Pronounciation and other finer control about edge-tts HOT 10 CLOSED

This function previously existed but was removed. It doesn't work properly due to Microsoft closing up the API and so I removed it. If you send XML that uses features not present in Edge browser the API will error out now.

from edge-tts.

It expects XML documents with only XML document formats that would be generated by Edge Browser. I believe it checks if it matches the template the browser uses.

from edge-tts.

Ah, thanks for the info.
If the API errors out, can you gracefully handle the error so it's a "might not work if you try it, but it's there if you want to try it"? Or was the problem that you kept getting issues opened for things that Edge didn't support and you didn't want to support that?

from edge-tts.

Well the issue was that the API didn't error out exactly, it just returned nothing. In any case it does raise an exception for when the API returns no audio.

Either way users of other libraries that depended on edge-tts made issues because the library that they were using was using custom SSML and was not working, so I removed it entirely.

I was never able to get it to do anything useful anyway. It used to support all the Azure Cognitive Services features but not anymore. It doesn't even support setting a custom pitch value.

from edge-tts.

Ah... That kind of sucks.
Let me see on Monday if I can find the team responsible for the Edge and see if they are planning to make the API more stable and support those extra features. Hold your fingers crossed, but temper your expectations :)
And thank you!

from edge-tts.

The API is stable, it's just that using the paid Azure Cognitive Services features no longer works. At the start they don't do any sort of filtering.

from edge-tts.

It's intentional, basically!

from edge-tts.

Oh, so they didn't shut it down, just moved it to a paid tier? Ok, I guess that makes sense... To be fair, it's one of the best TTS's around and the basica functionality is free so guess we can be happy with that at least :)

from edge-tts.

Well it was always a paid service. Azure Cognitive Services is a paid service for TTS generation. Edge browser's online TTS is basically just Azure Cognitive Services with restrictions.

Anyway if pronunciation and other fine control is available functionality in Edge, then I could probably add it to the library. Otherwise I guess this issue needs to be closed.

from edge-tts.

Got it, so it's essentially not even supposed to be free, and we just shouldn't talk about it and be happy it works :)
Cool, thanks, closing the issue.

from edge-tts.