• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Speech synthesis in games, what's happened to it?

Three

Member
Are there any games with it?
I know it still sounds somewhat robotic so probably doesn't make sense for most games but considering a lot of games actually have robots in them why don't we have this?

Portal, Horizon forbidden west for AI, and Detriot for example. Portal would have been ideal. When do you see speech synthesis being so convincing that we might get much smaller game downloads, AI responses and being able to hear our name in a game? I feel there has been no real push for this in the games industry.
 
Last edited:
edited: easier and quicker to have a person emote than ensure that the programming conveys the right information.
 
Last edited:
edited: easier and quicker to have a person emote than ensure that the programming conveys the right information.
In terms of AI responses that's true but we don't have speech synthesis for typed responses either which is a shame.
 
The only game I can think of that still actively uses it is the ARMA series.
And, without question, it is absolute shit.
I can see how it would be very useful in the future. From what we've seen thus far, though, it's so bad that it actually makes the game more difficult to understand. Just plain text would be better than the options currently available.


I'm curious if any games actually use it in a positive way.
 
Check the state of vocodes like the ones on http://fakeyou.com/
Yeah, none of them are a good substitute for an actual voice actor.
You're right, nothing will beat a voice actor but there are benefits to voice synthesis too like game size and dynamic responses (like NPCs referring to you by name) and it would make perfect sense for games with robots. There are ways to capture a voice actors performance and still have voice synthesis too. For example imagine a single recoring for NPC responses but different characters sound different saying the same thing (as you often get in games) . Respeecher for example does things like this. Voice cloning, listen to Obamas voice created from another recording. It's pretty convincing.

vocodes is only as good as its source material from which it recreates the voice so depending on what's out there of the original to recreate from you will get different results. A lot of the text to speech converters sound pretty natural though.
The only game I can think of that still actively uses it is the ARMA series.
And, without question, it is absolute shit.
I can see how it would be very useful in the future. From what we've seen thus far, though, it's so bad that it actually makes the game more difficult to understand. Just plain text would be better than the options currently available.


I'm curious if any games actually use it in a positive way.
I haven't heard it in ARMA. I guess we are a fair bit out for it to be adopted in games. Curious to see if it ever happens.
 
Last edited:
I just try it and, honestly,I don't think I can tell if the game manufacturer is using this feature in some everyday conversation
That's a fair point. They could actually be using it already and it's so good that I wouldn't even know if they are or not.
 
Just thought of another application too. Imagine this could be done in real time on a console.



Now imagine you are playing an online coop game where people sound like their character instead. The player playing Joel can talk and sounds like Joel talking with spatial accuracy, the person playing Ellie would sound like Ellie.
 
Last edited:
An earlier example could be how they used some of those for the vocals in the SMT Nocturne OST.
 
Audio footprints are so relatively small compared to general data-sizes that voice synthesis is mostly a redundant optimization in today's market.
 
tomodachi life virtual boy GIF
Tomodachi Life Nintendo GIF
Shocked Video Games GIF

Tomadachi Life. I like how that last one is clearly Bill Trinnen and Shigeru Miyamoto.
Shigeru Miyamoto Thumbs Up GIF by Gaming GIFs
 
Last edited:
There won't be long until the tech is advanced enough and so easy to use it'll become usable for game development.
Sure, a voice actor might do a better job, but that's a chunk out of the budget. Especially for Indie devs.

I could imagine future writers adding what 15.ai calls 'emotional contextualizers' (similar to stage direction) to the script and AI figuring out the rest with minimal fine tuning required, or game studios using one talented voice actor and voice conversion.
 
Top Bottom