I have audio files in a language that is spoken very quickly, and Aeneas is having some difficulties synchronizing the audio. (Not huge difficulties, but just from time to time - it still does amazingly well!)
But when I go to fine-tune the timings, I have two problems:
At first we were breaking phrases on commas, but decided there were too many commas, so we decided to reprocess without breaking texts at commas. So the phrases look like this:
4a|wo yʼan di ruute, yʼede:
4b|Nung ni tʼize oldiko, nʼun di yʼele wede i kʼize oldiko ye a beygung.
4c|Wo de kane an di ruute, anʼde:
4d|Kaye seezey ye, a wollo kee baa.
But on the fine-tune timings page, it comes up like this:
So it is breaking on commas here, and we have to ignore the text here and use the text breaks from the phrases file to do the timing.
I also noticed that often saving the timing file doesn’t “take”, but if I do it two times in a row, it seems to work consistently. What’s up with this?