How does a digital camera turn light into a picture?

Behind a digital camera’s lens is an image sensor covered with tiny light-sensitive sites. The lens forms an image across this array. During an exposure, photodiodes convert absorbed light into electrical charge. More light generally produces more collected charge, until the site reaches its capacity.

Circuits read the charge as an electrical signal. An analog-to-digital converter assigns it a number that can be stored and processed. Photons are tiny packets of light. In the invented example, half of 400 incoming photons produce collected electrons, giving 200 electrons. The assigned capacity is 300 electrons, mapped to the top of a 0–255 number scale. The linear rule therefore gives 200 ÷ 300 × 255 = 170. Above capacity, extra light cannot increase the stored amount, so bright regions can lose detail through clipping.

Color needs another step. Many cameras place red, green or blue filters over individual photosites. A common Bayer pattern has one red, two green and one blue filter in a repeating tile. Each site measures light through its own filter, rather than separately measuring all three color components at exactly the same point.

Processing combines neighboring samples to estimate a full-color image, then adjusts the result for display or storage. This reconstruction is powerful, but it has limits: very fine repeating patterns can create false colors. A photograph is therefore both a measurement of light and a carefully processed interpretation of those measurements.