TTS Speech Multilingual Voice Completely changed. Is there a way to use the previous models?
Alliosis
0
Reputation points
Last night I've been using en-US-AvaMultilingualNeural for multilingual translation and the voice was consistent. However, today, the voice completely changed when reading out other languages, and it sounds very inconsistent. (English is still the same)
Is there a way to use an older version of the neural voice? Or is there somewhere I can look to check on updates that were pushed out?