Showing who's speaking in live meeting captions

Captions are easier to follow when readers know who is talking, especially in a meeting with a full dais, staff reports and public comment. This article explains how speaker names work in SentraCaptions and how to set them up without slowing down your meeting.

Why speaker names matter

A hearing-impaired resident watching the chamber screen can't always see who has the microphone. A remote attendee on Zoom may only see a slide or a wide shot of the dais. Someone reading translated captions on a phone has even less to go on. Without names, a back-and-forth between a councilmember and staff can read like one long paragraph.

Names also help afterward. Transcripts that show who said what are much easier to use when staff prepare minutes. Transcripts are not official minutes, but they can help.

How it works

Step 1: voices are told apart

Speech recognition tells voices apart and labels them Speaker 1, Speaker 2, Speaker 3 and so on. It does not know who those people are. It is not identifying anyone automatically and does not use voiceprints.

Step 2: an operator adds names

On the live desk, an operator types a name for a speaker label, for example "Mayor", "Councilmember Rivera" or "City Manager". From then on, captions show that name when the speaker changes.

Step 3: names appear everywhere

Once a name is set, it shows on:

Because the name appears when the speaker changes, readers aren't distracted by a name on every line.

Check names after a break

Speaker labels can change after captioning restarts, for example after a recess. The voice that was Speaker 2 before the break may come back with a different label.

Make it a habit: after every recess or closed session, the operator watches the first few exchanges and re-checks the names. It takes a minute and keeps the rest of the meeting accurate. For more on breaks, see Pausing captions for recess and closed session.

Tips for the operator

  1. Prepare a name list before the meeting. Write down how you want each person labeled: title and last name for members, title for key staff. Decide in advance whether to use titles like "Chair" or full names.
  2. Use the roll call. Roll call is a natural moment to match voices to names, since each member speaks in turn.
  3. Add names to the glossary too. The glossary of local names helps speech recognition spell council members, streets, neighborhoods and programs right in the captions themselves. Speaker names and the glossary work together. See Caption glossary for local names.
  4. Keep public comment simple. Many agencies may prefer a neutral label such as "Public comment" rather than typing each speaker's name. Follow your agency's own practice for how speakers are named in the record.
  5. Re-check after breaks. As above, labels can shift when captioning restarts.

What speaker names don't do

It's worth being clear with staff and residents about the limits:

If your meeting needs more hands-on control, the optional live correction desk lets an operator review and fix captions before they go out, at the cost of a short delay. See Human review for live captions.

Speaker names with Zoom and transcripts

If you send English captions into Zoom, the names you set on the live desk show in Zoom's closed captions too. Long captions are sent to Zoom and YouTube as two-line pieces (about 84 characters) so they read well. For setup, see How to add live captions to Zoom public meetings.

After the meeting, transcripts are available in every language as TXT or SRT files, with the names you set. That gives the clerk's office a head start on the draft minutes and on subtitles for the posted recording. See Transcripts for minutes and records.

Learn more