Analog Television - Digital Video and Audio Broadcasting Technology - page 19

• Interference

• Intermodulation

Linear distortion is caused by passive electronic components. The amp-
litude or group delay is no longer constant over a certain frequency range
which is 0 ... 5 MHz in the case of video. Parts of the relevant frequency
range are distorted to a greater or lesser extent, depending on the characte-
ristic of the transmission link involved. As a result, certain signal compo-
nents of the video signal are rounded. The worst effect is rounding of the
sync pulses which leads to synchronization problems in the TV receiver
such as, e.g. horizontal "pulling" or "rolling" of the picture from top to bot-
tom. These terms have been known since the early days of television.
Changing of heads from field to field produces similar effects at the top
edge of the picture with some older videorecorders, the picture is "pul-
ling".
These effects have become relatively rare thanks to modern receiver
technology and relatively good transmission techniques. In the active pic-
ture area, linear distortion manifests itself either as lack of definition, rin-
ging, optical distortion or displacement of the color picture with respect to
the luminance picture.
Nonlinear distortion can be grouped into

• Static nonlinearity

• Differential gain

• Differential phase

With non-linear distortion, neither the gray steps nor the color subcarrier
are reproduced correctly in amplitude and phase. Non-linear distortion is
caused by active components (transmitter tubes, transistors) in the trans-
mission link. However, they become only visible ultimately when many
processes are added together since the human eye is very tolerant in this
respect. Putting it another way: "Although this isn't the right gray step,
who is to know?". And in color television this effect is less prominent, in
any case, because of the way in which color is transmitted, particularly
with PAL.
One of the most visible effects is the influence of noise-like disturban-
ces. These are simply produced by superimposition of the ever-present
gaussian noise, the level of which is only a question of its separation from
the useful signal level. I.e., if the signal level is too low, noise becomes vi-
sible. The level of thermal noise can be determined in a simple way via the
Boltzmann constant, the bandwidth of the useful channel and the normal