Is there any way to restore the ability to have manual text-to-speech xtts voice output recordings?

[Xeraster]

A while ago, you used to be able to go into Koboldcpp settings, configure text-to-speech for a locally hosted xtts coquI text to speech server and click "test". This opened a dialog where you could make the voice ai say whatever you wanted. This was the easiest way to use xtts BY A HUGE MARGIN compared to the next easiest way, which involves making your own webapp from scratch just to manage all this which is a monumental amount of work.

Well now that functionality has been taken away. Pressing "test" no longer gets you a dialog to make the sai say whateve you want. It just says "test, this is a test" and you don't get to modify what it says.

Please, is there ANY WAY to restore this? Because if not, I'll have to make my own xtts api manager webappp for this from scratch or pay for a online ai voice service.

[LostRuins]

Yes! The feature has not been removed, just relocated.

Now to add a custom speech, simply go to Add File -> Generate Speech

and you have many options to customize the generated speech

[LostRuins]

let me know if this solves the issue

[henk717]

@LostRuins might be a good idea to customize our test message with a message informing people where to find this. That way the test becomes the tutorial.

[LostRuins]

but the goal of the test message is to have something quick and short to test each voice.