Skip to main content

​
What Chief transcribes

During a Live Session, Chief transcribes two audio sources from your Mac: your microphone (you) and your computer’s system audio (everyone else on the call). Each spoken turn lands in the Transcript with a speaker label and a timecode range.

​
How speakers are labeled

Chief separates speakers automatically:
  • Your own microphone is always shown as You.
  • Other voices are labeled Speaker 1, Speaker 2, Speaker 3, and so on, each with its own color.
The numbering starts at Speaker 1 — there is no “Speaker 0.”
ℹ️ Good to know: “You” and “Speaker 1/2/3” are derived from the audio, not real names. When Chief hears an actual name in the conversation, it can use that name on an action item’s owner instead of a generic speaker label.

​
Reading the transcript

Each turn shows a colored speaker chip, the timecode range (for example 2:14–2:31), and the spoken text. Open a Live Session and scroll the Transcript panel to read it in order.

​
FAQ

Speaker separation is automatic and works from the audio itself. People who sound similar, talk over each other, or share a single device feed can occasionally be grouped. The label reflects the audio Chief heard, not a roster of attendees.
The transcript uses the automatic You / Speaker N labels. Where Chief picked up a real name in conversation, it can attach that name to an action item’s owner.
That means no audio has been transcribed. Confirm the Mac app has microphone and system-audio access and that the meeting audio is reaching your computer.