Pixiq

Digital image recordings contain a variety of non image-forming signals collectively known as noise. Typically, these unwanted artifacts take the form of white or randomly colored speckles. They can also occur as areas of colored blotches or band-like irregularities. These signals are most noticeable in smooth, evenly colored sections of a picture, especially the darker portions.

Certain shooting conditions in motion photography can increase the visual prominence of noise; in particular, setting high ISOs in order to use fast shutter speeds to freeze action. High ISO settings amplify the image-forming signal, and that also amplifies the appearance of noise as illustrated in the photo above, taken at ISO 3200. As a generalization, noise becomes a problem at ISO 1600 or above with D-SLRs, and above ISO 400 with point-and-shoot models. DSLR cameras perform better because they have comparatively larger light-collecting photosites on their sensors, resulting in a lower proportion of recorded noise signals. Another typically “noisy” condition occurs when long exposures are used to produce blurred motion effects. Long exposures measured in seconds tend to generate higher levels of heat at the sensor, leading to a condition known as thermal noise.

Higher noise levels also tend to occur when photos are underexposed, because photographers become overly cautious in trying to safely preserve all-important highlight detail. They will expose their image so that highlight areas fall short of the right axis on the recording’s histogram, making underexposure more probable. The result is a reduction in the amount of the image-forming light for shadows and darker areas, and thus a more prominent level of recorded noise. One strategy to avoid this is to “expose to the right.” This involves setting the exposure so that the highlights with details on the camera’s histogram are as close as possible to the right without clipping. This often results in a slightly overexposed scene, but with lower levels of recorded noise in shadow and middle tone areas. This slight overexposure can then be corrected in the computer.

There are also other approaches. For example, many D-SLRs have noise-reduction settings for use with long exposures. And then there are specialized noise-filtering programs that can be applied when processing the image file on the computer. Some of the most popular programs among photographers include Neat Image by Neat Image, Picture Code’s Noise Ninja, Nik Software’s Define, and Imagenomic’s Noiseware.