just curios; how''s speech synthesis-technology these days? can we use it in games any time soon?

hmmnoo. I wasn''t really concerned with the technical approach, but more a sort of wildly speculative look into the future.

I know there''s a company here in Norway that can "render" close to natural sounding speech, even with dialects and accents. I don''t know what the "source" looks like, but I concider it a fair guess that it is some sort of phoneme-system (like the old SoundOLE thing from Creative Labs). Nor do I know the processing requirements.

My point is that one day we should be able to do this in real time, and fast enough that we can ditch voice-actors in games. I just wondered if anyone had any knowledge of this whatsoever (unlike me), and could make a bet as to how far into the future we''re talking.

Oh well. I might be tripping, and it might never happen for all I know, but it would definitely be way cool.

There are a quite a few text-to-speech programs around. Here''s my guess about how it works.
They have a dictionary which translated known words to phonetics. If a word isn''t in the dictionary the pronunciation is guessed. If you''ve heard the Stephen Hawking (quadruplegic scientist) then you''ve heard one of these programs.
The better programs will let you change the voice characteristics and pitch, so you can make the program sing a song.
To do this when creating a game, you''d have to find some very technical info the internet or know a lot about sound & speech, as I doubt that the makers of any of the top programs would give away many ideas. You''d need a dictionary (words -> phonetics) for whatever language/accent you wanted to use. Each character would have a different set of voice characteristics. Even then it would sound natural. Voice actors aren''t likely to be replaced yet.