is given by

\[ r(m) = \frac{1}{L} \sum_{l=0}^{L-1} |X(l)|^2 e^{j \frac{2\pi}{L} lm}, \]

where \(m\) is the lag.

Public Functions

SpectrumToAutocorrelation(int fft_length, int num_order)

Parameters:

fft_length – [in] Number of FFT bins, \(N\).
num_order – [in] Order of autocorrelation, \(M\).

inline int GetFftLength() const

Returns:: FFT length.

inline int GetNumOrder() const

Returns:: Order of autocorrelation.

inline bool IsValid() const

Returns:: True if this object is valid.

bool Run(const std::vector<double> &power_spectrum, std::vector<double> *autocorrelation, SpectrumToAutocorrelation::Buffer *buffer) const

Parameters:

power_spectrum – [in] \((L/2+1)\)-length power spectrum.
autocorrelation – [out] \(M\)-th order autocorrelation coefficients.
buffer – [out] Buffer.

Returns:

True on success, false on failure.

class Buffer: Buffer for SpectrumToAutocorrelation class.