Acapela

Node module for supporting Acapela voices in an Electron app. Outside of Electron your mileage may vary. Windows only.

Requirements

Built on top of Acapela's speech engine. You'll need to raeach out to Acapela to get the needed libraries and license files. AcaTTS.dll, AcaTTS.64.dll and AcaTTS.ini are all required to be in the main directory of your app (not the library's directory, the Electron app's root folder). The app also expects to have the required /bin and /data directores in the same directory.

To compile the app you'll first need to get license files from Acapela and put those in the library's /include directory. Rename the .h file to license.h and then you can run node-gyp rebuild or however you build modules.

Usage

npm install https://www.github.com/coughdrop/acapela.git

The easiest way to use the library is to require extra-tts.js in the acapela module. If you require it in the app process then you can do things like the following:

var extra_tts = require('acapela/extra-tts.js');

extra_tts.getAvailableVoices({success: function(list) {
  console.log(list);
}});

extra_tts.speakText({
  voice_id: "<voice id from the list>",
  text: "hello my friends, I am speaking to you",
  success: function() {
    console.log("done speaking!");
  }
});

extra_tts.stopSpeakingText();

Chrome Memory Issue

Chrome apparently doesn't like downloading and unzipping really large files in a browser process, and on older computers also seems to struggle with the acapela libraries, it'll sometimes result in the entire window turning black. You can get around this in electron by requiring extra-tts.js on the main process, and extra-tts-ipc.js on the rendering process. You'll also need a little extra code on the main process, something like:

ipcMain.on('extra-tts-exec', function(event, message) {
  var sender = event.sender;
  var opts = JSON.parse(str);
  var args = opts.args;
  opts[0].success = function(res) {
    sender.send('extra-tts-exec-result', JSON.stringify({
      success: true,
      callback_id: opts.success_id,
      result: res
    });
  };
  opts[0].progress = function(res) {
    sender.send('extra-tts-exec-result', JSON.stringify({
      progress: true,
      callback_id: opts.progress_id,
      result: res
    });
  };
  opts[0].error = function(err) {
    sender.send('extra-tts-exec-result', JSON.stringify({
      error: true,
      callback_id: opts.error_id,
      result: err
    });
  };
  extra_tts[opts.method].apply(extra_tts, args);
});

ipcMain.on('extra-tts-ready', function() {
  event.sender.send('extra-tts-ready', 'ready');
});

License

MIT License

TODO

• add examples (in the mean time, you can see how we're using it in the TTS section here, https://github.com/CoughDrop/coughdrop/blob/master/app/frontend/app/utils/capabilities.js)
• specs (stop judging me, I'm not a native app developer)