Skip to content

Transcribe Streaming  >  Structures  >  AudioEvent

AudioEvent

Structure Class

AudioEvent dataclass

A wrapper for your audio chunks. Your audio stream consists of one or more audio events, which consist of one or more audio chunks.

For more information, see Event stream encoding.

Attributes

audio_chunk class-attribute instance-attribute
audio_chunk: bytes | None = None

An audio blob containing the next segment of audio from your application, with a maximum duration of 1 second. The maximum size in bytes varies based on audio properties.

Find recommended size in Transcribing streaming best practices.

Size calculation: Duration (s) * Sample Rate (Hz) * Number of Channels * 2 (Bytes per Sample)

For example, a 1-second chunk of 16 kHz, 2-channel, 16-bit audio would be 1 * 16000 * 2 * 2 = 64000 bytes.

For 8 kHz, 1-channel, 16-bit audio, a 1-second chunk would be 1 * 8000 * 1 * 2 = 16000 bytes.