Re: Making a Vintage Sounding TTS Voice
I just read up on the grapheme based language models in Festvox and they seem awesome when you are doing a language for which you don't have a lot of linguistic data. But in the case of Swedish, if I ever get far enough where I begin looking at making a Swedish model I think I will do it in the slower way, with letter to sound rules trained from a large lexicon. I found a huge lexicon for Swedish which contains over 700,000 words and is in the public domain, so I figured that might be a good starting point. Also, Swedish is my native language so mapping the formants for the various sounds shouldn't be too difficult. Actually, I'm rather looking forward to it so I hope I can get the actual synthesis part up to scratch. If I can, adding new languages and voices will be a blast.
Kind regards,
Philip Bennefall
-- Audiogames-reflector mailing list Audiogames-reflector@sabahattin-gucukoglu.com https://sabahattin-gucukoglu.com/cgi-bin/mailman/listinfo/audiogames-reflector
