This New Pixel Could Turn Screens Into Cameras

ETH Zurich researchers built a pixel that can both display and analyze light.

30 Jun 2026, 18:28 by Tudor Tarita · ZME Science

The Fourier pixel colored logo of ETH Zurich. The letter ‘E’ is roughly 1 millimeter tall on the camera. Credit: Glauser YM, Vonk SJW, et al.

Pixels are single-function. The pixels in your camera capture light, while the ones on your screen emit it. But what if a pixel could do both so you could record with your screen?

Now, researchers at ETH Zurich say they have built a new kind of pixel that does just that. The device, called a Fourier pixel, can emit light and read incoming light, including details ordinary pixels miss: brightness, phase, and polarization.

That could one day lead to camera-displays, holographic screens, and optical devices that adjust themselves as they sense the world.

How a Two-Way Pixel Works

The word “pixel” comes from “picture element,” a term that appeared nearly a century ago in the magazine Wireless World. Since then, pixels have defined modern life. On a phone or television, they create images. In a camera, they capture them. But they have remained specialists.

The ETH Zurich team, led by David Norris of the Optical Materials Engineering Laboratory, set out to break that split.

The new work builds on an earlier ETH Zurich study in which Norris’s group built “Fourier surfaces”—precisely shaped, wave-like surfaces designed with Fourier mathematics, a branch of mathematics that breaks complex patterns into waves. That earlier method gave the researchers a way to sculpt optical surfaces with nanometer-level accuracy.

In the new study, they used that fabrication approach to make a far more ambitious device: a pixel that can both generate and read complex light fields.

Watching Nanoscale Fluids Flow
Cicada wing destroys bacteria solely through its physical structure
New way of storing gas – the nanotech way
Economic inequality and decline fracture society, says a new study

A Fourier pixel turns incoming light into surface waves, guides them across a sculpted surface and scatters them back out as a chosen light pattern.. Credit:Glauser YM, Vonk SJW, et al.

Their Fourier pixel uses a carefully sculpted surface, shaped with tiny wave-like patterns only nanometers deep. When light reaches the chip, the pixel can turn it into a surface wave that travels across the material. At another point, that wave scatters back out as light. As the waves overlap, they interfere, reinforcing or canceling one another to form patterns, images, or beams.

Get smarter every day...

Stay ahead with ZME Science and subscribe.

Daily Newsletter
The science you need to know, every weekday.

Weekly Newsletter
A week in science, all in one place. Sends every Sunday.
No spam, ever. Unsubscribe anytime. Review our Privacy Policy.

Thank you! One more thing...

Please check your inbox and confirm your subscription.

That “full-field” control allows the pixel to handle three key properties of light: brightness, timing, and polarization. In other words, it can work with far more information than an ordinary pixel.

The study showed Fourier pixels generating vortex beams with doughnut-shaped centers, projecting ETH logos in different colors, focusing light into spots, and encoding two images in different polarizations. The same platform also worked in reverse, sensing phase and polarization from incoming light.

What Comes Next

The technology is still in its infancy. The team has so far made only very small arrays and full camera displays are still out of sight.

A modern screen or camera contains millions of pixels. The ETH group now wants to extend the method into matrices made of many Fourier pixels, closer to the architecture used in real devices.

“This ability is important because light can carry a lot of information if we can completely control all of its attributes,” Norris told Gizmodo.

The paper also points to possible uses in adaptive optics, holographic displays, optical communication, and quantum information processing. The researchers note that current sensors were designed for specific wavelengths, though clusters of Fourier pixels could collect color or spectral information.

For Norris, the achievement carries a simpler pleasure.

“Personally, I find the combination of simple math and precise fabrication extremely beautiful,” Norris told Gizmodo. “The math predicts some crazy wavy pattern for a specific optical output. We go and make that pattern on the surface and the device immediately creates the output that we want. In other words, math really works!”