Welcome to the world of Scilab Speech

Function List--Time Analysis

1．auto

2．energy

3．zcr

1．auto----autocorrelation function for the windowed data.

Calling Sequence:

　　auto(wavefile);

　　auto(wavefile, window_type);

Parameters:

　　wavefile: wave file. Vector or matrix.

　　window_type: desired window type. Interger.

Description:

This function is the biased autocorrelation function for the speech signal that has been windowed. It is useful for estimating the fundamental frequency of voicing for voiced speech segments. The default value of window_type is hamming.

Examples:

x=wavread('d:\test.wav');

y=auto(x);

y=auto(x, 4);

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　back to top

2．energy----calculate the short-time average energy of a speech signal.

Calling Sequence:

　　energy(wavefile);

　　energy(wavefile, window_type);

Parameters:

　　wavefile: wave file. Vector or matrix.

　　window_type: desired window type. Interger.

Description:

The energy function calculates the short-time average energy of a speech signal. It is useful for estimating word and phoneme boundaries.

Examples:

x=wavread('d:\test.wav');

y=energy(x);

y=energy(x, 4);

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　back to top

3．zcr----calculate the short-time average zero-crossing rate of a speech signal.

Calling Sequence:

　　zcr(wavefile);

　　zcr(wavefile, window_type);

Parameters:

　　wavefile: wave file. Vector or matrix.

　　window_type: desired window type. Interger.

Description:

This function calculates the short-time average zero-crossing rate which is used to estimate the frequency of voicing. It is also useful for estimating word and phoneme boundaries.

Examples:

x=wavread('d:\test.wav');

y=zcr(x);

y=zcr(x, 4);

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　back to top