Transcripts
Review speaker-labeled transcript segments, work with multiple recordings on one note, and add transcript text into your note.
The transcript is the record of what was said. It lives beside your note rather than in it — you write the note, Prismical keeps the transcript, and Enhance is what brings them together.
Opening the Panel
Click the chevron in the dock at the bottom of a note. The panel opens over the bottom of the window with the header Transcription.
It shares space with the Ask AI panel, so opening one closes the other. Expand it for a longer view, or close it entirely — the transcript is safe either way.
Reading It
Each line has a timestamp, a speaker label, and the text. Timestamps are times of day rather than offsets, so a line reading 9:41 AM tells you roughly when in the meeting it happened — useful for matching a transcript against a calendar. They mark the chunk of audio a line came from, so treat them as close rather than exact.
While you're recording, lines appear as people talk and the footer reads Listening and transcribing…. When you stop, they settle into a grouped log.
If a note has several recordings, each is grouped under its own divider — Recording 1, Recording 2 — with the time it started.
Empty note? The panel says "No transcript yet. Start recording to capture your meeting or voice note."
You and Them
Every line is labeled You or Them. The labels come from which source the audio arrived on: your microphone is You, your computer's audio output is Them.
That's a deliberate, mechanical rule, and understanding it saves confusion:
| Situation | What you get |
|---|---|
| A video call on a Mac with system audio | Your voice as You, everyone on the call as Them. |
| An in-person meeting | Everyone in the room is You — they're all coming through your microphone. |
| Recording on web or your phone | Everything is You, for the same reason. |
So "You" means came through my microphone, not was said by me. For an in-person conversation the labels carry no information — but Enhance still sorts out who said what from the content, which is what actually ends up in your note.
See System audio for why two sources make the difference.
What You Can Do With It
Open the recordings list in the panel footer for the actions on each recording:
| Action | What it does |
|---|---|
| Copy transcript | Copies the whole thing to your clipboard as Speaker: text lines, ready to paste anywhere. |
| Add this recording to the note | Runs Enhance on just that recording and folds it into your note. Recordings already folded in are marked with a check. |
There's no way to edit a transcript or correct a mis-heard word. The transcript is a record of what the model heard; your note is the thing you shape. If a name comes out wrong, fix it in the note — that's the version anyone will read.