General News - XVASynth Let's Modders Extend Bethesda Voice Sets

Some good news for fans of Bethesda modding as xVASynth lets modders extend voice acting of existing actors.

loading…

xVASynth is a machine learning based speech synthesis app, using voices from characters/voice sets from Bethesda games.

GitHub: https://github.com/DanRuta/xVA-Synth/
Patreon: https://www.patreon.com/xvasynth
Discord: https://discord.gg/nv7c6E2TzV
Twitter: @dan_ruta

More information.

It's an interesting idea, but the robotic results don't cut it, IMO. I'd just stick with text rather than that effect.

I wonder if anyone is working on something like this at a higher level - perhaps converting an actual voice performance into someone else's voice. I'd have thought it's at least conceivable, with machine learning and huge data sets.

Really cool. It’s raw sure, but for a mod this is great. I think in 5-15 years we’ll see this tech start to be implemented in actual games, sucks for voice actors sure, but it will allow smaller game company to make more policed games with much greater dialogue.

So wouldn't it be against certain copyrights to use someones voice against their consent? As you use pre-recorded audio from a voice actor and make new lines.