Google synthesis

Builds network request for Google Speech API V1 or Google's translation's speech synthesis API. Based of Zlargon's 'google-tts'.

Installation

npm install --save googlesynthesis

Usage

Constructor

// Import module.
const GoogleSynthesis = require('googlesynthesis');
// Create an instance.
const googleSynthesis = new GoogleSynthesis(
  console // Logger used for developing purposes, optional.
);

Methods

// Gets the url requests for getting the synthesized audio.
let urls = googleSynthesis.request(
	'Hello world', // Text to be synthesised.
	'en-GB', // Language code, default 'en-GB'.
	'UK English Female', // Name of the voice to be used, default 'Uk English Female'.
	0.5, // Pitch, default 0.5.
	0.5, // Playback speed, default 0.5.
	1 // Volume, default 1.
	);

Returns an array of urls since it only allows for a maximum of 500 characters per request.

// Gets the url requests for getting the synthesized audio, using the translate API.
let urls = googleSynthesis.requestTranslate(
	'Hello world', // Text to be synthesised.
	'en-GB', // Language code, default 'en-GB'.
	'1' // Playback speed, default 1.
	);

Returns an array of urls since it only allows for a maximum of 200 characters per request. To see which languages are supported see the Google Cloud documentation.

Various extra methods you most likely won't have to deal with

// Slices transcript into sections.
let slices = googleSynthesis.slice('Hello world', 8);
console.log(slices);

// Gets a key from translate.google.com.
googleSynthesis.key(function(key) {
  console.log(key);
});

Result of slices will be: ['Hello','world!']

Events

// Emitted when an error occured.
googleSynthesis.addEventListener('error', function(error) {
  console.error('Error', error);
});
// If you have set getKey to true in the constructor,
// then this event will emit when it has retrieved one.
googleSynthesis.addEventListener('key', function(key) {
  console.log('key', key);
});

Example

The following example is made for electron so the Web Audio API is available.

// Initialize module, see constructor section for more information.
const GoogleSynthesis = require('googlesynthesis');
const googleSynthesis = new GoogleSynthesis();

// Audio player.
const audio = new Audio();

// After this event is called the service can be used.
googleSynthesis.addEventListener('key', function(key) {
  // Get urls for the phrase 'Hello world!'.
  let urls = googleSynthesis.request('Hello world!');
   
  // Setup listener so it cycles through playing each url.
  let index = 0;
  audio.addEventListener('ended', function() {
    index++;
    if (index >= urls.length) {
      audio.removeEventListener('event', this);
	  return;
    }
    audio.src = urls[index];
    audio.play();
  });
  // Set first source.
  audio.src = urls[index];
  audio.play();
});

For another example see the Electron-VoiceInterfaceBoilerplate's output.js.

Troubleshooting

If the module suddenly stops working it might be because of several reasons

If you are using 'request()', you might have exceeded the maximum number of request to the speech API. You currently can't increase this limit.
If you are using 'requestTranslate()', Google changed the method of creating the required token for each request had changed.

RedKenrok/node-googlesynthesis