There is a tutorial for that on my channel, however the user Vul has made a pretty good one as well. This includes synthesis ai voices with existing models as well as training new ai models. The most important info is that this can operate offline (with nvida gpu) or on colab, and you will need about 5 minutes worth of audio dataset cut in 1~10seconds wav files to train the model well.
![[AI Voice Cover] Fluttershy sings Killing In The Name by Rage Against The Machine](http://i.ytimg.com/vi/dR3iq7Kbo98/mqdefault.jpg)
The bad timeline where instead of pursuing reformation the Mane 6 opted to simply enslave Discord
lol, treu
u mean the good timeline
And turn him into a hand puppet
Au where Twilight used that spell that pretty much removes Discords free will after all
If Discord comes through the portal to the Equestria Girls world, he well be just Bo Burnham. Also works the other way around!
bo burnham th3 draconequus
but wouldnt he appear as jhon DeLancy or his other character "Q"?
Yes…what?
Look at me!
Twilight has never sounded so scary!
The voice isn't quite as accurate as Fluttershy singing American Idiot, but still very entertaining
The discord is spot on tho right?!
I would love to see discord sing welcome to the internet
Discord and Celestia or Twilight or even Fluttershy singing “The internet is for pron”
@@noname-lp3mu i agree with that too but make no mistake welcome the internet by bo burnham suits discords personality very well
ruclips.net/video/k1BneeJTDcU/видео.html
YES! And with minor thematic lyric changes like Alastair's version. It would be perfect.
Funny thing, someone DID have discord sing it WITH fluttershy and it worked.
Chrysalis singing “The government can”
THAT DISCORD IS PERFECT
Wow the world sounds pretty neat
I wish the world was real that would be so cool
This is digustingly underrated
This is great! Thanks for using my model!
Yep, that's how the world works.
It crushes the honest and exalts jerks.
sounds good.
The Mane 6 meet Discord on Season 2 Episode 1 and 2 Return of the Harmony.
Who else wants a pair of headphones like twilight’s
04:13
Oh Princesses... 😨
What a fun fit.
that was a real trip
03:49
Yes, sir? But... Twi's a mare. 🤔
princess
2:57 😂
Can you do: ‘Fluttershy,’ ‘Rainbow Dash,’ & ‘Rarity’ sing: ‘We Can’t Stop’ by Miley Cyrus?
cute
Gostei
How to clone the voices?
There is a tutorial for that on my channel, however the user Vul has made a pretty good one as well. This includes synthesis ai voices with existing models as well as training new ai models.
The most important info is that this can operate offline (with nvida gpu) or on colab, and you will need about 5 minutes worth of audio dataset cut in 1~10seconds wav files to train the model well.