Transcriber

class Transcriber(default_microphone: str = 'list')

Bases: object

Real-time transcription of user microphone input

Methods

clear_recording_buffer	Clear the recording buffer
get_microphone	Get speech recognition microphone device
run	Return the most recent single phrase of microphone input

Attributes

record_timeout	How real-time the recording is.
phrase_timeout	Delay before ending user monologue.
energy_threshold	Microphone energy threshold for background noise.
default_microphone	Default microphone name.
phrase_time	The last time a recording was retreived from the queue.
last_sample	Current raw audio bytes.
data_queue	Thread-safe Queue for passing data from the threaded recording callback.
recorder	SpeechRecognizer to record and detect when speech ends.
temp_file	File for sending to OpenAI Whisper API.

clear_recording_buffer() → None

Clear the recording buffer

data_queue: Queue

Thread-safe Queue for passing data from the threaded recording callback.

default_microphone: str

Default microphone name.

energy_threshold: int

Microphone energy threshold for background noise.

get_microphone() → Microphone

Get speech recognition microphone device

last_sample: bytes

Current raw audio bytes.

phrase_time: datetime | None

The last time a recording was retreived from the queue.

phrase_timeout: float

Delay before ending user monologue.

record_timeout: float

How real-time the recording is.

recorder: Recognizer

SpeechRecognizer to record and detect when speech ends.

run() → str

Return the most recent single phrase of microphone input

temp_file: str

File for sending to OpenAI Whisper API.