A lot of the times, the tones/pronunciation is off, and sometimes, the voice sounds like it’s panting, squealing, or generally emoting in weird ways. The most recent example I’ve seen of this is the TTS for 隔天, but it’s happened a couple of times before as well. The only thing I can guess is that these pronunciation recordings are AI-generated and use some seriously weird training data. I wish they’d use something more professional.
What AI voice you have selected under the settings?