ikkonoishi's Journal: Searching for SHODAN

I'm currently looking for some free (beer and/or speech) programs for TTS that might be useful for making something that sounds like SHODAN of system shock fame.

If anyone has any experience with this sort of thing please comment because I do not have clue zero about this stuff.

Update: I have looked into source forge TTS systems, and have found a few things that look promising, but they all seem to use precompiled voice modules so they can't be modded easilly to do what I want. If I could find the proper voices it should be possible to play them in sequence and simultaneously in order to produce the various stops, starts, and tone shifts that SHODAN uses.

festival...

[Cryptnotic]

Try festival... see if it sounds like what you want it to sound like. It sounds very computery and unnatural. I'm not sure if that's what you want or not.

http://www.cstr.ed.ac.uk/projects/festival/ [ed.ac.uk]

Free, open source, not as good as AT&T.

Tough one...

[kerrle]

I know it seems like making a Shodan-like voice would be easy - and it probably would be easier than trying to make something completely human-sounding - but I don't think that a program would be able to accurately time the weird echos (and reverse-speaking) so that it was aesthetically right.

Making something that sounds like Morpheus from Deus Ex I think is doable - but Shodan would require a ton of variation and some pretty complex rules on how samples were put together.