Normalized autocorrelation in time domain.

[kostbill]

Hello.

I want to do this, however, I don't understand what normalization is.
Also, I don't quite understand how this will produce the pitch from a voice sample.

Can anyone help me here?

Bill.

 

[bulx]

Normalization is usually done for the sake of comparison with another signal, like scaling a signal so that it has unity power or so.

FT of autcorrelation gives PSD. (Wiener-Khinchin theorem). Pitch is a measure of spectral content, so thats probably the relation
-b

 

[kostbill]

I am sorry I don't follow. Can you suggest a tutorial?

Thanks.