CompositeVoicev0.1.1

NullOutput

No-op audio output provider that discards all audio.

Defined in: src/providers/output/NullOutput.ts:91

No-op audio output provider that discards all audio.

Remarks

NullOutput implements the Null Object pattern for the 'tts' + 'output' pipeline roles. Every method is a no-op, making it safe to use wherever an AudioOutputProvider is required but audio playback is not needed.

Common use cases:

Text-only agents — NullInput + LLM + NullOutput = ChatGPT-style text interface
Voice-in, text-out — MicrophoneInput + STT + LLM + NullOutput = user speaks, agent replies with text only
Server-side pipelines where only the transcription/LLM response matters
Testing pipelines without audio playback side-effects
Headless environments without audio hardware

Example

import { NullOutput, CompositeVoice, BufferInput, DeepgramSTT, AnthropicLLM } from 'composite-voice';

// Minimal server-side pipeline
const voice = new CompositeVoice({
  providers: [
    new BufferInput({ sampleRate: 16000, encoding: 'linear16', channels: 1 }),
    new DeepgramSTT({ apiKey: '...' }),
    new AnthropicLLM({ apiKey: '...', model: 'claude-haiku-4-5' }),
    new NullOutput(),
  ],
});

// Audio from TTS is silently discarded
await voice.initialize();

See

AudioOutputProvider for the interface contract
BrowserAudioOutput for the browser-side counterpart
BufferInput for the server-side input counterpart

Implements

AudioOutputProvider

Constructors

Constructor

new NullOutput(): NullOutput;

Returns

NullOutput

Properties

Property	Modifier	Type	Default value	Description	Defined in
`roles`	`readonly`	readonly `ProviderRole`[]	`undefined`	Pipeline roles covered by this provider. Remarks `NullOutput` covers both `'tts'` and `'output'`, so no separate TTS provider is needed. LLM text is still emitted via `llm.chunk` and `llm.complete` events — it just isn’t synthesized into speech.	src/providers/output/NullOutput.ts:108
`type`	`readonly`	`ProviderType`	`'rest'`	Communication type for this provider. Remarks `NullOutput` uses `'rest'` because it does not maintain any connection.	src/providers/output/NullOutput.ts:98

Methods

configure()

configure(_metadata): void;

Defined in: src/providers/output/NullOutput.ts:249

Configure the output format. No-op — audio is discarded.

Parameters

Parameter	Type	Description
`_metadata`	`AudioMetadata`	Ignored.

Returns

void

Implementation of

AudioOutputProvider.configure

connect()

connect(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:177

No-op — nothing to connect.

Returns

Promise<void>

Remarks

There is no TTS WebSocket or service to connect to in text-only mode.

disconnect()

disconnect(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:199

No-op — nothing to disconnect.

Returns

Promise<void>

Remarks

No TTS connection was established, so there is nothing to close.

dispose()

dispose(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:132

Dispose of the provider.

Returns

Promise<void>

Remarks

Clears the initialized flag. No resources need to be released.

Implementation of

AudioOutputProvider.dispose

enqueue()

enqueue(_chunk): void;

Defined in: src/providers/output/NullOutput.ts:258

Enqueue an audio chunk. No-op — audio is discarded.

Parameters

Parameter	Type	Description
`_chunk`	`AudioChunk`	Ignored.

Returns

void

Implementation of

AudioOutputProvider.enqueue

finalize()

finalize(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:167

No-op — nothing to finalize.

Returns

Promise<void>

Remarks

There is no TTS engine with buffered state, so finalization is a no-op.

flush()

flush(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:265

Wait for playback completion. Resolves immediately — nothing to play.

Returns

Promise<void>

Implementation of

AudioOutputProvider.flush

initialize()

initialize(): Promise<void>;

Defined in: src/providers/output/NullOutput.ts:122

Initialize the provider.

Returns

Promise<void>

Remarks

Sets the initialized flag. No resources are acquired since all operations are no-ops. If already initialized, this is a no-op.

Implementation of

AudioOutputProvider.initialize

isPlaying()

isPlaying(): boolean;

Defined in: src/providers/output/NullOutput.ts:295

Check whether audio is playing. Always returns false.

Returns

boolean

false — NullOutput never plays audio.

Implementation of

AudioOutputProvider.isPlaying

isReady()

isReady(): boolean;

Defined in: src/providers/output/NullOutput.ts:142

Check whether the provider has been initialized.

Returns

boolean

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

Implementation of

AudioOutputProvider.isReady

onAudio()

onAudio(_callback): void;

Defined in: src/providers/output/NullOutput.ts:212

No-op — no audio will be produced.

Parameters

Parameter	Type	Description
`_callback`	(`chunk`) => `void`	Ignored.

Returns

void

Remarks

The callback is accepted to satisfy the interface but is never invoked because no TTS audio is synthesized in text-only mode.

onMetadata()

onMetadata(_callback): void;

Defined in: src/providers/output/NullOutput.ts:225

No-op — no metadata will be produced.

Parameters

Parameter	Type	Description
`_callback`	(`metadata`) => `void`	Ignored.

Returns

void

Remarks

The callback is accepted to satisfy the interface but is never invoked because no TTS audio is synthesized.

onPlaybackEnd()

onPlaybackEnd(_callback): void;

Defined in: src/providers/output/NullOutput.ts:313

Register a playback-end callback. No-op — playback never ends.

Parameters

Parameter	Type	Description
`_callback`	() => `void`	Ignored.

Returns

void

Implementation of

AudioOutputProvider.onPlaybackEnd

onPlaybackError()

onPlaybackError(_callback): void;

Defined in: src/providers/output/NullOutput.ts:322

Register a playback-error callback. No-op — no errors can occur.

Parameters

Parameter	Type	Description
`_callback`	(`error`) => `void`	Ignored.

Returns

void

Implementation of

AudioOutputProvider.onPlaybackError

onPlaybackStart()

onPlaybackStart(_callback): void;

Defined in: src/providers/output/NullOutput.ts:304

Register a playback-start callback. No-op — playback never starts.

Parameters

Parameter	Type	Description
`_callback`	() => `void`	Ignored.

Returns

void

Implementation of

AudioOutputProvider.onPlaybackStart

pause()

pause(): void;

Defined in: src/providers/output/NullOutput.ts:279

Pause playback. No-op — nothing is playing.

Returns

void

Implementation of

AudioOutputProvider.pause

processChunk()

processChunk(_text): void;

Defined in: src/providers/output/NullOutput.ts:157

No-op — text is discarded, no synthesis occurs.

Parameters

Parameter	Type	Description
`_text`	`string`	Ignored.

Returns

void

Remarks

There is no TTS engine, so text chunks are silently discarded. This exists to satisfy the duck-type validation for the 'tts' role.

resume()

resume(): void;

Defined in: src/providers/output/NullOutput.ts:286

Resume playback. No-op — nothing is paused.

Returns

void

Implementation of

AudioOutputProvider.resume

sendText()

sendText(_text): void;

Defined in: src/providers/output/NullOutput.ts:189

No-op — text is discarded.

Parameters

Parameter	Type	Description
`_text`	`string`	Ignored.

Returns

void

Remarks

There is no TTS engine to receive text for synthesis.

stop()

stop(): void;

Defined in: src/providers/output/NullOutput.ts:272

Stop playback. No-op — nothing is playing.

Returns

void

Implementation of

AudioOutputProvider.stop

synthesize()

synthesize(_text): Promise<Blob>;

Defined in: src/providers/output/NullOutput.ts:238

No-op — returns an empty blob.

Parameters

Parameter	Type	Description
`_text`	`string`	Ignored.

Returns

Promise<Blob>

An empty Blob.

Remarks

There is no TTS engine, so synthesis always returns an empty Blob.