When I go to generate captions from a generated transcription, the result is always a overshoot of the actual clip, which needs to be fixed manually. Tho this has been much worse in previous versions, it still exsists none the less. Premiere clearly knows where each word is, as the follow along animation in the captions indicates, it knows when a clip ends, why does the caption then overshoot this?
Makes no sense to me and always results in a lot of busy work.