One optimization technique is to expose to the right. Note that by underexposing the shot, the photographer deprives himself of the best definition ranges of his camera. Areas of lower brightness are no longer recorded and all result in black pixels. The dynamic range of the photographed scene is divided into 12 equal fractions but the tonal range recorded in the first zone is expressed in 2048 light shades and the last (2048 times less bright) is expressed to one shade step almost black and twice as much almost black. The second contains half of the rest and so on. The first contains half of all the possible shades, or 2 12 / 2 (4096 / 2). The information captured on 12 bits is divided into 12 zones called Exposure Value ( EV). The maximum number of zones, from white to black, depends on the bit depth of the sensor (typically 12 bits, 14 for some cameras). The most subtle details of the bright areas are rendered while the definition of the dark areas is reduced to a few variations of hues. The mathematical consequence implies that each area receives half as many photons as the previous one and, therefore, contains half as much information. We classify the result of this capture in bright areas according to fractions (½, ¼, etc.) defining the tonal range of the photo. Thus some photosites are “filled” according to the maximum exposure capacity and others less. The exposure is determined so that they collect just enough before saturation, the clipping. I synthesize the various fundamental concepts of photography in a common presentation.Įach photosite of the sensor (each defining a pixel) captures photons.

Let us digress (optional reading) to return to sensor operation, how it collects light and why it results in noise.

Dxo prime