Bug#495583: RFP: hts -- HMM-based Speech Synthesis System

["Trent W. Buck" <twb@cybersource.com.au>]

Juhapekka Tolvanen wrote:
> Package: wnpp
> Severity: wishlist
>
> * Package name    : hts
> * URL             : http://hts.sp.nitech.ac.jp/
> * License         : BSD
>   Description     : HMM-based Speech Synthesis System
>
> I made a Google search in order to find some DFSG-free speech
> synthesis package that could speak Japanese language. This is one of
> my findings.

I would also like these voices to be packaged for Debian.

Upstream also provides six American English voices[1] in a compressed
format compatible with Festival 1.6.  IMO they sound far better than
the MBROLA and festvox English voices.

At [2] you can find instructions for installing these voices by hand,
on Ubuntu.  I successfully followed these instructions to get the
English voices on Debian, but I'd prefer proper packages.

[1] http://hts.sp.nitech.ac.jp/?Download
[2] http://ubuntuforums.org/showthread.php?t=751169