This post is also available in:
עברית (Hebrew)
Speakers have a basic privacy problem: sound spreads. Play music, an announcement or a conversation aloud, and people nearby can usually hear it too. Highly directional audio systems exist, but producing a narrow beam of sound typically requires complicated arrays containing many ultrasonic emitters, making the equipment more expensive and difficult to manufacture.
Researchers at Pohang University of Science and Technology (POSTECH) have developed a simpler alternative. Called MiPAL, the experimental speaker uses a single sound-producing element combined with specially engineered metamaterials to direct audible sound toward one listener while reducing leakage to people nearby.
The system is based on a parametric array loudspeaker, which uses ultrasound rather than directly projecting ordinary audible frequencies. Audible information is encoded onto high-frequency ultrasonic waves. As the beam travels through the air, nonlinear acoustic effects generate the sound that a listener ultimately hears.
Because ultrasound can be directed much more narrowly than conventional speaker output, the technique can create a focused audio zone. The problem is that single-emitter systems can still produce unwanted vibrations that send sound sideways.
The speaker tackles both parts of the problem with two different metamaterial structures.
According to Interesting Engineering, at the front of the speaker is an acoustic metasurface that acts somewhat like a lens for sound, shaping the ultrasonic output into a narrow beam aimed toward the intended listener. Behind the transducer, elastic meta-units suppress unwanted mechanical vibrations that would otherwise create lateral sound leakage.
Together, the structures allow a single emitter to provide directional audio without requiring the large arrays traditionally used for parametric speakers.
The researchers demonstrated sound across approximately 500 Hz to 10 kHz, covering more than four octaves of the audible range. In simulated aircraft-seating experiments using music and broadcast audio, the elastic structures substantially improved isolation between neighboring seats.
The metamaterials are also 3D printed, allowing engineers to modify their geometry for different environments. Designs could potentially be adjusted for varying seat spacing, listening angles or installation requirements without redesigning the entire speaker.
The technology could eventually support personalized entertainment on aircraft and public transportation, museum audio guides or announcements directed at specific individuals without requiring headphones.
There are also potential security applications. Directional audio could provide instructions or alerts to selected personnel in command centers, checkpoints or other sensitive facilities while limiting how widely the message is audible. However, practical security would depend on how effectively sound leakage can be controlled outside laboratory conditions.
The speaker remains a research-stage technology, but its main contribution is straightforward: instead of filling a room with sound and asking listeners to isolate themselves with headphones, engineer the sound field so the audio travels primarily toward the person who is supposed to hear it.
