Imogen Heap Instagram – Some of you may have noticed the voice in the @korus_ai remix platform video sounds familiar!
This is an AI voice model of mine created with @jenmusicai (part of ai.mogen character and functions) to explore scenarios together for permissions on AI voice.
It went like this:
1. I have a verified Auracle on @auracles.io as a music maker.
2. I created a recording of my voice via Auracle Voice (in dev) that was then verified and checked by the system as my own.
Now I have a home for my voice with info on what the voice allows (even if that’s nothing at all!). There is a folder with verified recordings of my voice should anybody or service want to use it for training.
3. Jen requested permission to make an AI Voice model and I accepted.
4. I’ve used this voice on my soon to be released last piece of this 3 part piece (after WHYDTM and Noise).
5. KORUS wanted to use an AI voice for their promo video.
So this is Ai.Mogen (the ai voice portion of Mogen) modelled by Jen, permissioned by me, via my Voice Auracle.
In time, I may well get it out there for anyone to have a play about with in general. Baby steps.
6. So… continuing on…KORUS had initially used an American accent character off the shelf AI voice (with text approved by me) that they sent to me as an audio file.
7. I ran the audio through my Jen voice model so the timbre of my voice transfers over the original (it’s a bit like, changing the Midi instrument sound of a keyboard part once somebody has played it).
8. Then I sent it back to Inder (CEO of Korus).
So that was the journey!
Sounds long but with permissions and all pieces in place this can automated and done in no time at all.
also… Can’t wait for you to play about with this thing that Jen have created called style filter! But more on that later.
Here’s some video of Jen Music CEO @shara.senderoff and I last month at @websummit | Posted on 06/Dec/2024 23:34:19
