How it works… The Problem of Genlock

This is the first of several articles to be printed in Fusion” about the engineering side of television broadcasting. Suggestions for future subjects will be welcomed. The word “genlock’ has been frequently used, perhaps far too often by people who do not really know what it means. Here the assistant head of engineering, basil bultitude, explains what it is, why it is used, and the advantages of other forms of generator locking.

To start with, the basic system of television transmission must be understood. A television picture is made up by one small spot of light travelling over the picture screen. It first starts at the top left-hand corner and rapidly travels across the screen in a series of straight lines. With each successive line the spot moves slightly down the screen, and eventually, when 405 lines have been completed, the whole screen has been covered and one picture is said to have been transmitted.

The time taken to transmit one picture is a twenty-fifth of a second. Because of flicker, inherent in this system, two half pictures each of 202½ lines, double spaced, are actually transmitted. Each half picture takes a fiftieth of a second to transmit and is termed ‘one frame’. The light spot varies in intensity related to the brightness of the original scene; although there is only one spot of light on the screen, it is moving so fast that persistence of vision of the eye enables us to ‘see’ a whole picture.

It is obviously necessary for the spots on all television sets to be in exactly the same area of picture at any instant. For this reason, at the end of each line and frame, synchronising pulses are transmitted. These synchronising pulses ensure that the spots in the receivers and in the cameras start a new line or frame at precisely the same time.

Because of interference from the electrical power mains which makes itself felt on the received picture, it is desirable that the frame speed be related to the mains speed (frequency), i.e. fifty per second.

To achieve this, the synchronising pulse generators (SPG for short) at the studio end are locked to the mains. One might think that if several SPGs were locked to the same mains supply, their outputs would be identical. Unfortunately this is not so.

The SPG mains locking system is not rigid but is in fact slightly springy. This means that two such generators locked to the same mains supply will be continuously varying slightly in speed, one with the other, depending upon the fluctuations of mains, and upon the ‘springiness’ of their lock.

Consequently, a picture generated by a camera fed from one SPG may be running at a slightly different speed from the picture of another camera fed from another SPG. This is insufficient to cause any noticeable mains interference, but it is sufficient to prevent superimposition or mixing of these two pictures.

This is, of course, a definite requirement for a television studio, and is achieved by driving all the cameras, etc, within one building from one SPG. At Wembley the SPG is situated in the central control room and at Television House in the master control room.

Sooner or later a situation must occur, when it would be desirable for a picture signal from Wembley to be superimposed upon a picture signal from Television House. The method of achieving this is ‘genlock’. In this system, the SPG at Television House is not locked to the mains but is in fact rigidly locked to the synchronising pulses from the Wembley generator.

Thus the degree of mains lock affecting the Wembley generator will also affect the Television House generator, and in consequence their synchronising pulses should be identical. The pictures from Wembley may then be superimposed on those at Television House. It is obvious that the Television House SPG may be locked to any incoming signal operating on British standards, but only to one at a time.

Thus, it would not be possible to mix pictures from Wembley, Television House and ITN together. If, when a Television House studio or telecine was on the air the SPG was then genlocked, a considerable disturbance would be seen on the pictures. In order to superimpose, one must genlock to a source before ‘taking it’, and because this cannot be done while Television House is on the air, a serious limitation is placed on its use. Attempts have been made to genlock more quickly by a system of automatic genlocking to reduce the picture disturbance. Nevertheless, a picture disturbance of up to four seconds may still take place. Automatic genlocking facilities are now available at Television House. The actual mixing or superimposition may only take place on the station whose SPG is genlocked.

It would be possible for Wembley to genlock its generator to ITN and Television House to genlock to Wembley, thus making a sort of ‘chain genlock’, then all three SPGS. would be running at exactly the same speed. This would mean extending lines from ITN to Wembley in order to carry the locking signal and would mean that ITN could only be mixed at Wembley.

This limitation occurs because of the different path lengths of the signals. The signals travel at roughly the same speed as light, i.e. 300,000,000 meters per second, and quite obviously a synchronising pulse that went from ITN to Wembley and then to Television House would arrive later than the picture signal that went from ITN direct to Television House. Within any television studio centre the path lengths of the various signals are carefully arranged to be the same, and this process is called ‘timing the station’.

From the above it will be seen that the genlocking system, automatic or otherwise, falls far short of the ideal of being able to mix anything anywhere. From our own company’s point of view, this imposes a very serious handicap. Our programmes may come from Wembley, Television House, outside broadcast vans and other programme contractors, all working, of course, on separate synchronising generators. The commercials always come from Television House, thus twice every fifteen minutes a different set of synchronising pulses are fed to the transmitter and of course to the home receivers. The receivers thus have to take up a new sweep speed suddenly, on the cut to and from the commercials and this nearly always results in a ‘frame roll’.

Towards the end of 1958, Associated-Rediffusion engineers attempted to rectify this situation by designing at new form of genlock called Slavelock. In this system the Wembley SPG was locked by a new method to the Television House SPG so that the two stations behaved, electrically, as one. This meant that the Television House SPG could then genlock to other sources, one at a time and the Wembley generator would automatically follow.

The necessary apparatus was built and in June of this year the two stations were locked for the first time by this method. Unfortunately the results were not perfect, the picture verticals from Wembley being slightly ragged. This was due to inaccurate timing. Work had to be stopped on this project because of other commitments, which is sad when we were so near our goal. However, perhaps in the future we may again start melting the solder on this apparatus.

The above is by no means the final answer, it only permitted sources to be mixed or superimposed at Television House and not, say, at Birmingham. Other broadcasting organisations throughout the world are working on the genlock problem. The BBC some months ago suggested that, if instead of locking to the mains, they locked their individual generators to a quartz crystal, they could simplify the problem.

Another idea of achieving a ‘synchronous network’ is that each incoming signal should go through a Standards Converter. A Standards Converter is basically a television camera looking at a television picture monitor. In this system the incoming picture is displayed on the monitor and the camera operated from the local SPG, thus the final picture from the camera is in lock with the rest of the station. This system has undoubtedly much to recommend it. The loss of picture quality in modern Standards Converters of good design is much less than is commonly supposed. However, it is certainly expensive; a Standards Converter costs approximately £20,000.

This is a complex story, and in order to try to explain it I have had to take some technical licence with the explanation, and for this I hope I may be forgiven by my engineering colleagues.