A full exposition of the ultrasonic technology used for directional sound is in the ICMC ’07 paper submission attached to this proposal. In brief, an array of ultrasonic speakers creates a beam of ultrasound which remains focused after being emitted. The ultrasound acts as a carrier for audible sound, which is ‘rectified’ by the air after the ultrasonic beam has propagated about a meter or so. Amazingly, the audible sound continues to move along the beam path, not spreading out as sound from a normal speaker would. This technique has become the basis for two commercial products as well as being an active area of research at DXARTS (Center for Digital Arts and Experimental Media), where we are working to develop an open-source, software version that has major advantages as a research tool to define new applications such as mobile wearable versions. That being said, once the system is optimized using the software version, as hardware or hybrid solution is envisioned for maximum portability in a wearable device in the short term as our current software-based platform runs on a Mac laptop running Supercollider 3, a powerful sound creation and manipulation programming language.

One of our current prototypes is a ‘Sound Glove’ that can be worn by a dancer or performer to project directional sound in an intuitive manner. Since the signal processing required for this application is simple we propose to do the encoding in a small hardware device and extend it so the dancer can use multiple arrays in a body suit to project sound in multiple directions simultaneously. Multiple channels of sound will be generated using our studio at DXARTS in Seattle by a musician and engineer who can incorporate sounds sent by the internet or cell phones as well as transmitted by a broadband link transmitting video and sound from the audience in Seoul. The multi-channel sound signal is then sent to Seoul where they are demultiplexed and transmitted by RF link or Bluetooth, and from there are encoded and sent to the speaker arrays. Due to the sheer amount of data and low latency required, we will require a minimum of at least 100Mb/sec bandwidth between DXARTS and the Art Center Nabi. We also require at least one (preferably two) live camera feed to be sent to DXARTS, with reasonably high fidelity stereo sound. The visual representation of the spiral laryngeal elements is produced by retroreflective patches attached to the body suit which become evident when stuck by lights behind the audience. The elements thus become visible as the arrays near them are directed towards the audience, providing the illusion that they are producing the sound.