Respeaking is a captioning technique that produces captions and transcripts through the use of voice-recognition technology. Captioners who use respeaking are called respeakers.
In order to create captions, a respeaker will listen to an audio source (for example, a recorded video or a live event via telephone connection) and speak into a microphone, repeating exactly what they hear and adding punctuation and sound effects as necessary.
The respeaker’s microphone is connected to specially-designed voice-recognition software, which turns the respeaker’s output into captions that then become visible on screen.
- For recorded captions, a respeaker will produce a transcript of the audio, then tidy up the captions, time them to be in sync with the video’s audio and produce a caption file.
- For a live event, the captions are broadcast via wi-fi to the viewer’s device in real time.