For example at this spoken Wikipedia audio at 11:15, the tool switches the voice and says it in what seems to be Spanish(?) despite that I selected English in both the input and output language and selected the en Brian voice.
This occurs roughly every 4th audio once (more likely for longer articles I guess) so it's not very rare. Another example can be heard here where the voice also switches and pronounces "formation" in French instead of English.
For the first file, the input srt file has (after linebreak) autoregressive (i.e. and in the second file it's (after linebreak) Formation (then another linebreak).
Please make sure the voice is not swapped and that the pronunciation is either always English or always English if the word exists in some English dictionary/….
For example at this spoken Wikipedia audio at 11:15, the tool switches the voice and says it in what seems to be Spanish(?) despite that I selected English in both the input and output language and selected the en Brian voice.
This occurs roughly every 4th audio once (more likely for longer articles I guess) so it's not very rare. Another example can be heard here where the voice also switches and pronounces "formation" in French instead of English.
For the first file, the input srt file has (after linebreak)
autoregressive (i.e.and in the second file it's (after linebreak)Formation(then another linebreak).Please make sure the voice is not swapped and that the pronunciation is either always English or always English if the word exists in some English dictionary/….