Black & White TV.



Before explaining how multiplexing is used to produce colour TV it is useful to consider a basic outline of how television itself works. To provide a moving black and white (TV engineers tend to call this ‘monochrome’ TV) picture we have to be able to record, transmit and display information about a two-dimensional pattern of brightness. Nominally, we have to specify how the brightness of every point in a picture varies in time in parallel. In practice, the technique called raster scanning is used to convert a series of still pictures into a single serial data stream.

A light sensor swiftly scans a predefines path over the picture and reads out how the picture brightness varies along each line in turn. The arrangement of lines and the order/speed/direction in which they're scanned is called the raster pattern used by the TV system. This scanning process means we can get from the sensor a single time-varying pattern which tells us what we need to know to ‘reconstruct’ the picture using a TV receiver. This is normally done by scanning an electron beam across a screen which is covered with a phosphor (a class of chemicals which fluoresce when you illuminate it with electrons). We use the signal coming from the TV transmitter at every instant to control the intensity of the electron beam.



In order for the system to work correctly we have to ensure that the raster patterns at the transmitter (in the video camera) and the TV receiver are the same. We also have to ensure that the two raster scans are correctly synchronised. The choice of raster pattern, how the brightness variations are modulated onto the transmitter, etc, vary from country to country. Here I will describe the system used in the United Kingdom.

The UK system uses a raster of 625 horizontal lines. The whole picture is transmitted 25 times per second, but the lines are scanned in an interlaced pattern. First the scan works its way down the screen and transmits the 1st, 3rd, 5th, etc, lines. This takes 1/50th of a second. Then the scan goes back to the top of the picture and works its way through the 2nd, 4th, 6th, etc, lines. The odd lines form one half-frame, the even lines form a second half-frame. Taken together they provide all the picture pattern information of one complete frame of TV signal. Interlacing means that, although the whole picture is only refreshed once every 25th of a second each locality is visited every 50th of a second. This helps reduce visible flicker effects.

Given the presence of the scan lines/raster and the series of 25 distinct pictures every second it is perhaps surprising that the flickering of the apparently ‘moving’ picture isn't painfully obvious. The illusion of a steady moving picture appears for two reasons. Firstly, the human eye/brain has a property called persistence of vision. The eye takes a finite time to notice sudden changes in brightness. The brain has had millions of years of evolution to teach it that physical objects don't keep vanishing and reappearing 25 times a second. Secondly, the phosphors use in TV's keep fluorescing for a short time after they're hit with electrons. This means they tend to maintain a reasonable brightness over one or two frame periods.


Taken together, these effects help to make flicker go unnoticed. Despite this, 25 pictures per second isn't really good enough to give a completely flicker-free illusion. The use of half-frames improves things because it makes objects flicker at 50 Hz, not 25Hz. However, one of the few advantages of USA/Japanese TV pictures over UK ones that they use a 60Hz half-frame rate which, being quicker, makes flicker less noticeable. Anyone who has used high-end multisynch computer monitors may compare their computer display with a normal TV. The multisynchs often use frame-rates (not interlaced) greater than 75 Hz. If you get close to a TV picture and compare it with the multisynch you can often see that the TV picture is flickering although you don't normally notice from the other side of the room!



Content and pages maintained by: Jim Lesurf (jcgl@st-and.ac.uk)
using HTMLEdit3 on a StrongARM powered RISCOS machine.
University of St. Andrews, St Andrews, Fife KY16 9SS, Scotland.