What improves speaker labeling the most
Speaker labeling works best when the recording has clear, separated voices and minimal overlap.
Best practices
- Reduce overlap: Encourage turn-taking and avoid people talking over each other.
- Stay close: Keep NotePin S close to the group (closer is better than relying on distance pickup).
- Lower background noise: Move away from loud music, wind, traffic, or crowded areas when possible.
- Don't block the mic: Avoid covering the device with clothing or placing it inside thick fabric.
- Keep positions stable (small groups): If possible, keep speakers in consistent positions instead of moving around a lot.
How to verify
Open the transcript and check whether speech is consistently separated into Speaker 1 / Speaker 2 / Speaker 3 rather than being merged or frequently switching.
Optional: use highlights for better outputs
Highlights won't "fix" speaker labeling, but in group settings they help make the summary more reliable by flagging key decisions and action items.
If labeling is still unreliable
Collect a short sample, speaker count, environment details (noise/room), wear style/placement, and firmware/app version, then escalate per support policy.