Currently, when viewing shared devices such as televisions, the audio volume must be set globally for all viewers. Both users with hearing impairment and users who prefer quieter sounds must agree upon a single global volume. Personal Sound Beam solves this problem by directing a narrow beam of sound to the user, allowing for custom, local volume control. We use an audio phase array speaker in order to achieve the local volume control. This technology also allows for custom audio experiences in a variety of applications such as in retail environments.

How it Works

Phased array speakers consist of multiple speakers placed a consistent distance apart. By delaying when the sound hits a particular speaker, we are able to alter how loud you hear the sound in each one, allowing for an immersive auditory environment. As shown in the figure below, depending on the frequency being played, you’ll hear different volumes depending on where you stand.

Alt text-1

If we shift the phase, by introducing a time delay to certain speakers, we can adjust the perceived intensity at a particular angle, as shown in the figure below. We can make it appear like something is moving past you by continually adjusting the time delay through each speaker.


We placed nine speakers inside of cardboard holders and placed them in a horizontal line. These were attached to a desktop computer which has a PCI breakout for high volume audio throughput. The desktop runs a program in C# which acts as the game manager and the server for every audio object. Each audio object is a python script which exchanges data over UDP in order to handle phase shifting for a specific audio file.

Challenges we ran into

Our speakers required external amplifiers in order for people to hear them. We had developed amplifiers using parts we had, but there was too much distortion. The parts lab at Drexel had the components, but it took us longer than expected to complete the amplification circuit (12 hours straight of soldering).

Accomplishments we’re proud of

There is an easy to use Python API that we developed which makes it easy to extend the work we did to a variety of audio setups and use cases.

What we learned

  • Read the datasheets
  • Boxes are good for carrying things
  • Radar technology is hard at 4:00AM

Built With

Share this project: