implementations

Implementation of 3D Positioning

• With Web Technology
  • In Browser
    • With 3D Web Engines
    • With pure Web Audio API support
  • Server side
    • Web Audio API for Node
    • With 3D Web engines
• ALTERNATIVE?
  • Use Node add on or write one in C++
  • Write Java application

---

With Web Technology

🚨 Hard to implement server side

In Browser

With 3D Web Engines

• 3D sound:
  • three.js
  • playcanvas * only for speaker * test code
  • Babylon.js
    • also supports HRTF for headphone
    • HRFT supporting Libraries (npm modules):

📎 Babylon js test project

With pure Web Audio API support

• Implementing Binaural (HRTF) Panner Node with Web Audio API
• Tutorial with code example
  • 3D positioning with Audio Web API

Server side

⚡ Motivation: To process incoming coordinates and to communicate back to multiple sound output streams / devices

Web Audio API for Node

• only a few features of the Audio Web API are available for Node js

📃 Source: NPM module 'node-web-audio-api'

What's implemented:

• AudioContext(partially)
• AudioParam (almost there)
• AudioBufferSourceNode
• ScriptProcessorNode
• GainNode
• OscillatorNode (coming soon)
• DelayNode (coming soon)

var API = require('web-audio-api')
    , AudioContext = API.AudioContext
//, AudioListener =  = require('web-audio-api').AudioListener  // !!! is missing!
    , fs = require('fs')
    , context = new AudioContext
    , Speaker = require('speaker')

console.log('encoding format : ' + context.format.numberOfChannels + ' channels ; ' + context.format.bitDepth + ' bits ; ' + context.sampleRate + ' Hz')
context.outStream = new Speaker({
    channels: context.format.numberOfChannels,
    bitDepth: context.format.bitDepth,
    sampleRate: context.sampleRate
})

fs.readFile(__dirname + '/../resources/audio/rain.wav', function(err, buffer) {
    if (err) throw err
    context.decodeAudioData(buffer, function(audioBuffer) {
        var soundSource = context.createBufferSource()
        soundSource.connect(context.destination)
        soundSource.buffer = audioBuffer
        soundSource.loop = true
        soundSource.start(0)
    })
    context.listener.setPosition(20, -5, 0); // not available in node-audio-web-api
})

🚨 Dead End: missing AudioListener class implementation!! ==> not possible to change the position of sound server side by using Audio Web API

With 3D Web engines

🚨 Dead End (similar as above)

• are all created for usage in browser
  • need the browser's 'windows' object
• would also use the Web Audio API at the end

---

ALTERNATIVE?

Use Node add on or write one in C++

• using OpenAL
  • open al using add on which we can specify for using multiple devices:
  • code : https://github.com/jefftimesten/node-openal

Editing / Customize Add-on code

openal.cc:

// Enumerate OpenAL devices
	if (alcIsExtensionPresent (NULL, (const ALCchar *) "ALC_ENUMERATION_EXT") == AL_TRUE)
	{
		const char *s = (const char *) alcGetString(NULL, ALC_DEVICE_SPECIFIER);
		while (*s != '\0')
		{
			cout << "OpenAL available device: " << s << endl;
			while (*s++ != '\0');
		} 
		// ...

according to tip on stackoverflow:

change ALC_ENUMERATION_EXT to ALC_ENUMERATION_EXT_ALL

change ALC_DEVICE_SPECIFIER to ALC_ALL_DEVICES_SPECIFIER!!

Thus, we get a list of all available devices!

🚨 BUG!: Mac OSX has an OpenAL bug due to old version used by CoreAudio

and differentiation between different devices in NodeOpenALDevice.cpp :

// ... 

NodeOpenALDevice::NodeOpenALDevice() {
	// differentiate between different devices here
	// e.g. device = alcOpenDevice((const ALCchar *) "Built-in Output\0");
	device = alcOpenDevice(NULL);
    if(device==NULL) {
		std::cout << "cannot open sound card" << std::endl;
		return;
    }
};

//...

Write Java application

• using OpenAL