Open Source Binaural Beats

Screenshot from Android Version

My first smartphone apps were these basic tools to allow users to create their own binaural beats. The app was called Binaural, and this past December, I made both the iOS and Android versions free for anyone to download and use. Now, I've decided to take this a step further and make the core of these apps open source! This means that you can go to my github page and download fully working sample apps that generate binaural beats and isochronic pulses. Then, you can use these algorithms in your own apps. If you find any of this useful, I'd appreciate you give me shout, and perhaps I'll feature your app on my Guided Meditation Treks blog!

Here's the source code for iOS.
Here's the source code for Android.

Or you can just download the app and not worry about how it works!

---

​Generating Binaural Beats With a Computer

Binaural beats are simply two discrete sine waves, each played in different ears. So, all the app needed to do is to generate a sine wave of a specific frequency in one ear, then a separate sine wave of a different frequency in the other. There are many different ways to generate sine waves, depending on our operating system and how you approach the problem, but the underlying idea is still the same. 

All I had to do was:

1. Generate two sine waves: one for each ear
2. Allow the user to manage the amplitude of the waves (for volume if needed)
3. Use a stereo "render" method, so that the left ear and right ear can have a different sine wave
4. Allow the user to manage the "carrier frequency" and "binaural frequency," then use these values to calculate the frequencies for the left and right sine wave:
   1. ​Left ear sine wave frequency = carrier frequency - (binaural frequency / 2)
   2. Right ear sine wave frequency = carrier frequency + (binaural frequency / 2)​
5. Since the human ear perceives higher frequencies as louder (for example, an 800hz tone is perceived as twice as loud as a 400hz tone at the same amplitude), I added a compensation to automatically adjust the amplitude when a user changes to frequency, based on the frequency they are changing from and the frequency they are changing to. So, if a user changes from 100hz to 200hz, I cut the amplitude in half, and when going from 200hz to 100hz, i double it.

​Generating Isochronic Tones With A Computer

​Unlike Binaural Beats, isochronic tones are a single mono sine wave. The difference is that this sine wave's amplitude is modulated across time. For example, an isochronic tone at 8hz would turn on and off 8 times per second. Therefore, the render algorithm looks a little bit different, and essentially fades the sine wave in and out very quickly.

What I had to do:

1. Generate a single mono sine wave, which is played into both ears at the same time
2. The sine wave's frequency equals the carrier frequency entered by the user
3. Using 48,000 samples per second, divide that in half, then divide it again by the user-entered "isochronic frequency." For example, if we have an isochronic frequency of 1 hz and 48,000 samples per second, then when we hit sample 0, the sine wav is ON and when we hit 24,000 samples, the sine wave is OFF. If our frequency is 2hz, then at sample 0 it's on, then at 12,000 it's off, then at 24,000, its on again, then at 36,000 its off again.
4. Due to rounding, I had to tweak it a little bit in case these numbers above didn't match perfectly at the top of each second.
5. I found that when I abruptly turned the sine wav off or on in the middle of its period, this would cause artifacts that resulted in a "chirping" sound. So, I had to add faders so that the wave was not abruptly turned on and off, thus preventing chirping by smoothly fading in and out during each period.
6. The same amplitude compensation for tone frequency changes as noted above.

Relevant Methods in iOS

For iOS, I started with some code from Matt Gallager to generate sine waves using Apple's Audio Units. This was my starting point for this app, though, in my later apps like the sleep app, I used a totally different method. Anyway, in Binaural, the  audio unit is generated in the RenderTone static classes of the two viewControllers, and the actual playback of the audio unit takes place in the renderTone() method of each of the respective viewControllers. You'll notice that the algorithms of RenderTone are significantly different between Binaural Beats and Isochronic Tones, based on the explanations above, but the only major difference between the two renderTone() methods is that binaural beats uses two channels per frame, and isoChronic uses only one.

Relevant Methods in Android

For Android, again, I started with some sample code to generate sine waves using android's AudioTrack object. Then, I made the relevant changes listed above. The android project has two relevant classes that contain the construction of these tones: Binaural.java and Isochronic.java Everything regarding the algorithms is nicely modularized in those two classes!

Here's the source code for iOS.
Here's the source code for Android.

Or you can just download the app and not worry about how it works!