Sony Patents a Camera That Adjusts Its Own Light Capture on the Fly
Cameras have long struggled to capture bright sunlight and deep shadows in the same frame. Sony's latest patent describes a sensor that monitors its own light exposure and adjusts how it reads each pixel before a single image is even saved.
What Sony's self-adjusting image sensor actually does
Today's image sensors commit to a fixed sensitivity setting at the start of each exposure. That's why photos taken looking out a sunlit window often show a bright, washed-out exterior and a dark, underexposed interior at the same time.
Sony's patented design adds a second, always-on monitoring circuit alongside the main pixel. That circuit continuously watches the incoming light and sends a live signal telling the main pixel how sensitive to be. If the scene is very bright, it dials the sensitivity down; if it's dim, the sensitivity goes up, all within the same shot.
The result is a sensor that can hold detail in both the brightest and darkest parts of a scene simultaneously, which is the definition of high dynamic range capture, achieved here through hardware rather than software tricks like merging multiple exposures.
a radiation sensitive circuit configured to continuously convert a first photocurrent generated by incident radiation into a pixel voltage signal and outputs a gain control signal based on a voltage level of the pixel voltage signal; …
Translation: A built in sensor monitors incoming light levels constantly to decide how much to boost the picture.
How the two circuits control each other in real time
The patent describes a solid-state imaging device (the silicon chip inside a camera) built from two cooperating circuits per pixel group.
The first is a radiation-sensitive circuit that runs continuously, even during the main exposure. It converts incoming light into a voltage signal called VPR and, based on how high or low that voltage is, generates a secondary signal called VGC (a gain control signal). Think of this as a real-time light meter that never stops reading the scene.
The second is the active pixel circuit, which does the actual image capture. During the exposure window, it collects charge from incoming light. Crucially, its conversion gain (how much electrical signal it produces per unit of light collected) is not fixed. Instead, it is set dynamically by the VGC signal coming from the first circuit.
- High ambient light: the gain control signal lowers sensitivity, preventing overexposure in bright areas.
- Low ambient light: the gain control signal raises sensitivity, pulling detail out of dark areas.
- Mixed scenes: each pixel region can be adjusted independently, frame by frame.
This is all done in analog hardware at the sensor level, before any digital processing occurs.
An active pixel circuit integrates a second photocurrent generated by incident radiation in an exposure period and converts a resulting electric charge into a pixel output voltage VSL at a conversion gain controllable by the gain control signal VGC.
Translation: The main pixel array gathers light and adjusts its sensitivity dynamically using those signals.
What this means for cameras in high-contrast scenes
For anyone who has ever taken a photo of a person standing in front of a bright window and gotten a silhouette, this is the problem Sony is addressing at the chip level. Existing high dynamic range techniques typically shoot multiple frames at different exposures and blend them together in software, which can cause motion blur artifacts when anything in the scene moves. A sensor that adjusts its own gain in real time, within a single exposure, sidesteps that problem entirely.
Sony's steady investment in image sensor hardware puts this in a broader context. Sony Semiconductor Solutions supplies image sensors to a large portion of the smartphone industry, so improvements at this level could eventually appear in cameras far beyond Sony's own branded products. The practical payoff would be cleaner photos and video in tricky lighting without any extra processing overhead.
That makes this Sony's 80th filing in our Display patent coverage since May, a count that includes dual-pattern focus work and daisy-chained chip control.
Getting this into a real camera requires new physical hardware baked into the sensor itself, not a software tweak. Sony would need to add a small monitoring circuit next to every group of pixels, one that watches incoming light continuously and tells the sensor how to adjust its sensitivity on the fly.
The good news is that Sony already manufactures sensors at a level of precision few companies can match, which closes some of the gap between a working prototype and something that ships in a box.
If they clear that manufacturing hurdle, the reward is a camera that handles bright and dark parts of a scene at once, in a single exposure, without stitching multiple shots together afterward. That stitching is why photos of moving subjects sometimes look slightly ghosted today, and this design would make that problem disappear by nature rather than by workaround.
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
17 drawing sheets from US 2026/0270581 A1 · click any drawing to enlarge
Want this weekly breakdown for a company we don't cover? Patentlyze Pro →