I’m a little confused at the inter-phrase break accuracy of Aeneas. The language I’m working in now has a fairly straight-forward orthography, and it seems like Aeneas does an almost perfect job of aligning the text with the audio. But I have selected the option to break in the middle between the phrases. Sometimes it does that very well, but quite often it will put the break right as the next phrase starts, or right at the previous phrase end, or sometimes even somewhat AFTER the next phrase has started. So it looks like Aeneas has correctly identified the phrases, but it can’t seem to put the break in the middle of the (relatively) silent audio clip between the phrases. Has anyone else seen that? This isn’t a huge deal, but it just seems like it could do a lot better fairly easily. I assume this parameter is getting passed to Aeneas correctly – I see “task_adjust_boundary_algorithm=percent|task_adjust_boundary_percent_value=50” as part of the command. It would be nice to have a bit more consistent “flow” from one phrase to another as the App reads the text, but I’m not really willing to manually fine-tune the timings of the entire NT.
You must be logged in to reply to this topic.
Except where otherwise noted, content on this site is licensed under a