The speech to text node accepts a Buffer of data representing a fragment of audio. This is supplied in msg.payload. When the node executes, the audio is parsed through a speech recognition algorithm and the corresponding text representation is returned in the new msg.payload value. The structure of the response is defined by the GCP Speech to Text API and is an instance of a RecognizeResponse object.
Processing an audio fragment takes time. A status visualization is associated with the node which is visible when the node is processing audio.
The node has configuration options including:
- Language Code - The language code to be used for processing. The set of allowed codes can be found here. The default is en-US.
- Sample Rate - The sampling rate of the audio input.
- Encoding - One of:
- LINEAR16
- FLAC
- MULAW
- AMR
- AMR_WB
- OGG_OPUS
- MP3