We use displays to present, check and adjust photos or videos. They are systems, specifically non-linear systems, with which digital image data can be displayed. Due to the nonlinearity, the digital luminance values of the input side are not in a proportionately linear relationship to the playback side. Considering it graphically, these systems produce a curve progression, which is referred to as the characteristic line. The gradient of the linear portion of this curve is defined as the gamma value.
Assume there is a picture signal with a linear behavior on the display (image top left), as would be the case for RAW images. The incoming data is depicted by the display. In this context, the input signal is charged with the characteristic curve of the display. Considering it graphically, the diagonal converts into a curve (image top right). The digital luminance values of the input side are thus in a non-linear relationship to those of the monitor output. The result is that the image appears quite dark as seen on the monitor.
The linearity on the output side must be preserved such that a true perception of the reproduction is ensured. In the case of a non-linearly operating display, it is necessary to compensate for the system’s distortion. This is achieved through the gamma correction of the linear input signal. This adjustment results in a non-linear behavior for the input signal (image bottom left), which is the inverse to that of the monitor (image bootom middle). Calculated together, the linearity on the output side of the monitor remains (image bottom right).
Many cameras can create images in RAW format. The generated data can be developed with a gamma correction by using different image editing programs and adjusted accordingly. The camera also uses a non-linear gamma correction for the internal generation of JPG formats. In addition to the true brightness appearing on a non-linearly operating display, the correction of the input signal provides an additional benefit. While the RAW data concede a bit depth 12 to 14 bits, JPG images have 8 bits. Instead of 4096 brightness levels (12 bit), the result with 8 bits is a maximum of 256 levels.
This is considerably less.
The cost of not using gamma correction in the generation of JPGs means a distorted perception of the brightness levels through the 8-bit conversion, which sometimes leads to a “step” effect. Adjusting the brightness distribution by means of adding in the gamma correction means that these effects are compensated.