Interface: BuiltInAdaptiveVadProviderConfig

Defined in: packages/agentos/src/io/hearing/providers/BuiltInAdaptiveVadProvider.ts:17

Configuration for the BuiltInAdaptiveVadProvider.

All fields are optional — sensible defaults are applied for 16kHz mono audio with 20ms frames (the standard voice pipeline configuration).

See

BuiltInAdaptiveVadProvider for usage examples See AdaptiveVADConfig for detailed VAD tuning parameters. See CalibrationConfig for environmental calibration settings.

Properties

calibration?

optional calibration: CalibrationConfig

Defined in: packages/agentos/src/io/hearing/providers/BuiltInAdaptiveVadProvider.ts:38

Environmental calibration settings. Controls how the noise floor and spectral profile are estimated from ambient audio.

See CalibrationConfig for available options.

---

frameDurationMs?

optional frameDurationMs: number

Defined in: packages/agentos/src/io/hearing/providers/BuiltInAdaptiveVadProvider.ts:30

Duration of each audio frame in milliseconds. Determines the number of samples expected in each processFrame() call. At 16kHz with 20ms frames, each frame is 320 samples.

Default

20

---

sampleRate?

optional sampleRate: number

Defined in: packages/agentos/src/io/hearing/providers/BuiltInAdaptiveVadProvider.ts:22

Audio sample rate in Hz. Must match the actual audio stream.

Default

16_000 (16 kHz, standard for voice processing)

---

vad?

optional vad: AdaptiveVADConfig

Defined in: packages/agentos/src/io/hearing/providers/BuiltInAdaptiveVadProvider.ts:46

VAD algorithm configuration. Controls speech detection thresholds, minimum speech duration, hangover period, etc.

See AdaptiveVADConfig for available options.