Hi VC_1, totally understand your situation, please be aware that we are constantly monitor the voice performance through different use cases and will provide update and bugs fix through time to bring the better result, and any configuration change, will introduce a new model variation. If the fix is creating new issue, we will also look into it and adjust asap, your details information is very critical for us to understand the issue on your end and compare with our observations.
If possible, please send an email with information below to this ttsvoicefeedback@microsoft.com
- Input SSML
- Audio samples before and now
- Issue description
- Expected result
- Your Azure subscriptionID
I have talked with the team who monitor this email to pay attention on your feedback, you can note that you have mentioned this bug to me on Blog post.
Hope this could help!
Garfield