Update README.md

0875d41 over 2 years ago

4.59 kB

license: cc

(THIS IS UNFINISHED!!!)

This was made with: https://www.ibm.com/demos/live/tts-demo/self-service/home

(IBM Watson text to speech)

This is a text to speech online generator for free. If you use this, refer to their terms of service: https://watson-developer-cloud.github.io/terms?name=Text-to-Speech%20Demo

OK, so you got your model with RVC. But you have another voice cloner as well. How can we transport the model to the other voice cloner? This is actually extremely easy. All we need is to create audio with our current models, and upload that audio!

The worst voice cloners work with 2 hours of audio... I know, they suck... Therefore, here is a 2 hour audio file of Allison from IBM (read their TOS.). Just convert this file to another character with RVC, then upload the converted file!

It actually took a very long time to generate this audio file... I think I will only make one...

So, download the sample audio FIRST. Then convert that, even though it's only a couple seconds.
Upload that sample audio to RVC and convert it. Most likely the pitch will be wrong. So keep changing the pitch until it's correct.
Finally download the FULL audio, and change the pitch to the right one.
Now convert the audio file for real with the new pitch changed.

Note: It will take about a minute and 30 seconds to upload the wav file. Please be patient.

Note 2: It will take probably around 1 hour to convert a 30 minute file.

Note 3: hm... It appears that uploading TOO MUCH audio will crash the program... I will find out the maximum amount of time allowed... EDIT: it's around 30 minutes max.

Note 4: it takes around 3400 seconds to turn 31:17 minutes of audio into another voice. That means for every second of audio, the program will take 1.8 seconds to convert it. (1.811401172 seconds to be exact...)

I will shorten these 2 hour files into 4 files of 30 minutes each, in the next update

Audio in .wav format:

Sample Audio:

https://huggingface.co/PhoenixStormJr/Use-RVC-for-other-voice-cloners/resolve/main/AllisonDEMO.wav