tts-loader

Very basic Webpack loader for text-to-speech.

This loader uses text-to-speech functionality to generate audio files with the contents of a passed-in text file.

Requirements

This loader currently supports macOS's built-in text-to-speech functionality (on Macs only) and Google's unofficially supported text-to-speech API.

In addition to the standard Webpack requirements, ffmpeg is required. On a Mac, it's easiest installed with Homebrew:

brew install ffmpeg

For other platforms, check out ffmpeg's Downloads page.

Example usage

Assuming you have a file called 'quickbrownfox.txt' with the text you wish to have spoken:

var quickBrownFox = require('tts-loader!quickbrownfox.txt');

That will return a array of file paths that can be used, for example, with Howler:

var foxHowl = new Howl({ src: quickBrownFox });
foxHowl.play();

Currently, the loader outputs the result in MP3 and OGG format.

Configuration

For greater configuration, it's recommended that you use a special naming convention for your text files and use the Webpack config files.

macVoice: A string to define which macOS voice to use. To get a list of the macOS voices on your system, run say -v '?' in Terminal. This defaults to the voice you have selected in your System Preferences.

drivers: An array of drivers to attempt. This defaults to ['mac', 'google'].

Sample configuration

{
    test: /\.speech\.txt$/,
    loader: 'tts-loader',
    query: {
        drivers: ['google', 'mac'],
        macVoice: 'Fiona'
    }
}

License

MIT license. See LICENSE.