Visual microphone listens using light instead of sound

Electronics

Technological Innovation Website Editorial Team - August 1, 2025

Schematic of a sound detection system using a low-cost visual microphone. The PDA is a single-pixel sensor, a photodiode. The DMD is a digital micromirror, a commercial component. [Image: Xu-Ri Yao/Beijing Institute of Technology]

Visual microphone

Researchers have created a microphone that listens using light, rather than directly detecting sound.

Unlike normal microphones, this visual microphone picks up tiny vibrations on the surface of objects caused by sound waves, and then transforms them into audible signals.

"Our method simplifies and reduces the cost of using light to capture sound, while also enabling applications in scenarios where traditional microphones are ineffective, such as conversations through a glass window," said Professor Xu-Ri Yao of the Beijing Institute of Technology in China. "As long as there is a passage for the light, sound transmission is not necessary."

It's a true technological marvel, but it's also one that reminds us that privacy is an increasingly unattainable issue. But the team assures that the photonic microphone can also be used for good.

"New technology could change the way we record and monitor sound, bringing new opportunities to many areas, such as environmental monitoring, security, and industrial diagnostics," Yao said. "For example, it may be possible to speak to someone trapped in an enclosed space, such as a room or a vehicle."

Researchers reconstructed audio signals by imaging the vibration of a paper card (ac). The results were improved with a signal processing filter to enhance the high-frequency (df) component. [Image: Xu-Ri Yao/Beijing Institute of Technology]

Hearing sound with light

Although several methods for detecting sound with light have been developed, these solutions require sophisticated optical equipment, such as lasers or high-speed cameras. The idea here was to use a computational imaging approach known as single-pixel imaging to develop a simpler and more cost-effective approach that makes optical sound detection technology more accessible.

Instead of the millions of pixels of traditional cameras, single-pixel imaging captures images using just one light detector—a single pixel. This makes it impossible to record an entire image at once, but the device is smaller, simpler, and potentially offers much greater sensitivity.

To achieve this, the light in the scene is modulated using structured, time-varying patterns, which is accomplished by a device called a spatial light modulator . The sensor—the single pixel—measures the amount of light modulated for each pattern. A computer then uses these measurements to reconstruct information about the subject being photographed.

To apply this single-pixel imaging technique to sound detection, the team used a high-speed spatial light modulator, capable of encoding the light reflected by a sound-impacted surface. The sound-induced vibrational motion on the surface causes subtle changes in the intensity of the reflected light, and these changes are captured by the single-pixel detector.

But, since the objective is not to photograph the object, the variations detected in the light are decoded into audible sound.

"The combination of single-pixel imaging with Fourier-based localization methods allowed us to achieve high-quality sound detection using simpler equipment and at a lower cost," said Yao. "Our system enables sound detection using everyday items, such as paper cards and sheets, under natural lighting conditions, and does not require the vibrating surface to reflect light in a specific way."

A sheet of paper also works, but with slightly worse results than paper. [Image: Wei Zhang et al. - 10.1364/OE.565525]

Tests and little data

To demonstrate the new visual microphone, the researchers tested its ability to reconstruct the pronunciations of numbers in Chinese and English, as well as an excerpt from Beethoven's To Elise .

A paper card and a plant leaf were used as vibration targets, positioned 0.5 meters apart, while a nearby speaker played the audio.

The system reconstructed clear and intelligible audio, with the paper card producing better results than the foil. Low-frequency sounds below 1 kHz were accurately recovered, while high-frequency sounds above 1 kHz exhibited slight distortion, although a signal processing filter improved the results.

Another advantage of using a single-pixel detector to record light intensity information is that it generates a relatively small amount of data. This means the data can be easily downloaded to storage devices or uploaded to the internet in real time, allowing for long-term or even continuous sound recordings—in testing, the data rate was 4 MB/s.

Other news about:

More topics