Live transcription
Text that appears while someone is speaking, a few seconds behind, instead of once the recording has finished.
Live transcription (or real-time transcription) writes down what’s said as it’s said. The system receives the audio in small pieces, proposes provisional text and confirms it as it gets more context, which is why the last few words may change a moment later.
It’s useful for following a talk without audio, for people with hearing difficulties to read what’s said, or for taking notes with the text in front of you. Because it works with less context, it’s usually a little less accurate than a transcript made afterwards.
In Echo, Live transcribes your microphone and, if you like, the sound from another browser tab (a video call). It records at the same time and, when you finish, transcribes the full recording again with speakers and a summary. You can mark moments and write notes with their minute.
You can also share live captions with a link: guests see them without an account and, if they choose, translated into their language. More on the transcription page.
