I'm trying Firefox on Android, and although I can hear the translation, it is really difficult, mainly because the translation doesn't go exactly paragraph by paragraph and therefore doesn't always properly sync. I suppose with podcasts it would be somewhat difficult to detect paragraphs exactly?

Some cool ideas for the spoken translation - allow the translation to come before the original audio - even better, maybe make it configurable. Imagine being able to hear the original, then translation, then the original again! Another thing that would be great is to hear the translation with different levels of volume - maybe a bit quieter, that way perhaps the brain neurons wouldn't be bombarded so intensively. Also, some quiet pauses between the original and translation would help as well.

No worries, cool app by the way. This is Firefox on Linux.