Speaker Detection

How It Works

lognow uses AI-powered speaker diarization to detect and label different speakers in your recording. Each speaker is assigned a label (Speaker 1, Speaker 2, etc.) and their segments are visually distinguished in the transcript.

Speaker detection works automatically — there's nothing you need to enable or configure.

What to Expect

• Speaker labels appear next to each transcript segment
• Color coding helps visually distinguish between speakers
• The system can detect multiple speakers in the same recording
• Speaker identification is based on voice characteristics, not pre-registered profiles

Accuracy Tips

Speaker detection works best when:

• Speakers take turns rather than talking over each other
• Each speaker has a distinct voice (pitch, tone, pace)
• Background noise is minimal
• Speakers are roughly equidistant from the microphone (or each have their own mic)

Limitations

• Speaker labels are generic (Speaker 1, Speaker 2) — the system doesn't identify people by name
• Very short utterances (a word or two) may not be reliably attributed
• Simultaneous speech (crosstalk) may be attributed to one speaker
• Speaker labels are consistent within a single session but may differ across sessions

Editing Speaker Names

After your recording is complete, you can rename speaker labels in the session view to replace generic labels with actual names. Click on any speaker label to edit it.