Live Captioning Methods

Live Captions for Conferences and Lectures

A planning guide for adding real-time text to in-person talks, from choosing a method and sharing materials to audio, space and backup arrangements.

On this page
  1. Start with who needs what
  2. Choose the method
  3. Get the audio right
  4. Share materials in advance
  5. Brief the speakers
  6. Plan the display
  7. Handle recordings and streams
  8. Have a backup plan
  9. Test and communicate
  10. Costs and expectations

Start with who needs what

Conference and lecture captioning is easiest to plan when you begin with the audience rather than the technology. Some attendees may be deaf or hard of hearing and rely on text to follow. Others may be non-native speakers, people in a noisy back row, or people who simply take in written words more easily. If you do not know who will attend, registration forms can include an optional question about access needs. Asking early gives more time to respond.

The next questions are practical. How many rooms are in use? How long are the sessions? Is each session a single speaker or a discussion? Are there pre-recorded videos in the program? Answers to these shape everything else.

Choose the method

The main options are a trained human writer, a respeaker, automatic captions, or a mix. At the planning stage, a few points matter most.

  • Specialized or high-stakes content usually calls for a human-led method.
  • Small, informal sessions with clear audio may be reasonable with automatic captions, provided expectations are clear.
  • Long days need more than one human captioner, since sustained live work is tiring, so that people can take turns.
  • Several rooms might use people in the main hall and automatic text in side rooms. This is a choice to state openly so that attendees know what to expect.

Remote captioners are common, working from audio sent over the internet. That can widen the pool of available people, but it adds a dependency on connections, so a backup route is worth asking about.

Get the audio right

Most live captioning problems begin with sound. Whoever produces the text needs a clean feed, ideally taken directly from the sound mixing desk or from the speaker’s microphone rather than a microphone listening to the room.

  • Use microphones for every speaker, including audience questions. A captioner cannot write what they cannot hear, and automatic tools cannot recognize it.
  • Ask panelists to use a microphone each, or share carefully.
  • Test the feed before doors open, with a speaker reading the type of material that will be presented.
  • Avoid background music and loud ambient sound where possible.

Poor audio quality will hurt automatic tools the most, but it slows human writers as well.

Share materials in advance

Preparation is one of the most valuable things you can give a captioner. Useful items include:

  • the schedule and session titles
  • speaker names, with correct spellings and pronunciations if unusual
  • slides, scripts or abstracts
  • a list of acronyms, product names and technical terms

For human writers, these let them load terms in advance. For automatic tools, many allow a custom vocabulary. A shared term glossary also helps any later cleanup.

Late changes happen. A short message when a speaker is swapped or a session moves keeps the captioning team in step.

Brief the speakers

Speakers have a large effect on caption quality. They can help by speaking at a steady pace, finishing one thought before starting the next, repeating audience questions before answering, and sharing their slides ahead of time. A brief note in the speaker instructions is usually enough.

Plan the display

Decide where the text will appear and who will be able to see it. Options include a screen beside the stage, a projection, a dedicated monitor for people who prefer it, or text on attendees’ own phones or laptops through a web page. The layout should be checked from the seats farthest away and from wheelchair spaces. Text size, contrast and line length matter as much for live text as for recorded subtitles. The placement of displays at events deserves its own planning, since it affects who can read the text comfortably.

If captions appear on a screen shared with slides, consider whether the text will cover important parts of the slide.

Handle recordings and streams

If the event is streamed or recorded, decide whether live captions will be included in the stream, and whether open captions or closed captions are right. Captions that attendees can switch off are usually more flexible than text burned into the picture. Afterward, plan for cleanup if the recording will be published, as live text is rarely ready as it stands.

Have a backup plan

Things go wrong at live events. Ask what happens if:

  • the captioner’s connection drops
  • the microphone feed fails
  • the captioner is unwell or delayed

A second captioner on standby, an alternative audio route, and a person responsible for contacting the captioning team during the event all reduce risk. Tell attendees where to go if the captions stop and who to ask.

Test and communicate

A short technical rehearsal is worth more than a long email. Try the real equipment, check delay and readability, and adjust. Then let attendees know that captions are available, where to see them, and that, being live, they may lag or contain errors. Some events also add a note to an accessibility statement, so people can find out in advance what will be provided.

Costs and expectations

Costs vary widely with method, event length and the number of people required, and prices change, so it is best to ask several providers for details and what is included. Accessibility requirements for events differ by place, and a qualified professional or the relevant authority can say what applies. Captions help many people but cannot be guaranteed to be perfect, so choose the approach that fits the risks of your content.

The complete guide to live captioning methods

How Live Captioning Works
Read the guide

Keep exploring

Guides from other topics

All topics →