InterScribe · Practical event resources
Event caption rehearsal sheet
Assign captioning responsibilities, test the source-to-audience path and record delay, fallback decisions and attendee feedback for each room.
Worksheet revision: October 12, 2026 · Free to use without an account
1. Make responsibilities explicit
A working keynote does not verify another room or a remote speaker. Complete one sheet per distinct delivery path. Confirm the support arrangement with the attendee or responsible access lead; a technical operator should not decide an accommodation while repairing a microphone.
| Owner | Decision / evidence required | Named person and backup |
|---|---|---|
| Producer | Programme, resources and authority to pause | |
| Access lead | Requested support and an agreed alternative | |
| AV lead | Microphones, outgoing feed, room/session mapping | |
| Caption provider / reviewer | Language, terminology and text quality checks | |
| Audience support | Joining, device check and feedback route |
2. Trace source → text → audience
Use a distinctive sentence containing a name, a number and a negative, then an ordinary question. Keep the expected words in the rehearsal brief. An input meter, a readable preview and an attendee’s successful task are separate observations.
| Checkpoint | Try this | Observed result / next action |
|---|---|---|
| Source | Speak at the lectern, handheld and remote microphone; include relevant playback. | |
| Text | Read the received captions; check meaning, names, numbers and “not”. | |
| Audience | Open the intended session on a separate device; check reading, joining and asking a question. | |
| Transition | Move to the next room/session and confirm the source belongs there. |
3. Record latency without promising a universal threshold
Measure the path the audience will use. Record a source cue and when that cue is heard or displayed at the attendee endpoint, with the clock or recording method. A connection gauge is not the same as end-to-end delay. Keep floor audio, captions, translated text and interpreted speech as separate routes.
| Route / device / conditions | Source cue and received cue times | Method / observed delay / task impact |
|---|---|---|
For example, a tester may record the source and attendee display together, identify the same cue in that recording and note the difference. Write down limitations of the method. Do not compare timestamps from unsynchronized devices as if they were precise. Repeat under representative room and network conditions and ask whether the observed behavior supports the actual participation task.
4. Practise a fallback before opening doors
Name the person who can pause or alter the programme. Agree what attendees will be told and where they can obtain help. A transcript tomorrow does not replace live participation today. Do not silently swap captioning for another format without discussing the person’s needs.
5. Ask for useful feedback
“Were you able to follow the parts you needed? Where did the text or joining route make that difficult? Was there a point when you could not ask or respond? What change would help next time?”
Offer a private, usable feedback route. Record the issue and task impact without collecting unnecessary personal information. Distinguish “the captions appeared” from “the attendee could complete the activity.” A successful rehearsal is evidence about that tested arrangement, not a compliance certificate or a guarantee for every session.