Speaker assignment improvements
1.3.3·
A reworked speaker assignment workflow for recording transcripts, with fully manual labeling, in-place reassignment, and protection for unsaved changes.
Improvements
- Manual speaker assignment for recording transcripts. Speaker labeling is now fully manual, giving you direct control over each voice and letting you merge any extra speakers the transcript detects into a single one.
- Reassign speakers directly in the transcript. Speaker badges in the transcript are now interactive, so you can reassign or merge a speaker in place without returning to a separate assignment section.
- Protection for unsaved speaker changes. A persistent save bar tracks your speaker assignments, flags anything unsaved, and keeps the panel open until you save or discard.
- Code-quality checks scan the right files. Corrected the developer code-quality checks so they scan only the intended files instead of raising false issues.