MrJokerRager
kiwifarms.net
- Registrado
- 2 de Nov, 2019
Someone made this cover for Jason Aldean.
Well that song has been reuploaded now.
Well that song has been reuploaded now.
AI-Generated Music (+ Parodies / Covers / Original Content) - The machines are learning art. Observe...
- Iniciador del tema Manimal House
- Fecha de inicio
-
🚨 On-and-off DDoS attacks. (July 25th: On-going connection issues.)
🧅 Tor is stable: https://kiwifarmsaaf4t2h7gc3dfc5ojhmqruw2nit3uejrpiagrxeuxiyxcyd.onion/ -
Want to keep track of this thread?
Accounts can bookmark posts, watch threads for updates, and jump back to where you stopped reading.
Create account
- Registrado
- 10 de Nov, 2021
Updated the OP with links to guides and cow models. Will do my best to update as more get made/discovered.
- Registrado
- 20 de Oct, 2018
The Metokur Voice Model is getting there. The actual voice isn't the issue - for whatever reason there is a hell of a lot of interference in the voice. I'm going to go over the sound samples again and refine more, but overall not the worst.
- Registrado
- 13 de Feb, 2022
Have you used Instrument HQ 2 or Vocs FT with "vocals" selected for Ultimate Vocal Remover? That usually does a pretty good job of removing any background noise from a person's voice for the most part.
- Registrado
- 20 de Oct, 2018
There's not actually any real background noise on Jim's mic. I'm not sure what's caused it. There was about 80 minutes worth of audio pumped in from his year of the chud streams with no background music.
I'm going to rerun it after reviewing the sample to see if there's something I missed.
The weird thing is also that it's not consistent. I ran the same settings on the same vocal recording of a song twice with exactly the same settings and one came out crystal clear while the other had a lot of interference.
- Registrado
- 13 de Feb, 2022
What pitch extraction algo are you using? Currently the "best" setting right now is rmvpe, but given Jim's mic perhaps Harvest might be a better option since from what I've read it's a bit more lenient in regards to noise on a person's voice. Also try going with 40k for the "Target sample rate" since 48k is more reserved for studio quality vocals since with standard vocals at 48k it starts to sound really fucky on the highend but with 40k it's a lot less harsh and robotic sounding.
hemmoroidlover55
kiwifarms.net
- Registrado
- 10 de Feb, 2023
i found this cover on youtube
- Registrado
- 21 de Jul, 2019
Maybe it's his shitty Logitech headset.
- Registrado
- 31 de Oct, 2022
Can anyone make a guide for how to make an AI voice model for use with the parodies, or link me to a good one(ideally something free and anonymous)? I want to make a PPP voice to have it sing some DSP detractor songs.
- Registrado
- 3 de Feb, 2023
Best I can do, My cobes model still sucks
- Registrado
- 13 de Dic, 2022
This is what all those idiots who are outraged by this milquetoast song hear.
- Registrado
- 26 de Jun, 2021