The inherent directivity (narrowness) of all wave producing sources depends on little more than the size of the source, compared to the wavelengths it generates. Audible sound has wavelengths ranging from a few inches to several feet, and because these wavelengths are comparable to the size of most loudspeakers, sound generally propagates omnidirectionally. Only by creating a sound source much larger than the wavelengths it produces can a narrow beam be created. In the past, loudspeaker manufacturers have created large speaker panels or used reflective domes to provide some directivity but, due to the sound's large wavelengths, the directivity of these devices is still extremely weak.
No loudspeaker can ever approach the directivity of Audio Spotlight technology.
Since the goal is a small loudspeaker but strong directivity, the only possible solution is to generate very small wavelengths - such as those of high-frequency ultrasound. The ultrasound used in Holosonic technology has wavelengths only a few millimeters long, which are much smaller than the source, and therefore naturally travel in an extremely narrow beam.
Of course, the ultrasound, which contains frequencies far outside our range of hearing, is completely inaudible. But as the ultrasonic beam travels through the air, the inherent properties of the air cause the ultrasound to change shape in a predictable way. This gives rise to frequency components in the audible band, which can be accurately predicted, and therefore precisely controlled. By generating the correct ultrasonic signal, we can create, within the air itself, any sound desired.
Note that the source of sound is not the physical device you see, but the invisible beam of ultrasound, which can be many meters long. This new sound source, while invisible, is very large compared to the audio wavelengths it's generating. So the resulting audio is now extremely directional, just like a beam of light.
Often incorrectly attributed to so-called "Tartini tones", the technique of using high-frequency waves to generate low-frequency signals was pioneered over forty years ago. Over the past several decades, many others have attempted – and failed – to use this technique to make a practical audio source.
Through a combination of careful mathematical analysis and engineering insight based on pioneering work at MIT in the early 2000's, the patented Audio Spotlight sound system has become the very first, and still the only, truly directional audio system which generates high quality sound in a reliable, professional package.
The actual measured sound fields of our commercial Audio Spotlight models are below. All measurements were done in an anechoic (reflection-free) room, with laboratory-grade microphones at 1kHz.
These sound fields show the remarkable isolation ability of the Audio Spotlight technology. Even one step outside the beam, sound levels are reduced by over 90%!
A few traditional loudspeaker makers have promoted so-called "directional" sound sources, to mimic the unique technology and appeal of the Audio Spotlight. But the physics cannot support the claims - all other sound sources are forever limited to the "wavelength versus size" physics constraint. Making a bigger louspeaker does increase its directivity, but only slightly. How close can regular speakers get to the Audio Spotlight?
For formal theory, with mathematical proof, please read the whitepaper "Fundamental Limits of Loudspeaker Directivity". Below are actual, physical measurements of Audio Spotlight systems, and similarly-sized "directional" traditional loudspeakers.
This speaker array is marketed as being "highly focused" while providing "isolated listening", though actual field specifications are conspicuously omitted from their datasheets. A line array like this is expected to be slightly directive in the long axis, but completely omnidirectional about the short axis. Here's how it measures, compared to an Audio Spotlight speaker of similar size:
1D Sound Bar
1D Sound Bar (vertical)
Speaker panels, whether they are made from a single radiating element, an array of small loudspeakers, or the opening of a dome, are acoustically equivalent, and are simply the 2D form of the Speaker Bar mentioned above. All have the same maximum directivity. In this case, the horizontal and vertical dimensions are equal. Therefore, the sound field pattern is identical to that of the Sound Bar’s largest dimension, but for both horizontal and vertical angles.
This one is also marketed as being "highly directional". Here's how it compares to the Audio Spotlight technology:
Clearly, no loudspeaker technology can remotely approach the directivity and isolation abilities of the Audio Spotlight.
Because the physics of sound creation is so different than any traditional loudspeaker, Holosonics had to make substantial advancements in several areas to bring this technology to successful fruition. Each advancement independently doesn't create success - all of the elements must be present for a high-performance, useful technology.
The mathematics of nonlinear wave physics is quite complex, and the specific governing equations addressing the conversion of ultrasound to audible sound are not solvable in closed-form. Solving these equations is essential for knowing how to synthesize the correct ultrasound signal, in real time, such that the desired audio signal is reproduced accurately. Beginning with the pioneering work by Dr. Pompei at MIT, Holosonics has successfully developed the proper mathematical basis for the ultrasound-to-audio effect, as well as realtime signal processing algorithms necessary for practical implementation. In addition, Holosonics has pioneered several processing techniques for enhancing the conversion efficiency, and ensure that only the absolute minimum of ultrasound is needed for reproducing the desired audio output, and for developing fast, efficient computational techniques to implement these algorithms on inexpensive DSP's.
Even when armed with the correct mathematical formula for synthesizing the ultrasonic signal, the ultrasound still needs to be reproduced accurately. High-fidelity audio requires high-fidelity ultrasound.
This presented a formidable challenge to Holosonics, as the only transducers commercially available were the common sonar-style piezoelectric devices, used in the prototypes of the earlier researchers in this field. While these can create ultrasound, and can be used to generate audio, the bandwidth is quite limited, and fidelity of ultrasound - and therefore audio - is abysmally poor. This was one major reason that no other company has ever had a successful product in this market.
Holosonics developed its own custom, proprietary transducer technology to address these difficulties. Based on a thin moving film, rather than piezoelectric ceramics, the transducer is light, inexpensive, extremely efficient, and has unmatched ultrasonic bandwidth and fidelity. With over 10 years in continuous development, it is an important part of Audio Spotlight technology.
Traditional architectures for audio amplifiers do not have nearly enough high-frequency bandwidth to properly recreate ultrasound. Further, the electrical load (impedance) of the Holosonics transducer is vastly different than that of a traditional loudspeaker. Therefore, existing audio amplifier technology and design techniques could not be applied to Audio Spotlight technology. Starting from first principles, Holosonics developed its proprietary, digital, high-efficiency ultrasound amplifier specifically to create high-power ultrasonic signals extremely accurately, and drive its transducer technology efficiently. The amplifier is so efficient, no heatsinks are needed, and very little energy is wasted.
Covering the above major innovations, and several others, Holosonics has over 25 issued patents in the USA and Internationally, with many others pending.
The technique of using a nonlinear interaction of high-frequency waves to generate low-frequency waves was originally pioneered by researchers developing underwater sonar techniques dating back to the 1960's . These early acoustics researchers successfully derived the formal mathematical basis for this effect and developed innovative sonar systems with more directivity and bandwidth than would otherwise be available. They called this device a parametric array.
In 1975, the first publication  appeared which demonstrated that these nonlinear effects indeed occur in air. While these researchers had not attempted to reproduce audio, they nonetheless proved that such a device may be possible.
Over the next two decades, several large companies, including Matsushita (Panasonic), NC Denon, and Ricoh attempted to develop a loudspeaker based on this principle. A paper describing one attempt was published in 1983 . While they were successful in producing some sort of sound, problems with cost, feasibility, and extremely high levels of distortion (>50% THD) caused the almost total abandonment of the technology by the end of the 1980's.
While a graduate student developing '3D Audio' at Northwestern University in the late 1990's, Joseph Pompei had similar ideas of using ultrasound as a loudspeaker, largely to overcome deficiencies he saw with traditional methods of sound reproduction. After performing extensive research on the idea, he discovered the large body of knowledge in the field of nonlinear acoustics, as well as the earlier attempts at using ultrasound as an audible source. Soon after arriving at MIT, his insight led him to identify – and subsequently rectify – the barriers which had plagued the earlier researchers. Through a combination of careful mathematical analysis and solid engineering, he was able to construct the very first, and still only, practical, high-performance audio beam system .
Audio Spotlight systems have been in use in thousands of installations all over the world since 2000. Customers include American Greetings, Best Buy, Boston Museum of Science, Cisco Systems, the Field Museum, the Guggenheim, Harvard Peabody Museum, Jack Morton Worldwide, Kaiser Permanente, Motorola, Science World BC, Tate Modern, Walt Disney, Western Union and the Yale Art Gallery.
1 Westervelt, P. J., JASA v35 535-537 (1963)
2 Bennett, M. B., and Blackstock, D. T., JASA v57 562-568 (1975)
3 Yoneyama, M., et al., JASA v73 1532-1536 (1983)
4 Pompei, F. J., Proc.105th AES Conv, Preprint 4853 (1998)