Apple is reportedly advancing efforts to introduce a next-generation image sensor for upcoming iPhones, designed to deliver dynamic range capabilities that closely mimic human vision. According to MacRumors, Apple has submitted a patent titled “Image Sensor With Stacked Pixels Having High Dynamic Range And Low Noise,” outlining a cutting-edge sensor system built to deliver superior dynamic range and minimal noise.
The design incorporates layered silicon architecture, advanced multi-level light detection, and integrated noise control directly on the chip—potentially enabling the sensor to achieve up to 20 stops of dynamic range.
iPhone camera vs human eye:
For context, the human eye can perceive roughly 20 to 30 stops of dynamic range, influenced by how the pupil adapts and how light is interpreted over time.
In contrast, most modern smartphone cameras typically manage 10 to 13 stops. If Apple’s new sensor performs as envisioned, it could not only leap ahead of existing iPhone camera technology but even outshine high-end cinema cameras like the ARRI ALEXA 35.
About iPhone sensor:
According to the patent, the sensor is composed of two stacked layers. The upper layer, referred to as the sensor die, is responsible for capturing light, while the lower logic die performs tasks like exposure management and noise reduction, handling the image processing workload directly on the chip.
A key innovation in the sensor’s design is a technology known as the Lateral Overflow Integration Capacitor (LOFIC). This system enables individual pixels to adaptively store varying amounts of light based on scene brightness—within a single frame. As a result, the sensor can effectively manage extreme lighting contrasts, such as capturing both a subject and a sunlit background, without sacrificing detail in either shadows or highlights.
Another critical element of the architecture targets minimizing image noise and visual grain. Each pixel is equipped with its own dedicated memory circuit that actively detects and neutralizes heat-induced electronic noise in real time. Notably, this noise correction happens directly on the chip, well before any software-based image processing or saving takes place.
