espeakng.js

Browser-based text-to-speech using eSpeak-ng. No servers, no APIs, pure JavaScript.

Features

• 🎯 Vanilla JavaScript - No frameworks or dependencies
• 🔊 Basic Audio Processing - Simple enhancements for clearer speech
• 🌍 100+ Languages - All eSpeak-ng voices included
• ⚡ Runs Offline - Everything works in the browser
• 📦 ~3 MB Total - Includes all voice data

Audio Processing

The library includes basic audio enhancements:

• High-frequency boost - Gentle emphasis for slightly clearer speech
• Noise gate - Simple threshold-based silence removal
• Volume control - Adjustable output level

Quick Start

// Initialize TTS
const tts = new SimpleTTS();

// Generate and play speech
tts.onReady(() => {
  tts.speak('Hello world!', (audioData, sampleRate) => {
    SimpleTTS.playAudioData(audioData, sampleRate);
  });
});

Installation

Download these files:

1. espeakng-simple.js - Simple wrapper API
2. espeakng.worker.js - Web Worker
3. espeakng.worker.data - Voice data (must be in same folder as worker)

Or use the minified version:

1. espeakng.min.js - Minified API
2. espeakng.worker.js and espeakng.worker.data

API

new SimpleTTS(options)

• options.workerPath - Path to worker (default: 'js/espeakng.worker.js')
• options.defaultVoice - Default voice (default: 'en')
• options.defaultRate - Default speech rate (default: 175)
• options.defaultPitch - Default pitch (default: 50)
• options.defaultVolume - Default volume (default: 1.0)
• options.enhanceAudio - Enable basic audio enhancement (default: false)

tts.speak(text, [options], callback)

tts.speak('Hello', {
  voice: 'en',      // Voice ID (see tts.getVoices())
  rate: 175,        // Speech rate (80-450)
  pitch: 50,        // Pitch (0-100)
  volume: 1.0,      // Volume (0-2.0)
  enhance: false    // Apply basic audio enhancement (default: false)
}, (audioData, sampleRate) => {
  // audioData: Float32Array with audio samples
  // sampleRate: Sample rate (currently 11025 Hz)
});

tts.getVoices(callback)

tts.getVoices(voices => {
  // Array of voice objects with identifier, name, and languages
  console.log(voices);
});

Helper Functions

// Play audio data directly
SimpleTTS.playAudioData(audioData, sampleRate);

// Create Web Audio API AudioBuffer
const buffer = SimpleTTS.createAudioBuffer(audioData, sampleRate);

Known Issues

• Sample Rate: The worker reports 22050 Hz but audio is actually 11025 Hz. The wrapper handles this automatically.
• Voice Names: Use simple voice codes like 'en' instead of 'en-us' for better compatibility.

Example

const tts = new SimpleTTS();

tts.onReady(() => {
  // List available voices
  tts.getVoices(voices => {
    console.log('Available voices:', voices);
  });
  
  // Simple speech
  tts.speak('Hello world!', (audioData, sampleRate) => {
    SimpleTTS.playAudioData(audioData, sampleRate);
  });
  
  // Speech with options
  tts.speak('Hello world!', {
    voice: 'en',
    rate: 200,
    pitch: 75,
    volume: 1.2
  }, (audioData, sampleRate) => {
    SimpleTTS.playAudioData(audioData, sampleRate);
  });
});

Demos

• simple-demo.html - Interactive demo with voice selection and controls
• demo.min.html - Minimal demo using minified version
• compare-demo.html - Compare with/without basic audio enhancement

Original API

For direct access to the eSpeak-ng worker, use espeakng.js:

const espeak = new eSpeakNG('js/espeakng.worker.js', () => {
  espeak.list_voices(voices => console.log(voices));
  espeak.set_voice('en');
  espeak.synthesize('Hello world', (samples, events) => {
    // Raw PCM samples
  });
});

License

GPLv3 - Based on eSpeak-ng