[hts-users:00432] Re: variablity of sound

["Heiga ZEN (Byung Ha CHUN)" <zen@xxxxxxxxxxxxxxxx>]

Hi

Mohammed Samir wrote:

> Is this mean that this sound is the voice of the
> person who record the training data.

Yes.

> Because I thought that after training, the MLSA filter can generate
> sound with any pitch or (frequency) that doesnt exist
> in the training sounds.

OK, I understand.
In hts voices for festival-1.4x distributed in the HTS website, you can change f0 range by changing configuration variables

hts_f0_mean
hts_f0_std

on festival as follows:

festival> (set! hts_f0_mean 100)
festival> (set! hts_f0_std 2)


The f0 values are modified using these variables as follows:

modified_f0 = hts_f0_std * original_f0 + hts_f0_mean

I don't know whether these variables are supported in the latest festival release :-(

Regards,

Heiga ZEN (Byung Ha CHUN)


--
------------------------------------------------
Heiga ZEN     (in Japanese pronunciation)
Byung Ha CHUN (in Korean pronunciation)

Department of Computer Science and Engineering
Nagoya Institute of Technology
Gokiso-cho, Showa-ku, Nagoya 466-8555 Japan

http://kt-lab.ics.nitech.ac.jp/~zen
------------------------------------------------