. .
.
Cepstral Analysis of Speech
.
.

 

Low Time Liftering

 

Write a Scilab program to demonstrate the effects of the cutoff quefrency of the low time lifter on the smoothing of the log-magnitude spectrum. For a frame of 40 ms voiced section of speech, calculate the cepstrum and apply a low time lifter with cutoff quefrencies of 10, 20, 40, 60, 80 and 100 samples, and show the effects of the low time lifter on the smoothed log magnitude spectrum.

 

High Time Liftering

 

Create a series of high time lifters of length L=10, 20, 40, 60, 80 and 100 samples (i.e., For L=10 sample high pass lifter, preserve the cepstral values from n = 10 and above, zeroing out the values between n = -(L-1) to n = (L-1)). For a frame of 40 ms voiced section of speech, calculate the cepstrum and lifter using each of these high time lifters, and plot the resulting log-magnitude spectrum on a common plot.

 

Pitch Estimation

 

Write a Scilab program for cepstral pitch period (F0) detector. Show a plot of F0 (in Hz) vs. time (in seconds).

 

Formant Estimation

 

Write a Scilab program for plotting formant (F1, F2,F3) contours of the given speech signal by cepstral analysis.

 

Cite this Simulator:

.....
..... .....

Copyright @ 2024 Under the NME ICT initiative of MHRD

 Powered by AmritaVirtual Lab Collaborative Platform [ Ver 00.13. ]