Sony Patents a Camera That Sees Close and Distant Objects Simultaneously
Most depth-sensing cameras have to choose between seeing objects up close or far away. Sony's new patent describes a chip that tackles both at once by splitting the sensor into two cooperating halves.
What Sony's split-sensor depth measurement actually does
Imagine you're trying to take a photo that needs to know the exact distance to both your face and the wall behind you at the same time. Current depth-sensing cameras often struggle here: the settings that work well for nearby objects tend to fall apart at longer distances, and vice versa.
Sony's patent describes a sensor chip split into two sections. One half uses a slow signal suited for measuring longer distances, and the other half uses a fast signal suited for closer objects. They work in parallel, taking turns so neither half sits idle, meaning the camera can capture a full picture of depth across the whole scene in one pass.
The result is a depth sensor that doesn't have to compromise. You get useful data across a wide range of distances without needing multiple separate cameras or switching modes between shots.
A time-of-flight image sensor circuitry, comprising an imaging unit including a first imaging portion and a second imaging portion, wherein the circuitry is further configured to: apply a low-frequency demodulation signal having a first state and a second state …
Translation: The camera uses a specialized sensor that switches between different signal speeds to measure how far away objects are.
How the two imaging halves trade demodulation signals
This patent covers a time-of-flight (ToF) image sensor, a type of depth camera that works by firing light pulses and measuring how long they take to bounce back. The time gap tells the camera how far away each object is. The challenge is that the signal frequency you use for this measurement is a tradeoff: low frequencies (slow pulses) can measure long distances accurately, while high frequencies (fast pulses) give finer detail at close range but get confused at distance.
Sony's design splits the imaging chip into two portions. At any given moment:
- The first imaging portion receives a low-frequency demodulation signal, which processes reflected light tuned for longer-range depth measurement.
- The second imaging portion simultaneously receives a high-frequency demodulation signal, handling closer-range detail.
The low-frequency signal has two states (think of them as two phases of a cycle). When the first portion is in phase one, the second portion runs the high-frequency signal. When the first portion flips to phase two, the arrangement shifts again. This interleaving means both halves are always active and no measurement time is wasted.
Demodulation here just means extracting the timing information from the returning light pulses. By demodulating two different frequencies in parallel across the split sensor, the circuitry captures both near and far depth data in a single image frame rather than taking two sequential exposures.
… when the low-frequency demodulation signal is applied in the first state to the first imaging portion, the high-frequency demodulation signal is applied to the second imaging portion.
Translation: The sensor splits its workload by sending different types of timing signals to two separate parts of the camera at once.
What this means for cameras that need to see depth clearly
Depth sensors are in more devices than most people realize: face-unlock cameras on phones, gesture controls, AR headsets, robotics, and automotive systems all rely on accurate distance mapping. The longstanding technical frustration is that a single frequency setting forces engineers to pick a range they care about most and accept worse performance everywhere else. A chip that genuinely handles both at once, without extra hardware, is a meaningful engineering fix for a problem that's been a quiet bottleneck in sensor design.
Sony Semiconductor Solutions is already one of the world's dominant suppliers of image sensor chips, so depth-sensing improvements from their labs tend to flow into a broad range of consumer and industrial products. This particular filing sits alongside other sensor and chip-level work tracked in the latest Big Tech patents covering depth sensing and computational imaging, a field where the gap between what sensors can theoretically do and what they actually ship doing has been stubbornly wide.
This is the 18th Sony filing we've tracked since July in our self-driving sensing race, building on earlier applications like an adjustable-focus scanner and one on flash-recovery LiDAR.
Depth-sensing cameras fail at a basic task remarkably often: measure something close and something far away at the same time, accurately. That gap costs real money in recalled safety systems, degraded photography, and the extra components manufacturers add just to paper over it.
Sony's answer is to divide the sensor into two halves working simultaneously, each tuned to a different distance range, so one chip does what previously required more hardware or slower operation. The match between problem and approach is tight, because the fix lives at the chip level rather than in software patches or mechanical add-ons bolted on afterward.
The honest unknown is whether the two halves disturb each other in ways that erode accuracy once the sensor leaves the lab. That question belongs to manufacturing, not to this document, and it is the only thing standing between a clever design and a broadly useful one.
There are more where this came from
We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.
The drawings
9 drawing sheets from US 2026/0251796 A1 · click any drawing to enlarge
Want this weekly breakdown for a company we don't cover? Patentlyze Pro →