Kell factor (technology)
The Kell factor , also known as the K factor , is an experimentally determined factor in the field of raster graphics , and derived from it in video technology and television technology , which indicates by how much higher a scanned image signal from an image sensor must be at a minimum in order to be used in image display devices such as a monitor with discrete picture elements ( pixels ) Stördarstellungen due to beat to be minimized. It lies in the value range between 0.5 and 1. The value 0.5 corresponds exactly to sampling with the Nyquist frequency (half the sampling frequency ), the upper limit value of 1 corresponds to the sampling frequency. It is determined through practical tests and selected as small as possible in order to minimize the required number of pixels per unit of time for a given bandwidth of the image signal in the discrete system. In signal theory, the Kell factor describes a type of undersampling .
The Kell factor is named after Raymond Davis Kell , who carried out the first experiments at Radio Corporation of America (RCA) in the 1930s and determined an experimental value of 0.64. However, Kell did not succeed in describing its experimental setup in a comprehensible manner. The later published work of Raymond Kell together with his colleagues A. Bedford and G. Fredendall in 1940 comes to a factor of 0.85.
In the case of display devices with fixed pixels, such as liquid crystal displays (LCD monitors) and CCD sensors as image recording devices, a comparatively high cell factor of 0.9 may be necessary in order to allow an almost interference-free display. Digital high definition television (HDTV) generally works with a basement factor of 0.9. In the case of cathode ray tubes , such as were previously used in television sets and in analog television transmission processes, a Kell factor of 0.7 is sufficient.
General
If a signal is sampled with a sampling frequency at least twice as high as the highest spectral components in the signal, according to the Nyquist-Shannon sampling theorem , an error-free reconstruction of the continuous signal course from the discrete individual values is always possible. This case would correspond to a value of 1 or more of the Kell factor, which in this case does not correspond to a reduction in the number of pixels per unit of time, which is why values of 1 and greater are irrelevant. The goal is to minimize the number of pixels required per unit of time with the same bandwidth, so only values less than 1 are important.
In a Kellfaktor less than 0.5, this corresponds to a sampling frequency which is smaller than the top spectral components in the signal, it comes to the occurrence of image frequencies , or aliasing , which make a correct reconstruction impossible.
In the range from 0.5 to 1, the range in which the specific Kell factor can be selected, overlay effects occur as a result of the phase position of the image signal in relation to the phase position of the sampling frequency, which are noticeable as a beat . These interferences vary depending on the image content and the type of display device. The image on the right serves as an example, which consists of alternating white and black vertical lines, so its upper limit frequency corresponds exactly to the sampling frequency, which is equal to a Kell factor of 0.5. The vertical lines in the image are slightly inclined towards the right edge of the image, which leads to different phase positions in relation to the exactly vertical scanning points. This inclination leads to a beating, which develops as a mixture, in this case these are gray areas, of different shapes. Since these gray areas are not part of the original picture content, they act as a disturbance. With a Kell factor of 0.66, shown in the second figure, these disturbances are still present, but much weaker. This corresponds to the trade-off between reducing the number of pixels on the one hand and the lowest possible interference on the other.
The picture disturbances resulting from a too low selected Kellfactor are similar to picture disturbances with the moiré effect , but not to be confused with them.
Individual evidence
- ^ RD Kell, AV Bedford, GLFredendall: A Determination of the Optimum Number of Lines in a Television System . RCA Review 5, July 1940, p. 8 to 30 .
literature
- Charles Poynton: Digital Video And HDTV - Algorithms and Interfaces . Morgan Kaufmann, Elsevier Science, 2003, ISBN 1-55860-792-7 .