AI Audio Quality Issue in Japanese

I tried the Japanese AI Audio with multiple characters, but the quality was extremely poor and not at a practical level. Regardless of how the text was input, there were issues with incorrect kanji readings and unnatural intonation that couldn't be improved. It's unclear whether this feature underwent final quality testing. The unnaturalness is particularly striking when compared to what's offered by Gemini and GPT. Are there any plans to improve the quality for languages other than English? While there may not be many Japanese users, I would like you to consider receiving proper internal feedback.