Speex dev - Sep 2006 - Speaker/Language-etc dependency of encoded data

Hello,

I noticed that for one specific Speaker, there are codebook entries in
all codebooks, that "fit" the speaker.
So if one had a look at a histogram of the used codebook line numbers
for one speaker, the histograms would look very much the same for
different speech samples (of course, the speech samples should be long
enough, more than a minute of speech ought to be sufficient).

I suppose that this has something to do with the voice of the speaker,
so the histogramm shape ist specific for one speaker.

I do not know if factors like the spoken language, tha fact if the
language is sung or not, etc have an influence too.

I have not tested this yet, either. I'll soon produce some test data
myself, at least do the singing and speak english and german.

Can you imagine factors that possibly influence the histogram of the
chosen codebook entries besides the voice of the speaker and the
language? Which of these factors do you think are worth examining what
their influence is?

Ciao,

Bj?rn


-- 
Good day for overcoming obstacles.  Try a steeplechase.

-- 
Important! Please recognize my new GPG Public Key!
                 Bj?rn Thalheim
gpg fingerprint: 2F22 AAEB 1818 1548 EC78  1AE8 9D2E FCB4 0980 28CC
   download key: wget http://www.ifsr.de/~bjoern/gpg/public_key.asc
       See also: http://www.ifsr.de/~bjoern/gpg/key.html

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
Url :
http://lists.xiph.org/pipermail/speex-dev/attachments/20060911/b5436f11/signature.pgp