Sony has unveiled a new type of stacked CMOS sensor that uses "two-layer transistor pixels" to double the light gathering capability. Typical image sensors have the light-sensitive photodiodes and pixel transistors that control and amplify the signal on the same layer. However, the new design puts the photodiodes on top and the pixel transistors below, "approximately doubling saturation signal levels," Sony said.
Sony pioneered stacked sensors that put fast memory and other electronics directly under the sensor, allowing for faster readout speeds and thus rapid burst shooting and reduced rolling shutter (jello effect) on cameras and smartphones. This latest sensor uses a similar idea, but packs the pixel transistors onto a separate substrate underneath the photodiode layer.
That means each layer could be optimized, allowing Sony to double the sensor's light saturation (well depth), or the amount of charge each pixel can hold. That in turn allows for around double the light-capturing capability.
Sony notes that because the transistor pixels sit on a separate layer, it was able to boost the amplifier transistors in size. That allows for a bigger signal boost, reducing noise when shooting nighttime or other images in dark locations. The increased dynamic range will allow for "high-quality, low-noise images even in low-light," according to Sony.
Sony specifically stated that the tech will allow for higher-quality smartphone photography. With double the light gathering capability, it will allow for much improved light sensitivity even in relatively small, high-megapixel sensors. Sony has yet to say when this tech will make it to smartphone or cameras, but it plans to further improve the design for both large and small sensors.