MicrophoneInput

Defined in: src/providers/input/MicrophoneInput.ts:149

Browser audio input provider that captures audio from the microphone.

Remarks

MicrophoneInput adapts the SDK’s existing AudioCapture to the AudioInputProvider interface, making microphone capture a first-class pipeline role. It delegates all operations to AudioCapture:

Raw ArrayBuffer chunks from AudioCapture are wrapped into AudioChunk objects with timestamps and sequence numbers.

Data-flow diagram:

Microphone ──getUserMedia──> AudioCapture ──callback──> MicrophoneInput
                                                             |
                                                        wrap as AudioChunk
                                                             |
                                                             v
                                                   onAudio callback
                                                             |
                                                             v
                                                     InputQueue -> STT

Example

import { MicrophoneInput } from 'composite-voice';
import type { AudioChunk } from 'composite-voice';

// Create with custom microphone settings
const input = new MicrophoneInput({
  sampleRate: 16000,
  format: 'pcm',
  channels: 1,
  echoCancellation: true,
});

// Use in a 5-provider pipeline
const voice = new CompositeVoice({
  providers: [
    input,
    new DeepgramSTT({ apiKey: '...' }),
    new AnthropicLLM({ model: 'claude-sonnet-4-20250514', apiKey: '...' }),
    new DeepgramTTS({ apiKey: '...' }),
    new BrowserAudioOutput(),
  ],
});

await voice.initialize();
await voice.startListening();

See

• AudioCapture for the underlying Web Audio API implementation
• AudioInputProvider for the interface contract
• MicrophoneInputConfig for configuration options
• BufferInput for the server-side counterpart

Implements

• AudioInputProvider

Constructors

Constructor

new MicrophoneInput(config?): MicrophoneInput;

Defined in: src/providers/input/MicrophoneInput.ts:211

Creates a new MicrophoneInput instance.

Parameters

Returns

MicrophoneInput

Remarks

Instantiates the internal AudioCapture with the provided configuration. The provider starts in an uninitialized state; call initialize() before use.

Example

// Default configuration
const input = new MicrophoneInput();

// Custom microphone settings
const input = new MicrophoneInput({
  sampleRate: 16000,
  format: 'pcm',
  channels: 1,
  echoCancellation: true,
  noiseSuppression: true,
  autoGainControl: true,
});

Properties

Methods

dispose()

dispose(): Promise<void>;

Defined in: src/providers/input/MicrophoneInput.ts:256

Dispose of the provider and release all resources.

Returns

Promise<void>

Remarks

Delegates to AudioCapture.dispose(), which stops capture, closes the AudioContext, and releases the media stream. After disposal the instance should not be reused.

Implementation of

AudioInputProvider.dispose

getConfig()

getConfig(): Partial<MicrophoneInputConfig>;

Defined in: src/providers/input/MicrophoneInput.ts:286

Get a copy of the current audio input configuration.

Returns

Partial<MicrophoneInputConfig>

A shallow copy of the current AudioInputConfig from the underlying AudioCapture.

getMetadata()

getMetadata(): AudioMetadata;

Defined in: src/providers/input/MicrophoneInput.ts:431

Get the audio format metadata for the captured audio.

Returns

AudioMetadata

The AudioMetadata describing the captured audio format.

Remarks

Returns metadata derived from the configuration: sampleRate from config (default 16000), encoding as 'linear16' (AudioCapture always outputs 16-bit PCM), channels from config (default 1), and bitDepth of 16.

Used by the pipeline to auto-configure the downstream STT provider via configureSTTFromMetadata().

See

• AudioInputProvider.getMetadata
• AudioMetadata

Implementation of

AudioInputProvider.getMetadata

initialize()

initialize(): Promise<void>;

Defined in: src/providers/input/MicrophoneInput.ts:237

Initialize the provider, making it ready for capture.

Returns

Promise<void>

Remarks

Sets the initialized flag. If already initialized, this is a no-op. The actual microphone permission and AudioContext setup happen in start(), not here.

Implementation of

AudioInputProvider.initialize

isActive()

isActive(): boolean;

Defined in: src/providers/input/MicrophoneInput.ts:390

Check whether the microphone is actively capturing audio.

Returns

boolean

true when audio is actively being captured from the microphone.

Remarks

Delegates to AudioCapture.isCapturing(). Returns true only when the capture state is 'active'.

See

AudioInputProvider.isActive

Implementation of

AudioInputProvider.isActive

isReady()

isReady(): boolean;

Defined in: src/providers/input/MicrophoneInput.ts:276

Check whether the provider has been initialized and is ready for use.

Returns

boolean

true when initialize has completed and dispose has not yet been called.

Implementation of

AudioInputProvider.isReady

onAudio()

onAudio(callback): void;

Defined in: src/providers/input/MicrophoneInput.ts:410

Register a callback to receive audio chunks.

Parameters

Returns

void

Remarks

Only one callback can be registered at a time; subsequent calls replace the previous callback. Must be called before start() so chunks are delivered.

Each chunk is an AudioChunk containing the raw audio data, metadata, timestamp, and a monotonically increasing sequence number.

See

AudioInputProvider.onAudio

Implementation of

AudioInputProvider.onAudio

pause()

pause(): void;

Defined in: src/providers/input/MicrophoneInput.ts:361

Temporarily pause microphone capture.

Returns

void

Remarks

Delegates to AudioCapture.pause(), which suspends the AudioContext. No audio data is delivered while paused. Resume with resume().

See

AudioInputProvider.pause

Implementation of

AudioInputProvider.pause

resume()

resume(): void;

Defined in: src/providers/input/MicrophoneInput.ts:375

Resume microphone capture after a pause.

Returns

void

Remarks

Delegates to AudioCapture.resume(), which resumes the suspended AudioContext.

See

• AudioInputProvider.resume
• pause

Implementation of

AudioInputProvider.resume

start()

start(): void;

Defined in: src/providers/input/MicrophoneInput.ts:321

Start capturing audio from the microphone.

Returns

void

Remarks

Delegates to AudioCapture.start(callback), providing an internal callback that wraps raw ArrayBuffer chunks into AudioChunk objects with a timestamp and sequence number before forwarding them to the registered callback.

A callback must be registered via onAudio() before calling start().

See

AudioInputProvider.start

Implementation of

AudioInputProvider.start

stop()

stop(): void;

Defined in: src/providers/input/MicrophoneInput.ts:346

Stop capturing audio from the microphone.

Returns

void

Remarks

Delegates to AudioCapture.stop(). The provider can be restarted with start().

See

AudioInputProvider.stop

Implementation of

AudioInputProvider.stop

updateConfig()

updateConfig(config): void;

Defined in: src/providers/input/MicrophoneInput.ts:301

Update the audio input configuration at runtime.

Parameters

Returns

void

Remarks

Delegates to AudioCapture.updateConfig(). Configuration changes take effect on the next call to start(). They do not affect an active capture session.