Text to Speech

[Steve Pomeroy <steve@staticfree.info>]

coi rodo,

I'm a lurker here (and in my study of Lojban as well) but I decided to put
a bit of effort into something for the community that I have experience
in. I noticed that there were no text-to-speech voices or engines out there
(besides a rudimentary concatenate-phonemes-together java program). I
hacked up the English voice from Festival (using the Spanish one to help me
along) and put this together:

http://staticfree.info/projects/lojban_festvox/

After listening to it and some recorded Lojban, I've determined that it's
fairly poor quality. (but hey, it's something). What really needs to be
done is the diphones that someone so generously recorded, need to be
timed. (see the bottom of the above page for details) I imagine this could
be done. 

I also have recently learned about implementing programming languages using
yacc and lex, so at some point I might take a stab at samtrosku:

http://www.lojban.org/wiki/index.php/samtrosku

Let me know if anyone is interested in development of either of these
projects, I think they'd be fun endeavors assuming I can find the time for
them :-)

-Xavier X Velcrar
http://staticfree.info/