Individual Speaker Detection
Projectile Objects
Otter.AI identifies separate speakers. I tried this with Descript and it blended speakers together. Sometimes I want to select and delete all of the "interview's" questions and edit only from the "subjects" dialoge.
Canny AI
Merged in a post:
Fix speaker detection in latest update
E
Eric Leone
Since the latest update in June 2024, my interview video clips that feature the interviewee and interviewer audio on the same track are not being identified separately. The previous version auto detected the unique voices and allowed you to name the speakers. This new version has no way to fix incorrect speaker identification without adding in the speaker name every time they talk on the composition.
Maybe if you add the speaker label in the composition one time, it can auto detect and update the rest of the composition? But it still would be quicker to identify speakers in the previous manner so you don't have to scroll through the entire composition to find where there's several unique speakers. Either way this simple, but essential feature needs to be fixed.