[MUD-Dev2] [Technology] Lend your ears to science!

[Mike Rozak]

Benjamin Tolputt wrote:
> Actually, any details as to the use of TTS in games (&/or the direction
> of companies where this might be possible) would be appreciated. Whether
> it belongs on this list is debatable, but I know I would be interested
> in hearing more...

The biggest hurdle to using TTS in your game is that you need gameplay that 
requires TTS. If your gameplay DOESN'T require TTS, you'll find that TTS 
quality won't ever be good enough, not matter how good it gets. Basically, 
your game has to be about talking (to NPCs).

TTS is an information channel that competes against the text information 
channel. If the only information supplied by the TTS voice is the same as 
the text that's spoken, then players might as well read the text (assuming 
their eyes aren't busy elsewhere). However, if TTS includes prosody that 
changes/enhances the meaning of the sentence, then it's more relevent. How 
something is said is often more important than what is said.

The next biggest hurdle is that your gameplay needs to work around the 
current limitations of TTS, just as 3D games work (and worked around) the 
limitations of 3D graphics. In the 90's, most 3D games were based on space 
combat, flight simulators, or in mazelike buildings, because those settings 
didn't require many polygons or textures. Likewise, contemporary TTS isn't 
very good at emotional speech, so conversations with NPCs will have to take 
place in calm situations. Drunk NPCs are okay. Crying NPCs can't be done.

Then, you run into all sorts of technical and licensing problems, which are 
minor issues compared to the above issues.