Elevenlabs.io - AI Generate anyone's voice with a minute of audio saying how much they hate trannies

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
So I caved and spent the money for an account after hearing positive feedback from a friend. I really wanted to test this with the English voice actor for Noctis from Final Fantasy XV, he's got a bit of a flat voice like Dagoth Ur and I thought his voice would translate well to AI. End result? Not bad.

Here's Noctis doing the Final Fantasy XIV Free Trial copypasta:



I was curious about the feasibility of AI taking over for shitposting, because I stumbled on a tweet where someone asked the character's voice actor to record the "women fear me, fish fear me" copypasta at a convention.

video source

And here's me having the AI replicate that same quote:


Noctis doing the "Shadow the hedgehog pissed on my wife" rant:


Noctis doing the "WHAT A SHIT WEBSITE" quote from the recently halal'd Shiloh "Oakly" Sellers thread:


Noctis doing the OrsonZedd "ontologically evil" comment in the Keffals thread:


EDIT: Here's Noctis expressing his urge to destroy all beans:
 
Last edited:
THIS TIME. THIS TIME. THIS TIME.
1677885399177.png
My problem downloading whisper and autoregressive is literally that the repo sites were down. No fucking wonder it wouldn't download.
Edit: Three more times and bingo
1677892663370.png
I think it's working because my CPU is pegged at 50% and the fans are going and the VRAM is all used up. Just loading in a voice to use, so not sure if I should be expecting GPU usage yet, but there is none.
Edit: Edit: I have coherent output. Fucking finally.
 
Last edited:
Has anyone tried putting a cat's meowing noises into this thing? I'm dying to see if that works and what the result sounds like.
What do you want it to try and say? I assume it would just turn into a default male/female voice but worth a try.
 
  • Like
Reactions: MrJokerRager
You know, this actually opens the door for some really creative applications for non-human characters in video games.

Like imagine having a water elemental npc. You can have some monotone girl's voice for it, or you can use various water ambiences to generate audio. Like:
  • A steady river ambience for her normal voice
  • Rain for a sad dialogue
  • A raging storm or crashing ocean waves for her angry dialogue.
 
For some reason female voices aren't working for me but males are. Do you need lots of samples cause I heard it doesn't even use them just the first one uploaded.
 
  • Like
Reactions: MrJokerRager
For some reason female voices aren't working for me but males are. Do you need lots of samples cause I heard it doesn't even use them just the first one uploaded.

Tested by giving Tortoise 30 seconds of source material. It's a bit funky, but seems to be working alright.

Then again, it did this horrifying shit when I fed it Ronald Reagan's voice:


So at this point I have no fucking idea wtf I'm doing. How much sample data are you feeding it?
 
Back