Tech Page: Getting Directions

PUT SOUNDS JUST WHERE YOU WANT IT
Author:
Publish date:
Image placeholder title
Image placeholder title

Fig. 1: The Audio Spotlight uses the interaction of ultrasonic sound waves with the air to reproduce audible sounds in a highly directional pattern. By contrast, a conventional speaker radiates sound in a wide pattern, often with side lobes that the Audio Spotlight avoids.

In most situations, electronic musicians want speakers to radiate sound evenly throughout a listening area, delivering a similar sonic experience to everyone within earshot. And most speakers do just that. But there are certain circumstances in which a highly directional beam of sound would be desirable — for example, playing dance music that only those in a certain area of a large space could hear, allowing others outside that area to talk without having to shout over the music.

This is the idea behind the Audio Spotlight from Holosonic Research Labs (holosonics.com). Based on underwater-sonar research from the 1960s, the Audio Spotlight was developed by company founder Dr. Joseph Pompei in the 1990s.

The original sonar research uncovered an interesting phenomenon: ultrasonic sound waves interact in a nonlinear manner with the medium in which they travel, distorting the original waveform. This distortion can include audible components, essentially transforming the medium itself into a highly directional speaker. The trick is to modulate the ultrasound so that the desired audio is heard.

Several large companies, including Matsushita (parent of Panasonic), Denon, and Ricoh, have tried to develop an audio speaker based on this principle, but they couldn't fully overcome problems such as cost, power, and high levels of distortion (which were greater than 50 percent THD). Pompei took up the challenge as a graduate student at Northwestern University and MIT, where he solved the problems that had plagued earlier attempts.

The Audio Spotlight starts with a normal audio signal, which is then processed in a manner that is precisely the inverse of the nonlinear distortion expected during playback. Next, a broadband ultrasonic carrier in the 65 kHz range is amplitude-modulated with the processed signal, amplified, and sent to an ultrasonic transducer, which converts the signal into sound waves. As the ultrasound interacts with the air in front of the transducer, the resulting nonlinear distortion demodulates the sound and counteracts the processing performed on the original signal, leaving just the intended audio, which the company claims is accurate to within a few percent THD.

But why is the sound so directional? Basically, the area of interaction between the ultrasound and air — which is long and narrow due to the directivity of high frequencies — acts like a very large end-fire line array of speakers, which is very directional because the size of the source is large compared with the audible wavelengths it produces. Another, less accurate analogy is a shotgun microphone, whose pickup pattern is highly directional because it's basically an end-fire array of mic capsules.

The nonlinear demodulation starts a small distance from the transducer, and the intended audio can be heard quite a bit farther away (see Fig. 1). By contrast, a conventional speaker radiates sound in a very wide pattern, often with side lobes (also depicted in Fig. 1). Side lobes do not exist in Audio Spotlight systems; the beam is much like that from a flashlight.

I asked Pompei if the Audio Spotlight could be used in a 2-channel sound system, but he says this is unnecessary. Because there is no room interaction, Pompei claims that sound from the Audio Spotlight tends to be perceived as “enveloping” the listener, almost like wearing headphones, reducing the need for 2-channel reproduction from regular speakers.

The potential applications are many and varied, from limiting the area in which dance music can be heard to individual audio kiosks in a museum or store to audio billboards that can be heard only in a particular spot on the street. Although it might not be directly applicable to studio or live sound reinforcement, it's a very interesting technology nonetheless.