Input & Microphone

These arguments control microphone input and related settings.

Arguments

Flag	Description
--microphone_enabled	Enable microphone input.
--list_microphones	List available microphones and exit.
--set_microphone	Set the default microphone by name or index.
--energy_threshold	Set the energy threshold for audio detection (default: 100).
--mic_calibration_time	Duration (seconds) for microphone calibration.
--record_timeout	Real-time recording chunk length (seconds).
--phrase_timeout	Silence duration (seconds) before starting a new transcription.
--mic_chunk_size	Number of audio chunks to collect before processing (default: 1).
--paddedaudio	Number of chunks to overlap from previous batch for better context.

Details & Examples

--microphone_enabled

Enable microphone input for real-time transcription.

--list_microphones

Lists all available microphones and their indices. Example:

python synthalingua.py --list_microphones

--set_microphone

Set the microphone by name or index. Examples:

python synthalingua.py --set_microphone "Microphone (Realtek USB2.0 Audi)"
python synthalingua.py --set_microphone 4

--energy_threshold

Energy level for microphone to detect audio (default: 100). Adjusts how sensitive the microphone is to sound. Higher values = less sensitive to background noise.

--mic_calibration_time

How long to calibrate the mic for background noise. Example:

python synthalingua.py --mic_calibration_time 5

--record_timeout & --phrase_timeout

Control how often the mic records and when a new line is started. Note that if --discord_webhook is enabled and --phrase_timeout is set to a value greater than 1, it will be automatically set to 1 second to avoid multiple webhook messages.

--mic_chunk_size

Controls how many audio chunks to collect before processing. Use with --paddedaudio for batch processing:

python synthalingua.py --microphone_enabled --mic_chunk_size 3 --paddedaudio 1

--paddedaudio

Number of chunks to overlap from the previous batch for better transcription context. Works with microphone input when using --mic_chunk_size > 1, and also with streaming (--stream) when using --stream_chunks:

python synthalingua.py --microphone_enabled --mic_chunk_size 4 --paddedaudio 2

This will process 4 chunks at a time, with 2 chunks overlapping from the previous batch.

---

Back to Index