At UCLA's Samueli School of Engineering, Professor Aydogan Ozcan and his team have solved one of the most stubborn puzzles in 3D display technology: how to project sharp, detailed images across multiple depth layers without them bleeding into one another. Using a combination of AI-trained software and precisely engineered optical surfaces, they've created a system that can encode and display 28 separate image layers in a single snapshot—a breakthrough that could transform everything from augmented reality glasses to advanced medical imaging.
The challenge they tackled is deceptively simple to state but fiendishly difficult to solve. When light travels through space to create volumetric displays, images meant for different depths tend to interfere with each other, a problem called diffraction-induced crosstalk. The closer you pack those depth layers together, the worse the degradation becomes. For truly immersive AR/VR experiences and holographic displays to work, viewers need accurate depth cues across the entire scene, yet current technology struggles to deliver that fidelity when layers sit only micrometers apart.
Ozcan's team designed an ingenious two-part system. First, a digital encoder built on a Fourier-based neural network analyzes the target image stack and extracts features across multiple scales and frequencies. This encoder then creates a single phase pattern—essentially a sophisticated optical instruction manual—that encodes all 28 images simultaneously. That pattern then passes through a passive diffractive optical decoder: structurally optimized surfaces that need no power and perform what the researchers call "depth-dependent field programming." As light propagates through these surfaces, they physically route image content to its designated depth while suppressing information leakage between layers.
The results, published in Light: Science & Applications, are striking. In numerical simulations, the team demonstrated that axial plane separations could be as tight as a single wavelength of light. They then built a physical prototype with a two-layer optical system operating in the visible spectrum and proved the concept works in practice. The measured intensity patterns matched their simulations closely and dramatically outperformed a baseline system without the diffractive decoder.
What makes this genuinely important is the practical scalability. Dr. Çağatay Işıl, Alexander Chen, Yuhang Li, and their colleagues didn't just create a one-off laboratory curiosity—they characterized the design factors that matter: decoder depth, diffraction efficiency, spatial light modulator resolution, and axial encoding density. That information provides a roadmap for engineers building future systems.
The implications ripple outward. Holographic displays could finally move from science fiction toward real products. Near-eye AR/VR headsets could deliver the depth perception that makes users feel comfortable rather than nauseous. Medical researchers could build volumetric microscopes that reveal three-dimensional cellular structures more clearly. Real-time 3D visualization and volumetric optical computing—using light itself as a computing medium—suddenly seem within reach.
Looking ahead, the team plans to extend the framework toward multispectral operation and multiperspective holography, with an eye toward fabricating those diffractive decoder surfaces directly rather than assembling them layer by layer. The result: compact, energy-efficient 3D displays that could finally make immersive digital experiences feel as natural as looking at the world itself.