AI voice singing feature

[tight goblet]

Fish audio should add a feature where we can use an AI voice model to make an AI cover out of.
Like you can use any audio file and make the desired ai voice do a cover of it.
It would actually help because you wouldn't have to take forever for someone else to do it for you, and it's also good for people who aren't confident enough or not motivated enough to sing or make someone else do it for them.

[daring mesa]

Look what shadow said:
#1482698351614103644 message

They are planning on adding speech to speech soon!

[tight goblet]

I already read that before you even told me about it, if you were wondering.

I'm excited for it really. So I can use it instead of weights

[daring mesa]

Yeah.

[nocturne hemlock]

Where is it? I can't find it anywhere

[daring mesa]

Actually, they just released the voice changer.
It's not the same as speech-to-speech. Fish Audio's voice changer prevents AI characters from singing.

Even though the voice changer is similar to speech-to-speech, it is actually way different.

[night sedge]

the voice changer is very disappointing, if this is what they're releasing as "speech to speech". I hope maybe it's two different projects and that a proper Speech to Speech is still coming. Current voice changer doesn't keep the emotions at all. I suppose it could be ok for TikTok style voiceovers, but I was hoping for an actual speech to speech.

[daring mesa]

Same.

[nocturne hemlock]

But that's what we all want, characters singing since Weights GG shut down last month

[night sedge]

I don't want to make them sing, I want to be able to read something in a certain emotion with sighs, breaths, laughing, etc, and have it convert it to the voice chosen. Like Eleven Labs' s2s feature.