Sony Patents an Image Sensor That Knows When to Stop Collecting Light
Most camera sensors collect light for a fixed window of time and hope for the best. Sony's new patent describes a sensor that watches what it's collecting and stops the moment something meaningful shows up.
What Sony's self-stopping image sensor actually does
Imagine a rain gauge that automatically seals shut the moment it catches enough water to give you a reading, instead of sitting open until a timer goes off. That's roughly the idea here.
A standard camera sensor gathers light for a set period, then hands off whatever it recorded. Sony's patent describes a photosensitive element that actively watches the incoming light during collection and cuts itself off as soon as it detects a specific signal. You get the reading the instant it arrives, not after waiting for a clock to expire.
This kind of behavior is especially useful in systems that send out pulses of light (like a lidar depth sensor on a phone or car) and need to time exactly when that light bounces back. Getting the timing right at the sensor level, rather than in software afterward, can make the whole measurement faster and more precise.
How the sensor detects a signal and cuts off exposure
The patent describes a photosensitive element (a single light-detecting cell, or a pixel in a larger array) that does two things at once: it integrates incoming light (meaning it accumulates the energy of photons over time, building up a measurable charge) and it monitors that accumulation for a recognizable signal.
Once that signal is detected, the element stops integrating immediately. The key word in the patent is when rather than after, the shutoff is event-driven, not timer-driven.
This is different from standard rolling-shutter or global-shutter designs, where every pixel in the array collects light for the same fixed duration regardless of what shows up. Here, each element can potentially stop at its own moment.
- Integrate incoming light into a charge over time
- Monitor the integrated signal for a target pattern or threshold
- Stop integration immediately upon detection
The practical result is a sensor whose exposure duration is determined by the physics of the scene, not by a predetermined timer, which matters most in time-of-flight (distance-measuring) applications where nanosecond-level timing accuracy translates directly into millimeter-level distance accuracy.
What this means for depth sensors and future cameras
Time-of-flight depth sensors are already inside phones (Face ID, portrait-mode depth mapping) and cars (lidar for driver assistance). The tighter the timing on when a reflected laser pulse is detected, the more accurate the distance reading. A sensor that stops collecting the instant it catches the signal avoids the ambiguity of figuring out when within a fixed window the signal actually arrived.
For Sony Semiconductor Solutions, which supplies image sensors to a large share of the smartphone industry, a more precise time-of-flight architecture could flow into next-generation depth cameras, industrial scanners, or medical imaging gear. The patent's claim is written at a fairly broad, foundational level, which means this could end up as a building block inside many different product types rather than one specific device.
The core claim here is narrow enough that it reads more like a foundational building block than a finished invention. Sony is essentially trying to own the concept of an event-triggered shutoff at the pixel level, which is a genuinely useful idea for depth sensing but not the kind of thing that directly translates to a specific product announcement.
Which company should we read for you?
We track 17 companies here. Pro is the same weekly breakdown for any company you choose, delivered privately. Type a name and we'll scope it and send you a quote.
Get one Big Tech patent every Sunday
Plain English, intelligent commentary, no hype. Free.
Editorial commentary on a publicly published patent application. Not legal advice.