Sony Patents a Chip That Captures Sharp Photos and Distance Data Simultaneously

By Patentlyze Team · Updated Jul 3, 2026

Most phone cameras need separate hardware to figure out how far away things are. Sony's latest patent describes a single stacked sensor that captures a full color image and a depth map at the same time, using the same lens array.

What Sony's stacked image-and-depth sensor actually does

Imagine trying to take a portrait photo where the background goes blurry, or using your phone to scan a room for AR furniture placement. Both of those tasks require your camera to know how far away everything is, not just what color it is. Right now, phones usually handle that with extra sensors or software tricks, which adds cost and complexity.

Sony's patent describes a sensor built like a three-layer sandwich. The top layer captures the regular color image you'd recognize as a photo. The middle layer captures the same incoming light but reads it differently to figure out distance. The bottom layer is a processing chip that handles both sets of data at once.

Because all three layers share the same set of lenses and are physically stacked on top of each other, the color image and the depth map are captured from exactly the same point of view, at exactly the same moment. That kind of precise alignment is hard to achieve when you rely on two separate sensors sitting side by side.

How the three substrates split light into image and depth signals

The patent describes a stacked photodetector built from three bonded silicon substrates, all sharing a single array of lenses at the top.

• First substrate (top layer): Contains standard image pixels, each with a photodiode that responds to the portion of light passing through the lens array designated for color imaging. These output the signals that become your regular photograph.
• Second substrate (middle layer): Contains depth pixels, each with a photodiode tuned to a different portion of the incoming light, including some or all of the same light the image pixels see. These output signals used to calculate how far away objects are, likely via time-of-flight or phase-detection principles.
• Third substrate (bottom layer): A dedicated logic chip with two separate processing circuits: one converts the image pixel signals into final image data, and the other converts the depth pixel signals into a depth map.

The key claim is the stacking order: lenses on top, image pixels next, depth pixels below that, and the processor at the base. Because both sensing layers look through the same lens array, there is no parallax error (the slight offset you get when two cameras sit apart from each other), which makes the depth data much easier to align with the color image.

What this means for cameras, AR, and 3D sensing

For anyone who uses portrait mode, AR apps, or 3D scanning on a phone, the quality of depth sensing directly affects how realistic the result looks. A depth sensor and a color sensor that are physically separated always see the world from slightly different angles, which forces software to compensate. A single stacked sensor eliminates that problem at the hardware level.

Sony Semiconductor Solutions is one of the largest image sensor suppliers in the world, selling chips to Apple, Samsung, and most major phone makers. A patent like this points toward next-generation sensors where high-resolution photography and accurate 3D sensing are handled by a single, compact component rather than a cluster of separate modules.

Editorial commentary on a publicly published patent application. Not legal advice.