Voice Separation & Diarization
NoteWave uses advanced acoustic diarization to automatically detect when different people speak in a meeting. Each distinct voice is assigned a label (e.g. "Speaker 1", "Speaker 2"), making multi-party discussions clear and easy to follow.
Renaming Speakers Live & Post-Meeting
Meeting owners can rename detected speaker labels at any time:
- Hover over or click any speaker label in the transcript
- Click Rename Speaker
- Type the person's real name (e.g. "Dr. David Miller")
- Confirm the change — NoteWave updates every occurrence of that speaker across the transcript timeline, participant roster, and exported documents
Merging Duplicate Speaker Labels
If a participant changed microphone positions, stepped away, or had varying audio levels, the AI may occasionally split their speech into two separate speakers (e.g. "Speaker 1" and "Speaker 3").
To merge them, simply rename the duplicate label to the existing person's name. NoteWave will recognize the match and merge all dialogue into a single unified speaker timeline.
Pre-Assigning Expected Attendees
When recording via meeting links or in the recording modal, you can type in expected attendee names beforehand. The AI will cross-reference attendee names with vocal introductions and calendar invite metadata to assign proper names right from the start.
Best Practices for Clean Separation
- Avoid cross-talk: When multiple people speak simultaneously over one microphone, the acoustic wave combines, making distinct voice isolation harder.
- Microphone placement: Place an omnidirectional conference microphone in the center of the table rather than near a single speaker or a laptop fan.
- Video call bots: When using NoteWave Zoom, Teams, or Meet bots, each participant's audio stream is separated natively by the meeting platform, resulting in nearly 100% accurate speaker identification.
Was this article helpful?
Your feedback helps us improve our documentation.