c2acr
Functions
-
int main(int argc, char *argv[])
c2acr [ option ] [ infile ]
-m int
order of cepstral coefficients \((0 \le M_1 < L)\)
-M int
order of autocorrelation coefficients \((0 \le M_2 < L)\)
-l int
FFT length \((2 \le L)\)
infile str
double-type cepstral coefficients
stdout
double-type autocorrelation coefficients
The following example converts the 30-th order cepstral coefficients in
data.cepinto the 15-th order LPC coefficients.c2acr -m 30 -M 15 < data.cep | levdur -m 15 > data.lpc
- Parameters:
argc – [in] Number of arguments.
argv – [in] Argument vector.
- Returns:
0 on success, 1 on failure.
-
class CepstrumToAutocorrelation
Convert cepstral coefficients to autocorrelation coefficients.
The input is the \(M_1\)-th order cepstral coefficients:
\[ \begin{array}{cccc} c(0), & c(1), & \ldots, & c(M_1), \end{array} \]and the output is the \(M_2\)-th order autocorrelation coefficients:\[ \begin{array}{cccc} r(0), & r(1), & \ldots, & r(M_2), \end{array} \]The definition of the cepstrum can be represented as\[ c(m) = \mathcal{F}^{-1} \{ \log |\mathcal{F} \{x(m)\}| \}, \]where \(x(m)\) is a signal, \(\mathcal{F}^{-1}\) and \(\mathcal{F}\) denote the DFT and the inverse DFT, respectively. From the definition, the relation between the cepstrum and the autocorrelation can be derived as follows:\[\begin{split}\begin{eqnarray} c(m) &=& \mathcal{F}^{-1} \left\{ \frac{1}{2} \log|X(\omega)|^2 \right\} \\ &=& \mathcal{F}^{-1} \left\{ \frac{1}{2} \log \mathcal{F} \{r(m)\} \right\}, \end{eqnarray}\end{split}\]where the Wiener–Khinchin theorem is used. Thus\[ r(m) = \mathcal{F}^{-1} \{ \exp (2\mathcal{F} \{c(m)\}) \}. \]Note that the imaginary part is zero.Public Functions
-
CepstrumToAutocorrelation(int num_input_order, int num_output_order, int fft_length)
- Parameters:
num_input_order – [in] Order of cepstral coefficients, \(M_1\).
num_output_order – [in] Order of autocorrelation coefficients, \(M_2\).
fft_length – [in] FFT length.
-
inline int GetNumInputOrder() const
- Returns:
Order of cepstral coefficients.
-
inline int GetNumOutputOrder() const
- Returns:
Order of autocorrelation coefficients.
-
inline int GetFftLength() const
- Returns:
FFT length.
-
inline bool IsValid() const
- Returns:
True if this object is valid.
-
bool Run(const std::vector<double> &cepstrum, std::vector<double> *autocorrelation, CepstrumToAutocorrelation::Buffer *buffer) const
- Parameters:
cepstrum – [in] \(M_1\)-th order cesptral coefficients.
autocorrelation – [out] \(M_2\)-th order autocorrelation coefficients.
buffer – [out] Buffer.
- Returns:
True on success, false on failure.
-
class Buffer
Buffer for CepstrumToAutocorrelation class.
-
CepstrumToAutocorrelation(int num_input_order, int num_output_order, int fft_length)