https://huggingface.co/WoomyPearl/RVC-Model-Palace/resolve/main/NASBAnnouncer.zip Locally trained, credit me if used!

Why is there so much background noise?

The original voice clips have echo and reverb

Couldn't you have removed it from the audio before training?

There are reverb and echo removal models out there

Is there a way to batch remove them

I got the voice clips from the VG Resource

You could put all the clips in audacity and export them as one I think

You can put all the voice clips into a single audio file and run it through UVR

I think that’s kinda painful for me to do

Since I have to then separate each individual voice clip

You could try to use this https://audio-joiner.com/

(Actually nvm I can export multiple in one wav)

You can use FoxJoy Reverb Removal or UVR DeReverb