From Baird to MPEG
1 Jul 2001 0 comments. tbs.pm/1968
Martin Fenton presents a history of video
Video tape recording as we know it today dates back to 1956, when both the BBC and Ampex Corporation launched their own magnetic recorders.
In order to appreciate this achievement fully, however, we must go back to a time when Britain owned over half of the world, and Queen Victoria was the British monarch.
Valdemar Poulsen, today acknowledged as the father of magnetic recording, demonstrated his magnetic wire recorder in 1898, only a few years after the debut of the moving picture, but the results were very poor, and the machine failed to be a success. The next notable magnetic recorder to be marketed is the Blattnerphone, which appeared in 1929.
It recorded on a steel band of quarter-inch width, running at a high speed. It is recognised as the first medium to be able to record for long durations – up to half an hour – but its main disadvantage was that it was quite an enormous piece of machinery, and the steel tapes were impossible to edit.
Nevertheless, the BBC bought and installed Blattnerphone equipment in its studios, merely as a means of time-shifting programmes or items without recourse to using discs.
Magnetic tape was first shown in 1927 as iron oxide deposits on a paper base.
The German AEG company demonstrated their Magnetophon tape recorder in 1935, to be used with a plastic-based tape that BASF had developed the previous year.
Tape recorders actually went on sale in Germany and the USA in 1937 – surprising considering that, toward the end of the Second World War there were reports that the Allied Intelligence Services were puzzled by the ability of German Radio to broadcast concerts of classical music late at night.
They were certainly not being broadcast from disc and it was hard to believe that it would be live at that time of night. As the Allies moved into Germany at the end of the war, a number of tape recorders and supplies of tape were discovered at the radio stations. When taken back to America for analysis, it was found that these machines were far superior to the pre-war instruments.
Even before the BBC started officially broadcasting television in August 1936, John Logie Baird and his engineers were experimenting with recording television signals. (For more on this topic, see Pictures of the Dawn.)
To quote Don McLean, the engineer responsible for recovering the recorded signals nearly seventy years later, “recording the television signal in the 1920s was much simpler than you might at first think. With only 30 lines per picture (television frame), the highest frequency present was low enough to be audible. The video signal could therefore be recorded as an audio signal onto disc.”
And so it was that Baird, using then-available home gramophone cutting equipment, recorded television signals onto 10″ shellac discs as early as 1928. What he couldn’t do was play the recordings back. It took until 1994, and digital techniques, to make that possible.
It is a commonly held misconception that television pictures exist from 1936. This is not true: the pictures normally shown from 1936 were staged for cameras for the BBC Film Unit, for a series of “demfilms” shown during the daytime on BBC Television, and were not recorded during the original broadcasts of these pieces. The earliest surviving piece of original BBC television dates from 1938, is silent, and was discovered in 1998 by Andrew Emmerson.
Initially, television was either broadcast live, or from film via Cintel “flying-spot” film scanners. Film recording was a much longer, drawn-out process than live broadcasting, particularly as the film had to be processed.
Baird’s original Intermediate Film Technique involved pointing specially adapted film cameras at the subject, running the film through the camera and then through various developing chemicals held in a box attached to the camera, broadcasting the “live” picture through a film scanner to the gallery some 60 seconds later, while the film was still wet. It would be vision-mixed from there as normal. The BBC quite rightly abandoned the Baird system in favour of the electronic Marconi-EMI system in February 1937.
In 1946, the standard for telerecording was set: a specially adapted lens was placed before a cathode ray tube, and the picture was optically recorded, with the synchronisation set through the original studio sync generator.
The sound was added via a direct line from the sound desk, rather than from the monitor, which produced whistle. Nearly all black-and-white television recordings seen today are telerecordings, because of two factors: firstly, for a long time, videotape was more expensive than film (assuming that each reel was to be used only once.)
Secondly, it was believed for a long time that videotape had no archival stability, and that for a recording to be preserved, it should exist on film. While some tape types have proved to last longer than others, many of the Quadruplex videotapes that have survived from the early 1960s look as good today as they ever did.
By 1957, with tape recorders having taken a strong hold of the sound broadcasting world, the BBC attempted to manufacture a similar machine that would electronically record television signals.
The result was VERA – the Visual Electronic Recording Apparatus. It was introduced to great fanfare by the BBC’s “Panorama”, but quietly disappeared shortly after. VERA was withdrawn because of the invention and success of the 2″ Quadruplex video tape recorder by the Ampex Corporation in America.
The performance of the Quadruplex (so called because it had four heads mounted 90 degrees apart on a head wheel) outshone VERA considerably.
This was due to the fact that the heads rotated, whereas VERA’s heads were fixed; where the BBC’s machine had to run at considerable speed to achieve the necessary tape-to-head speed to record and reproduce the high frequencies that make up a television picture, the Ampex deck could run at a steady 15 inches per second (the same speed as professional audio tape recorders of the time), with the head drum revolving at 250 revolutions per second (PAL).
Each of the four heads produced half of each field, and thus a quarter of each television picture. While VERA’s results weren’t truly awful, they tended to appear rather smeary, and were by no means indistinguishable from a live broadcast. It was also very heavy on tape consumption, restricting the length of the longest possible programme somewhat.
The Quadruplex machine was developed from an earlier Ampex design commonly known as “The Bing Crosby Longitudinal”. This machine now resides in the Ampex Museum of Magnetic Recording. Crosby, the highly popular crooner, had had links with Ampex since the mid-1940s, when he provided the financial backing for the then small company to develop the tape recorders found in Germany at the end of the War.
Little technical information is known of the Bing Crosby machine. The revolving head technique had yet to be devised: it is thought that the Crosby ran at around 360 inches per second, and therefore hadn’t the capacity for large recording durations.
The Quadruplex system used a transverse scanning technique, where the tape was curved into an arc around the head by the use of compressed air.
This was an adaptation of Ampex’s unsuccessful original “accurate scanning” technique, where three rotating heads passed over a flat tape. The Quadruplex system could also be made to run at 7½ inches a second, giving twice the recording time, with different heads and reduced performance.
So good was the Ampex system that it remained a broadcast standard for over thirty years. Due to the nature of transverse scanning, however, it was impossible to fast-wind through a tape while watching the picture without destroying either the tape or the heads. (Quadruplex does allow you to hear the sound while in fast wind, though, due to the fixed audio heads).
What you could do to Quadruplex was physically cut and splice the tapes, with the aid of a special editing microscope (in order to see where the vertical video tracks lay, and therefore where to cut). Duncan Wood used this technique from the third television series of “Hancock’s Half Hour” onward, cutting different takes of scene together during the between-scene black points.
The helical scan system was developed by Sony with Shibaden in the early 1960s. Adopting many of the same principles as the Quadruplex, the helical scan VTR scans the tape diagonally – the tape crosses the head to form a helix, hence the name.
This allows the tape speed to be reduced, thus preventing the quartered segmentation necessary in the operation of four-head Quad machines. An advantage of the helical scan VTR, with the tape wrapped around the head drum, was the possibility of fast-winding while watching the picture. The ability to physically cut the tape was lost, however, as the tape track was so long on the diagonal plane.
The first domestic video recorder launched in the UK was the Sony CV 2000 in 1964. This was a ½” reel-to-reel, helical scan, monochrome VTR, designed for home use. It didn’t have a built-in television receiver, although cameras were marketed. Shibaden and Ampex followed with their own models, neither of which were compatible with the Sony machine.
While the results were hardly broadcast quality, an advance had been made. Sony later brought out the CV-2100, a switchable 405/625-line adaptation of the CV-2000, which was also incompatible with every other format (including the CV-2000, due to different tape speeds being used.) Many of these machines were seemingly sold to Universities for educational purposes.
The real breakthrough came in 1971 (Europe) when the Sony U-Matic was launched. A helical scan, ¾” video recorder, the U-Matic was the first machine to put the tape into cassettes (although SMPTE cartridges for Quad Spot players had been available for commercial playout for some years, these were of limited length – less than five minutes.)
The U-Matic can perhaps be regarded as the most successful video format of all time; perversely enough, it was initially marketed for domestic use. It found great favour in educational establishments and within many companies who wished to show training films. In the professional sector, U-Matic machines revolutionised news reporting in particular. The portable machines (see below) made it possible to shoot a news report and have it broadcast immediately, without having to rush film through the developing process. U-Matic is often confused with the Video Cassette Recorder – a Philips invention.
The fact that the Video Cassette Recorder was given no other name doesn’t help (although machines are nowadays more commonly known by their model numbers, the first of which was the N1500).
While the cassettes were loosely based on the design of the Sony U-Matic cassettes, they are incompatible – the Philips cassettes being loaded with ½” tape. The N1500 ran for a maximum of one hour, as did U-Matic; its successor, the N1700 (also known as the Philips Long Play) halved the tape speed.
Thus, N1500 tapes will not play on an N1700 model. This ability to record such a large amount of data into such a small amount of space is due to azimuth recording, a development of the helical scan technique that debuted in the Philips N1700, and has become a standard feature of all domestic VCRs since.
It was an invention born of necessity, caused by decreasingly slower speeds and thinner tape densities; the guard band between diagonal video tracks became so narrow that the tracks were almost starting to overlap.
The deliberate azimuth error introduced between the tracks negated any adverse effect that could have been caused by track A being mis-read by head B.
In 1976, Ampex and Sony launched their 1″ C-format system, which would live side-by-side with Quadruplex in the professional sphere. Its main advantages over Quad were that it was helical scan (and thus tape search facilities were possible), it was considerably smaller (although you would still have to stand to operate it, bearing in mind that a Quadruplex machine could fill a small room, in a space-limited studio environment, the fact that it was possible to fit nearly two of these machines in the space occupied by one Quad), and it could produce a broadcast quality programme of 2 hours on one tape, compared to 1½ hours on Quad SP. 1976 was also the year that saw the start of the video war in Japan, with the launch of a machine known to millions the world over as VHS.
The Sony Betamax had been launched the previous year: it appeared to have it all – small cassettes, easy to use, complete with built-in tuner and timer (another N1500 first). The cassettes ran for the same amount of time as the N1500 – one hour – and were better quality than the slower N1700.
Both Betamax and the JVC rival, VHS, were launched in 1978 in the United Kingdom. VHS, or Video Home System (a less daunting rendering of the original acronym, which stood for Vertical Helical Scan) had a major advantage over Betamax: it had three-hour tapes; Sony Beta initially had only 60 minutes to offer.
There have been many reasons given for the fact that VHS won out over Beta and Philips’ V-2000 (launched in 1980), both of which were technically superior systems. In America, for example, a lot of pornographic films were shipped in VHS only, creating an instant, loyal market. However the fact is that the excellent V2000 came along far to late to have an impact, while in the US, Betamax marketing had already been brought to a halt by a legal case against it brought by the Hollywood studios, charging that Betamax owners should not be permitted to record copyright material. While this attempt to stem the tide ultimately failed, the legal wrangle kept Betamax off the shelves for over a year and allowed VHS, without an equivalent encumberance, to forge ahead.
In the UK, JVC was part of the Thorn-EMI conglomerate, which also included Ferguson – soon to become the largest manufacturer of video recorders in the United Kingdom. According to the 1982 publication, “Video Techniques” (Newnes Technical), marketing and licensing agreements had been made as follows by the time of publication: Philips VCR (N1500 & N1700)* : 2 manufacturers Grundig SVR** : 2 manufacturers Sony Betamax : 5 manufacturers, plus 2 more US-only firms JVC VHS : 9 manufacturers, plus 8 more US-only firms.
*Withdrawn from the market by 1982, but placed to provide a guide as to the comparative success of the other formats. **SVR was a further adaptation of the Philips VCR machine, that ran at half the speed of the N1700.
The quality, apparently, was considerably poorer than VHS, which was launched in the same year.
The “war” was re-run in lesser detail in 1987, when both Sony and JVC launched high-band variants of their flagship VCRs – Beta ED, and SVHS respectively. While neither dented the market much at the time, SVHS won out by virtue of the fact that the machine was VHS, and would play VHS tapes.
Mention should also be given to a product that never made it beyond the prototype stage. Toshiba’s LVR (Linear Video Recorder) used multiple video tracks with a small, fixed record / replay magnetic head.
The head used fixed scanning but moved down the tape to the next track once the end of the quickly-moving tape (240 ips with 300 tracks to a ½” tape) had been reached.
One problem was a very brief interruption to the picture every so often as the head moved down. It was due for launch in 1980 as a cheaper alternative to the other formats which used then-expensive helical scan technology, but despite a working prototype being shown to the press at the time, it never appeared. Toshiba was then Japan’s second largest manufacturer of Betamax decks, after Sony itself.
In conclusion, it would take a lot of words to assess just what the effect of video recording has had on professional practice. Even the “poor” systems, such as VHS, appear in the professional video chain at regular intervals: an off-line VHS editing suite can be constructed for less than £500.
The invention of the U-Matic portable made it possible for news crews to record outside footage with synchronised sound quickly and efficiently. And, of course, the invention of the Quadruplex made video recording, time shifting and editing possible.
While analogue recording may currently be losing out to digital techniques, many of the technological advances made along the way are being implemented in the new generation of digital machines.
- Anonymous, “Voice Records – The Blattnerphone” from B.B.C year book 1933 (Richard Clay & Sons, Suffolk, 1932) Out of print
- Baird, J L “Television in 1932” from B.B.C year book 1933 (Richard Clay & Sons, Suffolk, 1932) Out of print
- Cain, J The BBC: 70 years of broadcasting (BBC, London, 1992)
- Morita, A Made in Japan (further details unknown)
- Trundle, E Television & Video engineer’s pocket book (Butterworth – Heinemann, Oxford, 1992) ISBN 0-7506-0677-0
- White, G Video Techniques (a Newnes technical book) (Butterworth & co., Southampton, 1982) ISBN 0-408-005206-8
- Wilmut, R F Tony Hancock – Artiste (Methuen, London, 1978)
- www.urbanlegends.co.uk accessed 7 December 1999
- www.staddiscombe.freeserve.co.uk accessed 7 December 1999
- www.lionlmb.org accessed 18 December 1999
- www.islandnet.com/~ianc/ accessed 3 January 2000
- ac.acusd.edu accessed 5 January 2000
- www.bbctv-ap.freeserve.co.uk accessed 13 December 1999. Arthur Dungate’s web page. Arthur worked at Alexandra Palace and Lime Grove during the 1950s and 1960s, and kindly allowed me to interview him on the subject of VERA and the Cintel scanners at the BBC.
- www.billabong.demon.co.uk accessed 28 December 1999
- www.dfm.dircon.co.uk accessed 16 December 1999. Don McLean’s pages on Baird’s Phonovision disc recordings
- batc.org accessed 10 December 1999. The home of the British Amateur Television Club. Thanks to Andrew Emmerson for details of the 1938 telerecording.
- www.jvc.co.uk accessed 3 June 1998. Thanks to Gary Broadbent for confirming the origins of the VHS name.
- www.sony.com accessed 14 November 1999.
- www.playaway.org.uk composed 1995. While this is my own site, I mention it here as I have paraphrased part of this self-written article in paragraphs 2 – 6 of this essay.