photosites

A photosite is the light-sensitive region on a digital image sensor that converts incoming photons into an electrical signal. It is the basic unit of a pixel in both CCD and CMOS sensors, and its size largely governs sensitivity, noise performance, and dynamic range.

When photons strike a photosite, they generate electron-hole pairs. The electrons collected in the photosite's potential well form a charge proportional to light intensity. This charge is read out, amplified, and converted by an analog-to-digital converter to produce a digital pixel value. If too much light arrives, the well saturates, clipping the signal.

The efficiency of photon-to-electron conversion is described by quantum efficiency. The fraction of available photons that generate charge depends on wavelength and sensor design. In color sensors, a color filter array (often a Bayer pattern) samples red, green, and blue at different photosites, and demosaicing reconstructs full color.

In CMOS sensors, each photosite may include amplification and other circuitry in the same pixel, enabling lower power and faster readout; CCD sensors transfer charge across regions to a single readout node. The choice of architecture affects noise, speed, and image quality.

Factors influencing performance include fill factor, microlenses to funnel light onto the photosite, and anti-blooming structures to limit charge spillover. Noise sources include shot noise, dark current, and read noise. The dynamic range reflects the ratio between the largest non-saturated signal and the noise floor.

Photosite size, or pixel pitch, balances resolution and light sensitivity: smaller photosites yield higher resolution but lower light-gathering capacity. Advances aim to increase quantum efficiency, reduce noise, and extend dynamic range for photography, video, and machine vision.