Television, the Early Years

This article is based on a series that appeared in Electronics Australia in July and August 1973.

At the age of 14, inspired by this series of articles I attempted to build a replica mechanical television based on the early scanning disk principle, assisted by my school friend George Mathews.

Unfortunately we could never get our contraption to work properly, but it "looked the part" and the local Sunday Times newspaper wrote an enthusiastic article about our attempts. We enjoyed celebratory status at City Beach Senior High School for a short while and displayed the equipment on parents night.

This page has been compiled by Peter May using text from the abovementioned series and other historical references.

Historically, one of the most interesting aspects is the period over which television experiments were conducted. We are inclined to think of TV as a purely modern concept, yet ideas and experiments go back for over 100 years.

As early as 1839, Becqueral had observed that some materials undergo chemical changes when exposed to light and also could be made to produce weak electric currents. From this discovery came the first crude photo-electric cells.

In 1847 Bakewell devised a system for transmitting still pictures over telegraph wires. This incorporated the vital principle of scanning the subject in a series of lines. This was not television as we know it, but rather the related technology of facsimile. Nevertheless, it was a start.

In Paris by an Italian, Giovanni Caselli, to provide 1863 a public facsimile service was installed between Marseilles and a public service similar to that provided for written messages by the conventional telegraph system.

Considering the crude equipment with which the designer had to work, the results which have been preserved are of quite remarkable quality, and a tribute to his imagination and resourcefulness.

But it was one thing to transmit a single picture, taking as much time as was necessary, and quite another to transmit a moving image, essentially in step with the original. In 1873 came another discovery of importance; the light sensitive properties of selenium. Selenium had been isolated in 1817 by the Swedish chemist, Baron Jons Jacob Berzelius, but its light sensitive properties were not realised at the time.

In 1873 a telegraph operator, Joseph May, was stationed at the cable station at Valentia, a small island off the coast of Ireland. Rods of selenium were then used as resistors, and May noticed that they changed their value under the influence of strong sunlight.

This discovery led to the commercial production of photocells having greater sensitivity than anything previously available. This, in turn, led to a number of proposals as to how a true television system might be constructed.

One of the earliest suggestions appears to have been made around 1875 by an American, G. Carey. He proposed a large screen, or mosaic, of photocells onto which would be projected, by means of a lens, the scene to be transmitted. From each cell a wire would run to a mosaic screen at the receiving end, made from an equal number of lamps. Thus each lamp at the receiver would be controlled by its companion cell at the transmitter, and light brightly, dimly or not at all, according to the amount of light falling on the cell.

Naturally, the smaller the cells and lamps and the larger their number, the better the image quality. At the same time, as the number of picture elements was increased, so would the number of wires connecting the two stations. Even a very elementary system, consisting of, say, elements, would call for 900 30 rows of 30 elements each, or a total of 900 circuits between stations; a prohibitive number, and for a very crude result.

Nevertheless, others toyed with the same idea and around 1877 a Frenchman Senlecq, is supposed to have made a working model.

Impractical though the concept was, it did envisage the vital requirement of breaking the image into elemental areas, each to be transmitted and displayed as a separate entity.

The flaw in the scheme was the proposal to transmit all the information simultaneously, over separate circuits. Most subsequent ideas envisaged the scanning principle transmitting the information from each elemental area sequentially, and with sufficient rapidity to allow persistence of vision to ignore the sequential action. The main difference between the various schemes was the manner of scanning.

Incidentally, it is interesting to note that the mosaic idea was tried many years later by both Baird (England) and Jenkins (USA) but with the essential difference that the elemental areas were scanned sequentially using a mechanical commutator.

In 1884 Nipkow, of Germany, suggested a television system using a scanning disc; a device which came to be known as the Nipkow disc. There were many variations on this idea but, basically; the scheme was very simple. A metal disc was punched with a series of holes in a spiral pattern. Commencing near the outer edge, the spiral made a single turn towards the centre, each hole being closer to the centre by approximately its own diameter.

As envisaged by Nipkow, the scene to be transmitted would be projected onto the disc by means of a lens. Behind the disc would be a selenium cell. As the disc rotated each portion of the scene would be scanned in a series of parallel (though slightly curved) lines. The output of the cell would thus vary according to the light intensity of each part of the scene.

At the receiving end the current from the cell would be used to control a polarising prism, interposed between a light source and an identical disc. The prism would modulate the light and, assuming the receiving disc was running synchronously and in phase with the transmitting disc, an image of the scene would be observed by looking through the holes in the disc at the modulated light.

Unfortunately Nipkow was unable to make it work. It is doubtful whether enough light would have reached the cell to produce any worth-while output but, in any case, its response would have been far too slow. At the receiving end it is doubtful whether the prism arrangement - suggested originally by Faraday - could have been made to work, and certainly not with the minute signals likely to be generated.

In 1884 Hertz, and later Hallwachs, Established the basic principle of the modern photocell; the emission of electrons from certain metallic surfaces under the influence of light. Although it appears that some cells were produced not long after this discovery, and exhibited an excellent response time, their sensitivity was far too low to be of any practical use. Remember, there were no valves or other suitable amplifying devices available at this time.

And so television, even in a crude form, remained a tantalising dream; tantalising because scientists felt that they had the method, but not the means. The idea was far ahead of the technology.

But in spite of the seemingly insurmountable problems there was no shortage of new ideas. Various workers continued to suggest alternative schemes, most of them aimed at improved scanning systems. Some were even beginning to appreciate the limitation of the cumbersome mechanical systems.

One of these was a Russian professor, Boris Rosing. In 1907 he proposed, and apparently built a television system. Whether the system worked or, if it did, how well, is not clear. What is more important, however, is that Rosing suggested a cathode ray tube as display device at the receiver; the first such suggestion as far as is known. (The cathode ray tube was developed in 1895 by Sir J. J. Thomson, for completely different applications.)

The transmitter system was also novel in that, while still mechanical, it employed an ingenious system involving two mirror drums. A basically similar system was to be used in practical experiments by several workers, particularly Baird, many years later.

Sweep signals for the cathode ray tube were generated by coils mounted adjacent to the drums and permanent magnets mounted on the drums. These were conveyed to the receiver over separate lines.

As with Nipkow's system, Rosing's design envisaged scanning the image by dissecting lens. While the concept was advanced, it seems doubtful whether he was able to achieve any success, considering the limited photocell devices which were available.

But Rosing's name and work are important for another reason. Many years later, in the USA, a scientist named Zworykin developed a camera tube the iconoscope. Zworykin was a Russian who migrated to the USA and had worked in Rosing's laboratory. It is claimed that it was here that he conceived the idea of his camera tube, though it was many years before it became a practical reality.

Also in 1907, and before Rosing's patent was published, an English scientist, A. A. Campbell-Swinton, suggested a system of "Distant Electric Vision" using the cathode ray tube. In 1908 he wrote a short article for the magazine "Nature", describing the system. This aroused a good deal of interest and he was urged to present his ideas in greater detail. This he did in the same publication in 1911.

Considering when it was written, Campbell-Swinton's is a truly remarkable paper. It proposed cathode ray tubes, at both receiver and transmitter. With regard to the latter, it describes a variation on the tube which makes it remarkably like the present day camera tubes.

The paper also revealed a keen appreciation of the problems which had been encountered so far. and the manner in which these would be overcome by the proposed scheme. It discusses the storage effect, which is a vital part of present day camera tubes, and the provision of suitable persistence in the receiving tube screen.

Many years later -- 1938 -- both J. L. Baird the English pioneer, and J. D. McGee of the famous Marconi-EMI team, paid tribute to Campbell-Swinton's camera tube concept. Baird said "This transmitting device is doubly interesting, as a modification of it is used in present day practice."

McGee, speaking of the Emitron, which he helped develop, said, "This transmitting tube was developed in the EMI laboratories directly from Campbell-Swinton's original suggestion."

But was the Emitron the only descendent of Cambell-Swinton's suggestion? His contemporary, Rosing, would almost certainly have learned of the suggestion.

And Zworykin worked in Rosing's laboratory. Did Campbell-Swinton's idea also find practical form in Zworykin's iconoscope, many years later?

Campbell-Swinton himself tended to modestly dismiss his ideas as "only an effort of my imagination, a suggestion of the direction in which experiment might possibly secure what is wanted."

Once again the idea was ahead of the technology. Among the vital links still missing was a suitable amplifying device, the valve being several years away.

Thus, over a period of some thirty years, suggestions had ranged all the way from obviously impractical mosaic systems through a variety of mechanical scanning systems, to a completely electronic system.

Rather strangely, history was to largely repeat itself over the next thirty years. Granted, workers moved from mere suggestions to practical experiments, but they were to go through the full range of mechanical scanning ideas again, before finally settling for the all electronic system.

Following Cambell-Swinton's suggestion, the subject stagnated for several years, due largely to the intervention of World War I and the settling down period which followed it.

In 1922 the subject was taken up again by man whose name has become synonymous with the development of television, at least in England and, to some extent, throughout the world: John Logie Baird.

There has been much controversy over Baird's role in the development of television. His opponents have always been quick to point out that not one of his technical developments found its way into the system we know today, and this is completely true.

But if Baird did nothing else, he stirred public interest to the point where the subject could no longer be ignored. There were plenty who either did not believe that the technology was possible, or that, even if it was, that it was desirable. But controversy only aroused public interest all the more so that, eventually, some kind of a development was almost inevitable.

Later, controversy was to rage again on a different plane; the mechanical system versus the all-electronic ones; the high broadcast band versus VHF. In fact, they were really all variations on the same theme, but different people saw them in different ways.

From his early childhood Baird had shown a scientific bent and, while still at school, had experimented with selenium cells in an effort to produce talking pictures and television. While these efforts were doomed to failure, they indicated a rare depth of thought in one so young.

It was in 1922, at the age of 34, that Baird approached the subject of television in earnest. In spite of suggestions of earlier workers such as Campbell-Swinton, he elected to investigate the mechanical systems which had been suggested to date. Among other things, these had the advantage of simplicity and low cost.

And cost was important. Baird had little capital, no equipment, and only an attic in which to work. To tell the full story of his struggle would take a whole chapter, and chronological account of his achievements.

It should be recorded, however that Baird worked against almost overwhelming odds; poor health, lack of capital, active opposition in some quarters, apathy in others. At one stage he was literally starving in his attic workshop; denying himself proper nourishment in order to carry on his work. Between 1922 and early 1924 Baird worked long hours, using crude, improvised equipment in an effort to achieve any result at all. And it wasn't until early 1924 that he transmitted over a distance of a few metres, the first crude image. It was simply the outline or shadow of a Maltese cross but, crude though it was, it was a big step forward.

It is only fair to record that others were working along the same lines at this time, particularly Jenkins of the USA. Apparently he achieved a similar order of performance at about the same time. Other workers were experimenting in Austria and France.

From the transmission of mere shadows Baird progressed to images of objects by reflected light. But for some reason which is not recorded the system was still not able to reproduce a range of tones, only blacks and whites.

It took almost another two years (October 1925) for Baird to achieve what he regarded as "true television" the transmission of an image - the head of a ventriloquist's doll -- with a useful range of light and shade to give modelling to the features.

Following this Baird gave many demonstrations to scientific bodies, the press, and the general public. He also conducted a number of experiments and demonstrations involving variations on the original theme.

Initially, at least, Baird appears to have followed Nipkow's and Rosing's lead in that he used his scanning system to dissect the image before it reached the photocell. While he had a tremendous advantage over his predecessors, in that he had much improved photocells, and valves to amplify their output, the overall optical inefficiency of this arrangement must have been immediately apparent.

The main problem was the amount of light needed on the subject. The brilliance, and the heat which accompanied it, was more than most subjects could stand for more than a few minutes. Somewhere along the line of investigations - though it is not clear exactly when, or by whom - an alternative idea was tried.

This was to use the disc to control the light source before it reached the subject. Thus the subject, normally in near darkness, was scanned by a spot of light which swept across it in a series of parallel lines. The photocells - and there could be as many as needed now - viewed the scene as a whole, but responded at any instant to light being reflected from that particular part of the scene actually being lit by the spot.

Ultimately large banks of cells were used, supported so as to view the scene from different angles. In this regard, it is interesting to note that there was a complete transposition of roles. The light beam was really playing the role of the camera, since banks of photocells, on the other hand, played the role of lamps.

This effect was quite genuine, and the banks of cells were positioned exactly if they were floodlights. If part of a scene was too dark, an extra bank of cells would be directed to it in order to brighten the reproduced image.

Obviously, the system had weaknesses. A major one was that subjects could only be televised indoors, where the ambient light could be controlled. Nevertheless, it formed the basis of most experiments over the next few years.

In 1926 Baird obtained permission to make an experimental transmission over a BBC transmitter, with the strict proviso by the BBC that the experiment was to remain a secret. In December 1926 he demonstrated television using invisible (infrared) light to illuminate the subject.

In early 1927, American experimenters succeeded in transmitting television images over 320 kilometres of telephone line.

A few months later Baird demonstrated the transmission of images between London and Glasgow -- a distance of some 700km.

In the same year he operated his own transmitter on about l500kHz and transmitted television images to a monitoring point about 20 kilometres away.

Later Baird established a short-wave transmitter in Surrey, the main object being to being to bridge the Atlantic. This was achieved on February 9, 1928. The signals were originated in his London studio, sent by landline to Surrey, radiated on about 6.7MHz, and received by an amateur a few miles from New York.

In early 1928 he transmitted images to the Cunard ship "Berengaria" in mid-Atlantic. In June 1928, he demonstrated the television of objects in daylight, but there is no record of how he achieved this, or why it was not more universally adopted.

In July. he demonstrated colour television, and in August, stereoscopic television. Impressive though this record is, it must realised that all these experiments employed very low definition standards.

The picture was made up of 30 lines, with a repetition rate of 12.5 pictures per second. Mechanical scanning was used at both the transmitter and receiver.

Looking back, it is difficult to get a fair impression of the quality of these low definition reports of demonstrations by Baird and others, but it appears that the writers were often carried away by the fact that a moving picture of any kind was transmitted. They often used such exaggerated phrases as "perfect definition" and "leaves nothing to be desired" which makes the rest of their descriptions highly suspect.

Such photographs as remain are usually in the form of half tone reproductions, and have therefore suffered one further order of degradation. In addition, a moving image tends to gain slightly in definition and is also subjectively more acceptable.

The most accurate description would seem to be that, assuming a head and shoulders picture, the person's features would be recognisable by those who knew them -- but only just.

By 1928, in spite of the limitations of the low definition system, Baird and his supporters felt that it should be given a trial as a public service, via the BBC. Initially, the BBC was less than enthusiastic feeling that the quality was not good enough, particularly as such transmissions would have to be limited to the 9000 Hz modulating frequency permitted for British broadcasting stations.

But the Baird camp persisted and, in March 1929, the BBC consented to an official test. The result was acceptance, in principal, of Bairds suggestion. After somewhat protracted negotiation, an agreement was reached whereby Baird was given five half-hour sessions a week for experimental transmissions.

These transmissions commenced on September 30, 1929. Initially, only one transmitter was available, so there was no accompanying sound. In March 1930 a second channel was provided, allowing simultaneous transmission of vision and sound. Vision was transmitted on 870 kHz and sound on 1330 kHz.

In 1932 Baird's apparatus and engineers were taken over by the BBC and his studios transferred to Broadcasting House. These transmissions continued until April 1934, when they were curtailed to two half-hour sessions a week.

The advent of these transmissions created a great deal of interest, particularly among experimenters. They found that, for the expenditure of a few pounds, they could convert the BBC signals into moving pictures.

Results could be obtained with equipment which was almost unbelievably simple. A neon tube, a disc (which could be home-made), and a small electric motor were the basic materials. The neon tube was connected in place of a conventional radio receiver's loudspeaker, the disc mounted on the motor shaft, the neon tube mounted behind the disc, and the experimenter was in business. A typical advertisement of the day offered a complete kit for £5-9-6 which could be assembled in 45 minutes!

The neon lamp was a special design using a relatively large rectangular plate as the main electrode. When viewed through the holes of the disc, running at the correct speed (750 rpm) and in proper phase, it created the image as its light output fluctuated in sympathy with the incoming signal.

Results from such simple equipment were strictly limited. In addition to the fundamental limitation of definition, the picture was small, about 45 mm x 85 mm, the light output quite low, and the light neon pink rather than white.

There was also the problem of synchronisation. Initially this was left entirely to the viewer, who had to juggle the speed of his motor, using variable resistors and/or friction brakes, to bring his disc into step, in both speed and phase, with the scanning cycle from the transmitter.

Later, a serious attempt was made to transmit synchronising information along with the picture. This was done by simply masking the top and bottom of the picture slightly at the transmitting end. Vertical scanning was being used, so this produced a brief period of no signal between each line.

With 30 lines at 12.5 picture per second, there were 375 lines transmitted every second with the same number of breaks between them. Thus, regardless of the picture content, and the other frequencies which it produced, because of the "black" pulses, the television signal always contained a prominent 375 Hz signal. This became the synchronising signal.

At the receiving end this signal was extracted and used to drive a small synchronous motor. This unit was additional to the main driving motor which was typically a series wound, brush type, "universal" motor, of about 1/30 horse power. The synchronous motor was mounted on the same shaft and was powerful enough to slow down or speed up the main motor once the latter was running at approximately the correct speed.

As well as holding the receiving disc in synchronous speed, the synchronous motor could be used to adjust the phase, or "frame" the picture. This was done by rotating the outer case of the motor, carrying the coils, thus causing the rotor to follow it.

The synchronous motor was connected in series with the neon lamp directly in the plate circuit of the output valve. The motor windings were bypassed with a capacitor to provide a path for the higher video frequencies.

Although the scheme worked reasonably well, it was not without its problems. Since there was nothing to differentiate between the black of the picture border, and a black which could occur within the picture, the system sometimes became confused. Large black areas at the top of, and running out of the picture were particularly prone to cause of loss of sync.

Many schemes were devised to produce a bigger and brighter picture. The basic approach was to use a relatively high powered lamp (100W projector type) in place of the neon lamp. Since filament lamps cannot be modulated at the frequencies involved, a variety of light valves was devised.

One of the most popular used a pair of Nicol prisms and a Kerr cell. One prism polarised the light beam which would then pass through the second prism only if its polarisation was undisturbed. The Kerr cell uses an electrostatic field to rotate the polarisation thereby effectively modulating it.

Such a cell was commonly used in con junction with a mirror drum scanner. The mirror drum system, though more complex, was more compact. A typical drum would measure about 150 mm in diameter and 40 mm wide. Around the outside of the drum was fastened a total of 30 mirrors, one for each line. The modulated light beam from the light valve was focussed onto these mirrors.

The drum was mounted on its vertical axis so that, as it rotated, the beam was swept vertically from the top to the bottom of the screen. To provide the vertical scan, each mirror was set at a slightly different alongside the preceding one.

The beam from the drum was normally directed onto a ground glass screen, which was viewed from the opposite side. The result was a much larger, brighter picture composed of white light. Kits for this type of receiver were also marketed, particularly by Baird (Baird Television Ltd) and sold for about £25.

Transmitting equipment was somewhat similar. For studio work a mirror drum scanner converted the light from an arc lamp into a series of lines which scanned the collected by banks of photocells. For films a simple disc scanner was used.

The most difficult image to handle was the outdoor scene. Such a scene could not be scanned with a beam of light, and attempts to direct the image onto photocells after dissection had not been very successful, due to the enormous light loss involved.

This led to the intermediate film system. The scene was photographed with an ordinary movie camera, typically mounted on top of a van for sporting events. A light tight tube ran from the camera into the van which was, in effect, a film processing plant.

There was no takeup spool on the camera, the film running straight from the gate into the tube and thence to the processing tanks. High speed developer and fixer were used and the processed film, still wet, was taken straight to the scanning gate, which was under water. The film was then run into a tank of water until such time as it could be dried and spooled.

Typical delay time between camera gate and scanning gate was 30 seconds, although the German Fernseh A.G company managed to reduce this to 10 seconds.

A variation on this idea, which was tried with limited success, was to pass the film through a stripping bath after scanning, remove the emulsion, then pass it through an emulsion coating bath, dry it, and pass it back to the camera. Thus only a relatively small loop of film was required.

Regardless of the particular system, the sound was recorded on the film at the same time, to delay it by an equal amount. It was generally claimed at the time that the small delay was of little consequence if an image of an important event could be transmitted.

A similar system was used to present large screen television pictures in a theatre. The television image was recorded on film, the film processed at high speed, and passed straight into a conventional projector.

There were many minor variations of these basic television systems suggested and tried between 1929 and 1934 and it is impossible to detail them all. What we have described should give a fair idea of the state of the art at that time.

In fact, time was running out for all such low definition systems. The truth was that, in terms of television research, they were only the tip of an iceberg. Because they had the advantage of technical simplicity and low cost, and had arrived, the low definition systems had generated an enthusiastic following and created plenty of publicity.

But results were too poor to be taken seriously. Only the most ardent enthusiast could derive any real entertainment from such a system, once the novelty had worn off.

Several groups of workers, in both Britain and the USA, had long appreciated this, and had set their sights on a much higher standard; the so called "high definition" system. So, while the low definition systems temporarily captured the public imagination, the high definition enthusiasts were working hard, with little publicity, toward a goal which they believed would be really worthwhile, even if it took longer to develop and cost more to implement.

The high definition system was an ambitious concept, with many problems to be overcome. Probably the most basic was to find a substitute for mechanical scanning, since the size and complexity of the latter grew enormously as the number of lines was increased. This, combined with their unsuitability for outside broadcasts, ruled them out. Campbell-Swinton's original suggestion based on the cathode ray tube was the logical alternative, but converting this general concept into practical hardware was a formidable undertaking, even allowing for the enormous technological advances which had occurred in the intervening years.

Then there was the equally basic problem of the bandwidth required. The low definition experiments had used the broadcast or short wave bands and had not generated modulating frequencies much above 10 kHz; 15 kHz at the most in one or two special cases.

Because an increase in the number of lines requires that the definition within each line be upgraded proportionately , bandwidth increases roughly as the square of the increase in line numbers.

Thus, while the highest possible number of lines was desirable, this requirement had to be balanced against the bandwidth problem. To make matters worse, no one was quite sure just how many lines would be needed for an acceptable picture. Suggestions ranged from 60 lines to 120, 180 and 240.

There was also the problem of handling the vision (video) frequencies before modulation and after detection. Even l0 kHz for a low definition image presented problems. When someone suggested a 100 line system at 25 pictures a second it was calculated that the video frequencies would be in excess of 300 kHz; a formidable figure at that time.

The suggested solution to the spectrum space problem was to use the ultra shortwave (VHF) bands; There was plenty of room there, the main problem being that they were not well understood at that time. Even while advocating their use, engineers realised that a lot would have to be learnt before it was practical. Much the same applied to the video frequency problem, though they were more confident of coping with this.

Nevertheless, a lot of real progress had been made, particularly in the USA by Zworykin of RCA. The general TV scene in the USA had run roughly parallel to that in Britain, with Jenkins, and others experimenting with mechanical systems. But as early as 1923 - only one year later Baird commenced his experiments - Zworykin had begun work on an all electronic system. By 1929 he was able to stage a demonstration which proved that such a system was possible, even though it was still very much in the laboratory stage.

In Britain the high definition champions were Electrical Musical Industries LTD (EMI) and the Marconi Company. In the 1920s the Marconi Company had carried out extensive work on the development of a high speed facsimile system. In so doing they had encountered, and solved, many of the problems inherent in the transmission of high definition television signals.

EMI had also been conducting television experiments. Like everyone else, they faced the vexed question of "how many lines?" and, in 1931, developed an experimental 120 line film scanner. To test it, and the 120 line concept, under working conditions, they needed a VHF transmitter. They ordered it from the Marconi Company.

Thus began a close liaison between these two companies; one which was to have a marked effect on subsequent events.

Zworykin's work, particularly that associated with an electronic camera tube, and his demonstration in 1929, had aroused a good deal of interest in the EMI team. They even considered buying the knowledge which RCA had already acquired. As it turned out, the price was too high. and they decided to "go it alone".

In the team which shouldered this task were a number of engineers worthy of special mention: J.D McGee, A.D Blumlein, C. O. Browne, N. E. Davis, E Green. Blumlein is generally regarded as one of the most brilliant engineers of his time. Tragically, Blumlein and Browne, another brilliant worker, were killed in a plane crash while engaged in advanced radar research during the war.

The team's major effort, involving mainly McGee and Blumlein, was undoubtedly the production of a camera tube, similar to that being developed by Zworykin, which they called the Emitron. The patent for this was issued in 1932. (Zworykin published a description of his tube, which he called the Iconoscope, in 1933.)

While McGee appears to have been largely responsible for the Emitron itself, Blumlein concentrated on devising circuits which. would compensate for the many deficiencies of this type of tube (which it shared with the Iconoscope). Later, Blumhein was also responsible for developing the television waveform on which the complete EMI television system was based. A patent for a much improved version of the Emitron was issued to McGee and Blumlein in August 1934.

Other members of the team concentrated on the standards to be adopted, and the hardware to achieve it all. The number of lines was a particularly contentious issue. In finally opting for 405, the team seemed to be flying in the face of commonsense. There were many prophets of doom who maintained that the bandwidth required could never be handled; that 180 or, at most 240 lines was not only adequate, but all that could be achieved in practice.

In spite of the sceptics, by 1934 the team felt that they had a viable system; a system which they could offer to the authorities with the assurance that it would achieve lasting public acceptance. As it turned out, the year 1934 was a significant one in television history; one in which many far reaching decisions were made.

Early in 1934, Baird demonstrated a 180 line system on VHF. The vision signals were transmitted on 50 MHz and the sound on 48 MHz. Mechanical scanning was used at the transmitter but the picture was displayed on a 12 in diameter cathode ray tube.

Around the time Baird was demonstrating his 180 line system, EMI and the Marconi Company were making the logical move in view of their common interest in television, and the broad background of experience which, jointly, they possessed. They joined force to form the Marconi-EMI Television Co Ltd.

About March 1934 the BBC decided to reduce the 30 line transmission to two half hour sessions a week.

In May 1934 the British House of Commons appointed a Television Committee, "To consider the development of television, and advise on ... the relative merits of the several systems ..." Blumlein, among others, gave evidence before the committee to the effect that Marconi-EMI could provide a 405 line all electronic system. Baird Television PTY LTD also gave evidence, proposing a 240 line system.

In January 1935 the report was presented to the Postmaster General. Among other things it recommended a high definition system of "not less than 240 lines with a minimum picture frequency of 25 per second."

In August 1935 the BBC announced acceptance of the recommendation. Both systems were to be given a trial, operating on alternate weeks. Baird was to use 240 lines, 25 pictures per second; non interlaced, with mechanical scanning. Marconi-EMI was to use 405 lines, 25 pictures per second interlaced, and all electronic scanning.

Each system was to have its own studio equipment and vision transmitter. A third transmitter, common to both, was provided for the sound. Vision was to be radiated on 45 MHz and sound on 41.5 Mhz.

As a result of this announcement the pressure was really on for both organisations. Baird, still backing mechanical scanning, had to produce a 240 line system capable of handling large studio scenes, film scanning, and outdoor subjects. The fact that he did produce such a system is a tribute to his ability and determination.

EMI faced a similar situation. In early 1935 they had only a few prototype units in existence, from which they had to produce a complete working system ready to go to air. The fact that they did this in 18 months is a tribute to not only their technical ability, but to their unbounded enthusiasm and willingness to work as a team.

The BBC was also under pressure. Alexandra Palace had been chosen as the site for both studios and transmitters mainly because it was already 100 metres above sea level, and height was important for good VHF coverage. A major job was to strengthen one of the building's corner towers and then to erect a steel tower, another 100 metre high, above it.

Internally the building had to be converted into studios, dressing rooms, administrative offices, transmitter rooms etc. Then the two companies had to supply, install and test their equipment.

By late 1936 both systems were installed and working and both went on the air for experimental periods. The system was formally opened on 2 November 1936. There were in fact two ceremonies, one transmitted on each system. After that the two systems operated week about for the trial period.

(The receivers of the day had to be fitted with a time base changeover switch in order to accommodate the two different line frequencies; 6000 Hz for the Baird system, 10,125 Hz for the EMI system.)

Both systems provided similar facilities, though by quite different methods. Both could televise live studio production, sound films, and outdoor scenes. In regard to the latter, both suffered a common limitation; the distance between the camera and the transmitter.

Initially, the maximum length of cable between the camera and the transmitter had to be limited to about 330 metres. This meant that outdoor scenes were limited to those that could be staged in the grounds of Alexandra Palace. These were devoted largely to demonstration of various sporting techniques, golf, tennis, riding etc, by well known experts in these fields.

Apart from this problem, which was soon overcome, the Marconi-EMI system faced no serious limitations. The Emitron camera was just as much at home out of doors as it was in a moderately well lit studio.

Two film scanners were provided so that, as in normal theatre practice, a continuous program could be presented from standard reels of film. The scanners were standard intermittent film projectors, slightly modified, and arranged to run synchronously with the waveform generators. They simply projected the film image onto the mosaic of the Emitron.

The Baird system was not nearly so versatile and, while they could perform all these functions, many of them had to be done the hard way. For small scenes, such as three-quarter length presentations, or interviews involving two people, the "flying spot" system was used.

This was similar to that used for the for the old 30 line system, but upgraded to 240 lines. The light source was a powerful arc lamp, and scanning was by means of a disc punched with 240 holes in a spiral pattern, running at 1500 rpm in a vacuum.

For larger scenes, it was difficult to obtain good light coverage using this method, and it was necessary to employ the intermediate film system, which has already been described. This system was also used for outdoor scenes.

For film scanning, involving either rapidly processed intermediate films or standard commercial films a disc scanner was used with a novel modification. Rather than a spiral, the holes were arranged in a circular pattern, providing horizontal scanning only. The vertical scanning wall provided by the film, which was moved through the gate continuously, rather that intermittently as for normal projection. This system of continuous motion is still used today, in conjunction with the cathode ray type flying spot scanners.

By February 1937 the BBC had made it's decision; it favoured the Marconi-EMI system. On 5 February 1937 the Baird system was dropped and the world's first regular public TV service settled down to the more or less routine job of supplying programs and learning what could be done.

Even at this early stage the system had so proved itself that Blumlein, representing Marconi-EMI, went to the USA to negotiate a free exchange of "know how" with RCA. By this time RCA were only too happy to agree.

The next big step war to shake oft the shackles imposed by the limited length of cable between camera and transmitter. In this case there was a particular incentive; the coronation of King George VI was set down for May 12, a mere three months away. Everyone was anxious that this should be the first outside broadcast.

To do this it was necessary to develop a complete mobile TV station; cameras, control room, VHF transmitter, power plant etc, to relay the signals to the Alexandra Palace transmitter.

The system was developed and built by Marconi-EMI and delivered to the BBC just in time for the coronation. It was built into three large vans. One van carried the camera control equipment capable of handling three Emitron cameras, plus the sound circuit, control, and microphones.

The second van carried the modulation amplifiers, a 1 kW VHF radio transmitter (85 MHz), and a demountable directional aerial. The third van carried generating plant for where AC mains were not readily accessible.

As well as providing a radio link, the system could work into a cable link where one was available. A suitable cable had been provided between Alexandra Palace and Broadcasting House, and to a number of points in the West End of London.

Although the equipment was delivered to the BBC only days before the coronation, with little time for testing, they decided to attempt televising the coronation procession. In spite of these hazards, and poor light caused by overcast weather, the experiment was a complete success. As a precaution both the cable link end the radio link were used, one to serve as a backup if the other failed.

From this point on development was more of an evolutionary nature, but much progress was made before the outbreak of World War II in 1939 closed the system for the duration.