An audience at a performance

Revoicing and Voice Recognition


Live Subtitles can be produced using a speech-to-text reporter as outlined here . Voice recognition software is becoming increasingly sophisticated and can now be used to provide access for some live events, using a revoicer.

Revoicers, also known as respeakers, repeat clearly what is being said during unscripted events using special software that’s trained to recognise their voice. Their speech is then converted into text which appears on a display device.

This type of access has successfully been used for post-show talks.

Things to consider:

  • For this kind of access to work well, revoicers should be seated in an area separate to the event where they can hear clearly, but can’t be heard by the audience. This means that a good sound feed is essential.
  • Revoicers also need to pare down (edit) the live dialogue or conversation, which means the text that appears isn’t verbatim, although it will always give a good idea of what’s being said.

Voice recognition

Voice recognition can be used without a revoicer, but the results from this, in English at least, are not very good at present. YouTube has an automatic captioning feature on its site now, which is an excellent development, but it’s recommended that you correct any errors in the automatically generated subtitles before relying on them. Sometimes they only bear a passing resemblance to what is actually being said, like this:


This video tells you how you can fix the automatically generated captions.