• Instagram
  • LinkedIn
  • Theresa Ambat

Making Music with... code?

Updated: Dec 13, 2019

This quarter I had the opportunity to do a quarter-long independent study through the DXARTS (Digital Arts and Experimental Media) department at the University of Washington. This study consisted of learning about real-time sound synthesis/processing, MIDI and OSC messaging, and audio spatialization using the Ambisonic Tool Kit.

I composed this piece using a programming language called SuperCollider.

I wanted to explore the idea of "moving water," by creating a computer music piece, manipulating synthesizer parameters using data from the motion of water. My original intent was to do a live performance using bowls of water, accelerometers, and recordings of a choral piece. Each bowl of water would have an accelerometer in it and I would move the water with my hands as if it were an instrument. The data from the moving accelerometers would then be inputs to granular synthesis parameters (fundamental frequency, frequency envelope, and quality), which would then modulate the vocal recordings.

Accelerometer + Arduino for capturing water data.

Sadly, this did not end up being my final product as I struggled to get the technology working. With the help of Marcin Paczkowski (a research scientist at DXARTS), I was able to get Arduino and SuperCollider to communicate, but I struggled to implement the data well enough to produce a musically quality piece. Not only that, I had a limited amount of time to waterproof the hardware itself. Thus, I decided to put aside the accelerometer idea for later so I could focus on the composition more.

Despite this fallback, I still wanted to incorporate the idea of "moving water" and chant into my piece. I wrote a short choral piece on Psalm 42:2--"As the deer longs for streams of water, so my soul longs for you, O God." This was my first time writing a choral piece, but with the help of my experience in Catholic church choirs + my music theory class, I was able to write a pretty nice sounding chant (at least, I think it sounds nice).

Of course, if I was going to compose computer music piece about water, I needed to have sound samples of water. One Saturday I spent the day recording various sounds of water in bottles and pots around my house (I even recorded my toilet flushing), as well as samples of a small creek a few blocks away from my apartment.

After recording my sounds, I threw them all into SuperCollider, and wrote the piece over the course of about 3 weeks. I even got to play around with the Ambisonic Tool Kit and listened to my piece in the DXARTS 3D sound lab. So, instead of just presenting my piece in stereo, it was presented in 3D. One of my favorite parts of the piece is when there's a sort of "shooting star" sound that travels upward in front of you, which is then followed by rain sounds that travel down the speakers all around you.

Working in the DXARTS 3D sound lab.

I found this project to be a really rewarding experience. I had never worked on a school project for such a long amount of time (10 weeks) and so this was a great way for me to exercise time management. A few things I hope to do with this next quarter would be to implement accelerometer data to simulate the motion of water in sound (again, not necessarily the sound of water itself, but its kinetic properties) write more of the Psalm 42 chant, explore the various sound worlds in the piece for longer periods of time, use different types of synthesis besides granular, and be more intentional about using the Ambisonic Tool Kit in my composition. You can see the process for creating it on my instagram story.

You can listen to the piece on my BandCamp. Please listen to this on headphones for the best experience. Of course, you won't get the same "3D" effect as if you were in the 3D sound lab, but I've decoded it to binaural to simulate the experience over headphones.