A parametric array does not generate sound directly – it uses ultrasonic waves to reproduce sound. Ultrasonic waves modulated with an audio signal are generated by ultrasonic transducers. As they pass through air or encounter objects, they de-modulate and sound can be heard.
Diffraction of ultrasonic waves is much lower than sound waves because of their higher frequency. So the ultrasonic waves – and the sound coming from the parametric speaker – spread very little and are confined to a narrow beam that can reach tens of meters.
Neurotechnology has implemented innovative hardware solutions that overcome some of the major obstacles that have prevented widespread adoption of parametric sound systems. De-modulation is a non-linear process, and complex audio signal preprocessing must be implemented in order to invert this non-linear process and achieve high-quality and high-volume sound. Researchers at Neurotechnology developed algorithms that address these issues and designed patent pending ultrasonic transducers specifically suited for parametric sound systems. This led to the creation and release of the first Focusonics speaker: Model A.
With Neurotechnology's extensive experience in computer vision, we can also offer custom solutions that use the Focusonics directional speaker in combination with person detection and recognition for targeted advertising, personalized greeting services or automated warning systems. For example, a camera can be used to detect a person passing by or entering a shop and determine their sex, age or other attributes. The Focusonics speaker can then be used to transmit a personalized advertising message to this person without disturbing others.
PatentsO. Putkis, G. Vanagas, M. Mikolajunas and D. Virzonis. WO2020026005 - Method for generating parametric sound and means for carrying out said method. International patent application.