StreamRecorder

WebSocket audio streaming microservice. Accepts raw WAV audio via WebSocket, converts to MP3 with ffmpeg, and forwards to the backend API.

How It Works

1. Client connects via WebSocket to /ws-stream with auth token and chat ID
2. Client streams raw audio data (WAV) as binary messages
3. Client closes the WebSocket when recording is done
4. Server converts WAV to MP3 in-memory via ffmpeg and forwards to backend API

Routes

Route	Description
GET /	Health page with API documentation
GET /ws-stream?auth=<token>&chatId=<id>	WebSocket endpoint

WebSocket Parameters

• auth — JWT token (required), sent as Bearer token to backend
• chatId — Chat identifier (required), alphanumeric/hyphens/underscores, max 128 chars

WebSocket Close Codes

• 1000 — Normal close or idle timeout
• 1009 — Stream exceeded maximum size

Configuration

Variable	Default	Description
BACKEND_URL	http://api:4000/messages	Backend API endpoint
MAX_FILE_SIZE	10000000	Max input size for ffmpeg conversion (bytes)
MAX_STREAM_BYTES	10000000	Max bytes accepted per WebSocket stream
CHUNK_TIMEOUT_MS	2000	Idle timeout between chunks before auto-close
FETCH_TIMEOUT_MS	30000	Timeout for backend API request
MP3_BITRATE	64k	FFmpeg MP3 encoding bitrate

Development

# Run with watch mode
deno task dev

# Run tests
deno task test

Docker

docker build -t stream-recorder .
docker run -p 8000:8000 -e BACKEND_URL=http://api:4000/messages stream-recorder

Client Example

const ws = new WebSocket(
  'ws://localhost:8000/ws-stream?auth=YOUR_TOKEN&chatId=YOUR_CHAT_ID'
);
ws.binaryType = 'arraybuffer';

const stream = await navigator.mediaDevices.getUserMedia({ audio: true });
const recorder = new MediaRecorder(stream, { mimeType: 'audio/webm' });

recorder.ondataavailable = (e) => {
  if (ws.readyState === WebSocket.OPEN) {
    e.data.arrayBuffer().then((buf) => ws.send(buf));
  }
};

recorder.start(250); // send chunks every 250ms

// Stop recording
recorder.stop();
ws.close();