Error 13 in text-to-speech systems typically refers to an issue with processing or interpreting the text. It could be due to various reasons such as unsupported characters, formatting, or even the structure of the text itself. The fact that the same neural voice works on other text files but not on this particular one indicates that there might be something specific about the problematic file causing the error.
Review the text in the problematic file for any unusual characters, formatting, or structural issues. Look for special characters, unsupported symbols, or any uncommon formatting that might be causing the error.
Compare the content and formatting of the problematic file with the ones that are working fine. This might help you identify any specific differences that could be causing the error.
And try preprocessing the text by removing any unsupported characters, ensuring consistent formatting, and structuring the text in a simpler way to see if that resolves the issue.
After doing these steps and the issue still persists, please include more information regarding this use case.
Thanks Poala. Is there a list of unsupported characters per Voice (even within same language code) ?
If there are unsupported characters, both voice generations should have affected? Why would en-US-News-K work but not en-US-Neural2-J ?Is there any diagnosis or debug methods to see what could possibly be wrong?