Vibrating TV panel makes voices seem to come from speakers’ mouths

Show summary Hide summary

Researchers at the Korea Institute of Machinery and Materials (KIMM) have built a display that makes sound appear to come from the exact spot on screen where the action occurs. The prototype aims to eliminate the familiar mismatch between picture and sound on large displays, potentially changing how we experience movies, signage, and live visuals.

Why on-screen sound location matters

In everyday life we localize sound instinctively: a dropped object or a speaking person draws our attention because sound usually comes from the source. Most TVs and monitors, however, separate audio and visuals. Speakers or soundbars sit apart from the image, so voices or moving sounds often seem displaced.

Audio systems such as stereo or surround sound create convincing directionality, but they still rely on speakers positioned around the viewer. Precisely matching perceived sound location to moving on-screen events remains difficult, especially on very large panels.

Turning the panel into a speaker

KIMM’s approach places multiple actuators on the back of the display so the screen itself generates sound. The team built a roughly 100-inch prototype using a transparent LED film over an aluminum panel, driven by 80 voice coil motors (VCMs) mounted behind it.

Rear view of a display with multiple actuators mounted behind the panel
Actuators mounted behind a display vibrate the screen to produce localized sound.

Each motor vibrates the panel like a loudspeaker diaphragm, pushing air to produce sound. But rather than relying on a single vibrator, the system coordinates many actuators to concentrate acoustic energy in selected zones of the surface.

Engineering a movable sound field

Vibrations in a solid panel spread and interact, which makes localizing sound challenging. To address that, researchers studied how different panel areas respond to actuator inputs and used the Moore-Penrose pseudo-inverse to compute the voltages that produce a focused vibration field.

Rather than simply driving the actuator nearest a desired point, the system combines signals across many VCMs. That coordinated control creates a localized sound source that can be shifted in real time to follow moving objects on screen.

Stability and AI-based tracking

Real-time operation introduces extra complications. Overlapping vibrations can reinforce or cancel each other, and some actuators transmit energy more effectively than others. KIMM reduced instability with a selective excitation algorithm that omits poorly transmitting actuators from control calculations.

Acoustic camera heatmap showing localized sound spots on a screen
Acoustic-camera image showing how the prototype concentrates sound on specific screen areas.

For practical use, the display also incorporates AI for sound-source separation and localization. The machine-learning component identifies where a sound-producing object appears in the video, isolates its audio from background sources, and routes that sound to the matching area of the panel automatically.

Prototype performance

Tests on the prototype reported a high degree of precision. The team measured up to a 35 dB difference in vibration amplitude between active and suppressed zones and reported 99.9% accuracy in choosing intended sound-emitting areas. Acoustic-camera measurements showed the localized sound could track objects moving across the screen.

The researchers note the system performs best in the low- and mid-frequency bands, effectively localizing vibrations below 500 Hz.

Modular design and possible applications

KIMM adopted a modular-grid architecture so panels can be combined or separated depending on screen size. That design could scale the technology from conventional home TVs to large LED signage and media installations.

Driving multiple modules at once can reproduce stronger low-frequency effects, the team says, making the system useful for exhibitions, performances, media art, and in-vehicle displays as well as commercial digital signage.

Limits and open questions

While the prototype narrows the spatial gap between sight and sound, it does not necessarily replace dedicated loudspeaker systems. Reproducing very deep, high-output bass and the full dynamic range of a separate subwoofer-based setup remains a different engineering challenge.

Still, by aligning audible directionality with on-screen action, the technology promises a more natural and immersive experience on large displays. Dr. Dongjoon Kim, a senior researcher at KIMM’s AX Convergence Research Center, said the team plans to expand the approach to modular LED tile displays and develop it into a broadly applicable core technology.

Publication and institutional backing

The work appears in the Journal of Mechanical Science and Technology and was conducted under the National Research Council of Science & Technology framework.

Give your feedback

Be the first to rate this post
or leave a detailed review



MOTORCYCLE MAGAZINE is an independent media. Support us by adding us to your Google News favorites:

Post a comment

Publish a comment