Hallo, I have just signed up to the news group, but I read to news from time to time whenever I face problem in order to solve my problem. I am thankful to this helpful forum. Now my problem is: I have a for example 65536 or 2^16 in lengthen speech signal. I want to segment the speech signal in the middle iteratively and dyadically i.e. 32768 32768 and then again 16384 16384 and 8192 8192 and then again 4096 4096 and final the last block has 256 and 256, That means now I reach each block is as 256 lengthen signal. Now I want to apply window function to each block which is the same as the listed paper (E. Wesfried and M. V. Wickerhauser, ?Adapted local trigonometric transforms and speech processing?, volume 41 (IEEE Transactions on Signal Processing) (1993) and then I want to apply folding at each edge of the block, then multiply each folded block by DCT-IV.
Here each block will generate a set of spectral coefficients that I want to process further. There is a software wavelab from Stanford, but I never understand how to apply these to the speech signal.
I will appreciate any help to my long existing problem, I can provide more information if I am required.