Fullscreen HTML5Med Res (??MB)HTML4

The History of Videotape – part 1 Quadruplex By Ian Batty, Andre Switzer & Rod Humphris Analog videotape is now obsolete. But it was state-of-the-art for many decades, and during that time, a video recorder was arguably the most advanced piece of electronic equipment in many homes. The history of video recording is quite fascinating, and this series of articles provides an in-depth explanation of how it came about and changed over the years. www.historyofrecording.com/ampexvrx1000aniv.html A udiotape recording and playback predate videotape, with early magnetic recording of audio demonstrated in 1898. Oxide tape was invented in Germany in 1928. By the time serious work on videotape recording started in the 1950s, audiotape was already widely used. Audiotape use amplitude-based recording; a stronger signal creates proportionally stronger magnetic patterns on the tape. Audio signals are in the frequency range of 20Hz to 20kHz, a range of ten octaves or three decades. This is not especially difficult to achieve with magnetic tape. Videotape, however, needs to cover 44 Silicon Chip the range of 60Hz to at least 4.2MHz for the US NTSC standard, or 50Hz to 5MHz for CCIR/PAL (see Fig.1). This is a range approaching 17 octaves. That’s a much bigger challenge. On playback, tape head output doubles for every doubling in frequency (ie, output increases at 6dB/octave). Let’s say that we can get away with a video signal that has a signal-to-noise ratio (SNR) of 40dB. From 50Hz to 5MHz, the signal ratio due to the 6dB/ octave effect is 100dB! That means that our tape system SNR needs to be at least 140dB (Fig.2). That is simply not possible. So video signals cannot be recorded and played back using Australia’s electronics magazine conventional amplitude recording. Another reason why amplitude recording cannot be used for video is that any tiny variations in tape-to-head contact (dropouts) would severely affect the replayed picture (Fig.3). Variations in the tape’s oxide layer would also cause major visual disruptions, especially if the signal level falls and the synchronising signals cannot be detected. Tape-to-head speeds Tape systems work well up to a frequency where the wavelength of the recorded magnetic pattern approaches the width of the tape head’s magnetic siliconchip.com.au Fig.1: the recording bandwidth needed for a direct (linear) analog transcription of standard audio and video (PAL) signals. The horizontal axis is logarithmic; video covers 16.5 octaves (five decades) while audio covers 10 octaves (three decades). The BBC’s Video Electronic Recording Apparatus (VERA) was an attempt to record video onto tape in a similar manner to audio. It used stationary heads and a very high tape speed, necessitating huge tape reels. Despite their size, each reel only lasted 15 minutes! Source: www.vtoldboys.com Ampex’s Harold Lindsay (left) and Alexander M. Poniatoff (right) with the well-regarded Ampex 200 audiotape recorder. Source: www. historyofrecording.com Fig.2: the signal from the tape head increases by 6dB for every doubling in frequency. This shows the impossibility of recording a video signal directly to tape, since to avoid saturation at 5MHz and signals below 50Hz being lost in the noise, the system would need an impossibly high dynamic range of 140dB. gap. At precisely one wavelength, the signal on one side of the head has the same amplitude and polarity as that on the other side. With no difference in the magnetic field, there is no output from the head. So the combination of head gap width and tape speed determines the frequency at which head output falls to zero, and thus the maximum recordable frequency. For the NTSC limit of 4.2MHz and a practical head gap of only 2.5µm, the required tape speed is 21 metres/sec (2 × 2.5 × 10-6 × 4.2 × 106 × 103). That’s the entire length of an old-fashioned 2400 foot/731m reel in about 35 seconds! It’s siliconchip.com.au worse for the CCIR/PAL bandwidth of 5MHz, needing a tape speed of 25m/s, giving a reel playtime under 30 seconds. So it is not practical to use linear tape recording for video recording. VERA Despite all these apparent problems, some hardy folks did give amplitude recording a try. The BBC’s Video Electronic Recording Apparatus (VERA) from 1952 took on the challenge, using stationary heads and a very high tape speed. Unable to accommodate the required 405-line standard’s bandwidth of 3MHz with amplitude recording, Dr Australia’s electronics magazine Peter Axon’s team ingeniously split the entire signal into three bands. Band A contained signals 50Hz~100 kHz (including synchronising signals), frequency modulated onto a 1MHz carrier. Band B contained signals 100kHz~3MHz using amplitude modulation. Band C frequency-modulated the audio signal onto a 250kHz carrier. Splitting the video bandwidth did allow the 405-line bandwidth of 3MHz to be accommodated, and demonstrated the principle of recording video on tape. VERA’s development lasted until 1956, by which time US company ...continued on page 48 March 2021  45 A A Timeline Timeline of of Videotape Videotape Recording Recording 1956: Ampex VR-1000A The VR-1000A was the first of Ampex’s 2-inch quadruplex recorders (www.flickr.com/photos/82365211<at> N00/2215654688/). Prior to this Ampex had worked magnetic tape systems that were based off the Germans’ work on the Magnetophon. 1965: Ampex VR-5000 One of the first Type-A format VTRs, 1-inch tape, one head and helical scan (www.ebay.com/itm/273727570578). 1969: Philips LDL-1002 Has a recording time of 45 minutes and runs from a 50Hz AC synchronous motor (https://commons. wikimedia.org/wiki/File:Philips_ ldl_1002.jpg). 46 Silicon Chip 1956: RCA TRT-1B RCA’s first workable video tape recorder. Recordings made on the TRT-1B were also compatible with the earlier Ampex VR-1000A (www. lionlamb.us/quad/rca.html). 1965: Sony CV-2000 The world’s first consumer videotape recorder (https://youtu.be/ wHiBxlhzgyY). 1969: Akai VT-100S Records up to 20 minutes onto 1/4inch tape and has a separate camera unit with a built-in mic (https://youtu. be/iaPAyVcXz_0). The difference between the VT-100 and 100S was the inclusion of a stop-motion feature. Commercial Equipment Australia’s electronics magazine 1958: BBC VERA Here is the first live demonstration of VERA in 1958 by Richard Dimbleby: https://youtu.be/YCyxPLXLaKA Source image: http://archive. totterslane.co.uk/tech/vera.htm 1967: Sony DV-2400 The “Portapak” was the first consumer-oriented portable videotape recorder and could record up to 20 minutes (https://en.wikipedia.org/ wiki/File:Sony_AV-3400_Porta_Pak_ Camera.jpg). 1969: IVC 800 A 1-inch videotape colour recording/ playback machine (https://youtu.be/ EIhI85cHIfg). It also has slow motion playback and two audio tracks. Consumer Equipment siliconchip.com.au 1971: Sony VO-1600 The first video cassette recorder; it used Sony’s U-matic system and had a TV tuner (www.labguysworld.com/ Sony_VO-1600.htm). 1975: Sony SL-7300 The first standalone Betamax player, it was called the SL-7200 in America (http://takizawa.gr.jp/uk9o-tkzw/tv/SL6300.pdf). 1976: JVC HR-3300 The first VHS recorder, it could hold two hours of footage per cassette (https://en.wikipedia.org/wiki/ File:JVC-HR-3300U.jpg). 1983: Sony BMC-100P The “Betamovie” is an early camcorder for the Betamax format (https://en.wikipedia.org/wiki/ File:Sony_Betamovie_BMC-100P.jpg). siliconchip.com.au 1972: Philips N1500 This was the first device to use the commonly known VCR format (https:// en.wikipedia.org/wiki/File:N1500_ v2.jpg). 1976: Ampex VPR-2 1974: Sony VO-3800 The first portable U-matic recorder. While it records in colour, it can only play back in black & white, and needs a separate power supply to display colour (www.labguysworld.com/ Sony_VO-3800.htm). 1976~85: Bosch BCN 52 Two Ampex VPR-2s that used 1-inch Type-C videotapes which replaced quadruplex (www.vtoldboys.com/ hw1980.htm). 1976: Sony BVU200 The Sony BVU200 was one of the first “broadcast video” U-matic players before being replaced by Betamax. 1985: Sony Handycam The first Video8 camcorder which succeeded the Betamax-based models (https://en.wikipedia.org/wiki/ File:Handycam-dvd.JPG). Australia’s electronics magazine A 1-inch Type-B recorder with digital timebase corrector (TBC) playback, slow motion and visible shuttle. https://commons.wikimedia.org/wiki/ File:BCN_52_type_B_VTR.jpg 1999: Sony DCR-TRV103 The first Digital8 camcorder. Outside of Sony the only other manufacturer of Digital8 devices was Hitachi. March 2021  47 While there will be some difference in playback signal level between sync tip and peak white frequencies (due to the 6dB/octave effect), these will be removed by the limiting amplifiers used in FM receiver/playback systems. Ideally, the playback response will be flat from 50Hz to 5MHz, the required range of 100,000:1 or five decades. FM signals are recorded at tape saturation level. This ensures a high playback signal, but also removes the need for the tape biasing critical to amplitude systems. Rotating heads Fig.3: a simulation of what you could expect to see upon playback of a linearly recorded video signal due to small variations in the head-to-tape distance and variations in the properties of the tape’s oxide layer. In this example, you can see a large-scale dropout at the top and a few one-line dropouts near the centre. This image was taken from the 1923 episode “Felix the Ghost Breaker” of Felix the Cat (https://archive.org/details/FelixTheCat-FelixTheGhostBreaker1923). Ampex had successfully demonstrated its superior and revolutionary quadruplex system. Already obsolete, VERA first went to air in 1958. VERA’s high tape speed of 5m/s meant that a 520mm diameter reel of tape (over 4.2km!) only ran for some 15 minutes. The American experience was similar to the BBC’s. Bing Crosby Enterprises, owned by popular entertainer Bing Crosby, was already using Ampex 200 audio recorders in their studios. One was modified for a tape speed of 360 inches/s (over 30km/h!), and did play back a grainy image. The Radio Corporation of America (RCA) also demonstrated a linear system. Like VERA, these systems used stationary heads, high tape speeds, and gigantic reels of tape. These linear, amplitude-based systems could not be made practical. The solution: frequency modulation Conventional amplitude modulation must always occupy a bandwidth of twice the highest modulating frequency. Also, it’s impractical to use a modulating frequency more than a fraction of the carrier frequency for AM. Frequency modulation (FM) can occupy any required bandwidth (Fig.4). 48 Silicon Chip Narrow-band FM (NBFM) occupies a bandwidth that’s a fraction of its highest modulating frequency, while broadcast FM uses a bandwidth that’s five times its highest modulating frequency. It’s also possible to frequencymodulate close to the carrier frequency. Video frequency modulators commonly use a carrier frequency of a few MHz for the synchronising signal frequency (synch tip) level (zero signal volts), and a carrier frequency some two to three times that for peak white level (one signal volt). The actual rate of modulation (corresponding to the frequency of the modulating video signal) is accommodated by circuit design. Additionally, frequency-modulated systems are highly immune to variations in signal amplitude. This means that tape dropouts and other imperfections will have much less effect in frequency-modulated recording systems. Could we have linear AM systems for total frequency modulation and overcome the signal quality and bandwidth problems? Maybe. But that would leave the 20km/h-plus tape speeds that made these systems impractical. The solution is rotating head mechanisms. A rotating head moves relative to the tape, as well as spooling from the supply to takeup reel. This was Ampex’s stroke of genius. The magnetic track could lie at a slant angle across the tape, with multiple tracks in parallel (see Fig.5). This means narrow tracks, and narrowing the magnetic track makes the SNR worse. But frequency-modulated systems do not respond to noise for signals of moderate strength, so the designers can define a track width that gives an acceptable SNR for the frequency modulated record/playback system. The tape heads were mounted on a spinning disc, running almost at right angles to the tape’s direction of travel (Fig.6). Known as the headwheel, its rotational speed easily allowed writing/reading speeds across the tape in the metres/second range. This allowed the tape transport’s longitudinal speed to be greatly reduced, giving the practical, standard speed of 15ips or 381mm/s. Readers may anticipate the need for high-precision control of tape speed Fig.4: the basic principle of encoding an analog video signal using frequency modulation (FM) which makes recording it onto tape a much simpler affair. This is essentially the same approach used in analog TV broadcasting. Australia’s electronics magazine siliconchip.com.au Fig.5: the Ampex quadruplex videotape layout. The tape is moving horizontally while the head is moving vertically, so the video tracks are laid down at an angle. The audio, cue and control tracks are laid down in the traditional method, along the length of the tape. and head positioning. These are done by servomechanisms. Servos will be described fully in the following article. Ampex quadruplex Alexander M. Poniatoff founded Ampex in 1944, using his initials, and ex(cellence) for the name. Releasing the high-performing Ampex 200 audio recorder in 1948, Poniatoff and his company anticipated the use of tape recording for television, beginning experiments in 1952. Ampex’s 1956 demonstration of their VR-1000 rendered other designs obsolete, and “quad” would become the industry standard. There were two complications, however. First, although the tape could be wrapped to conform to the circumference created by the spinning heads, any wrap over 90° was impractical. Fig.6: the Ampex quadruplex head mechanism. The head is in the centre while the vacuum shoe, which keeps the tape in contact with the head, is at left. Source: https://youtu.be/ fpBRuheelu4 siliconchip.com.au But, since the video signal is continuous, there must be continuous head-totape contact. So the head wheel was designed to carry four heads, with the tape wrap a little over 90°. This guaranteed continual headto-tape contact, and head switching could be done electronically. The tape was made to conform to the arc of the heads by a curved “shoe”, aided by a vacuum system. The shoe is visible to the left in Fig.6. The head rotational speed was dictated by the minimum acceptable headto-tape speed to give sufficient record/ replay bandwidth, and this meant that only some 16 picture lines could be written or read in one head scan. This meant that any mistiming between heads would distort the picture – an effect known as head banding. To prevent track-to-track interference, unrecorded guard bands were left between each recorded track on the tape (see Fig.7). Also, during playback, it was vital that the heads aligned accurately to the centres of the transverse tracks. The audio was recorded on a linear track, just as with a conventional audio recorder. A control track with alignment pulses was added, and on replay, these were detected and fed to the head servomotor to ensure accurate head tracking and correct picture re-assembly. The high tape-to-head speed, combined with frequency modulation, gave the full video bandwidth without any band-splitting (as in VERA), and high immunity to tape defects. The transverse recording brought two further benefits. Firstly, tape stretch, a serious problem with linear recording, was minimised by the nearvertical track angle. Since the heads were servoed to the index pulses, these would separate or close up as the tape Fig.7: when the “Magna-see” slurry was applied to a quad tape, the video track strips became visible. Each strip encodes 16 lines of video. As there is a gap between the strips, it is possible to cut and splice quad tape by hand. You just need to know exactly where to cut! Australia’s electronics magazine March 2021  49 stretched, keeping the head scanner aligned to the centre of each track. Secondly, each track contained a complete number of picture lines, and it was possible to ‘expose’ these with a fine magnetic slurry called Magnasee, as shown in Fig.7. So editors could visually locate end-of-frame edit pulses and successfully cut-and-splice an original tape with no visual disturbance to the replayed picture. (16 x 64). But that isn’t good enough. Videotape itself is not rigid – it will suffer stretch errors that even the most aggressive servos cannot correct. No mechanical servo can respond with microsecond accuracy, at microsecond intervals. Even errors in the tens of nanoseconds (10-8 seconds) will be evident if the VTR’s output is put to air. Timebase correction VTRs are mechanical gadgets with two critical electromechanical servo systems. The tape transport servo controls the tape speed, and this determines whether the off-tape video will exactly match the vertical rate of station syncs. If this isn’t done, the VTR video will roll vertically and cannot be put to air. The headwheel servo controls the headwheel’s rotational speed, and this determines whether the off-tape video will exactly match the horizontal rate of station sync. If this is not done, the VTR video will slide horizontally, or be offset left or right compared to station sync and cannot be put to air. Remember that in the late 1950s, digital technology was restricted to massive computers the size of a small bus. So the solution was to use an array of switchable delay lines to ‘juggle’ the replay video’s timing, and force it into exact synchronism with the station references. These analog timebase correctors (TBCs) used selectable delay lines with periods from 125 nanoseconds, augmented by a continuously-variable secondary system. Yes, analog TBCs were large, expensive and complex, but videotape could only replace film if the VTR’s playback images could be made to follow station sync. Timebase errors So, Ampex’s VT-100 could record and play back high-quality video. And the playback picture looked fine on a monitor connected directly to the VTR. But it proved impossible to feed that replay video into a studio system for broadcast for reasons relating to station synchronisation. Every TV station has a master reference that generates sync pulses (station sync) for the cameras, the vision mixers and other program sources, ensuring that every image is framed exactly. Every image is absolutely ‘in-sync’ with every other, so that any superimposing (such as a crossfading from one camera to another) shows the two images blending without one ‘drifting’ over the other. This was never a problem with putting film to air; “telecine” used a TV camera that viewed the image from an ordinary movie projector that ran the film, and that TV camera was locked to station sync. But the VTR’s playback signal was not in sync with the station. We can design a servo system that forces the VTR’s tape transport to run at precisely the station’s 50Hz frame sync rate. We can also add a headwheel servo to make sure the headwheel scans exactly 16 lines in 1024 microseconds The analog TBC circuit (see Fig.8) comprises, first, a stepped, digitallycontrolled delay line from 0.125µs to 63.875µs. The coincidence detector senses the time error between the station sync and the off-tape video. The coincidence detector’s control output sets the switchable delay line to a delay which is some multiple of 0.125µs. The output is now stable in time, but it may not be exactly in-phase with the station sync, and this would give an image slightly displaced to the left or right relative to an image from a studio camera. The second stage in the process uses analog processing: the analog coincidence detector sends a control signal to a continuously-variable (analog) delay line. This allows the TBC to ‘trim’ the video output so that it is precisely in phase with station sync. The VTR’s output could then be mixed with any other station source (such as a camera), and show no displacement error across the screen. If this sounds complicated, you’re right. And recall that this was implemented in valve technology. RCA’s TRT-1, competitor to the Ampex machines, is the size of six refrigerators! Over time, design advances reduced quadruplex technology in size and improved video quality. NTSC and PAL colour systems were designed for monochrome compatibility. As quad machines had always had the capability of recording the entire video bandwidth, this meant that they could record and play back colour video too. Timebase correction was vital for successful colour operation. While monochrome systems could tolerate timing and phase errors, the NTSC colour system transmitted colour infor- The TRT-1, RCA’s first 2-inch VTR, took up six full racks (the three racks shown here are half the machine). Each was about the size of a domestic refrigerator. TBCs were required to interface VTRs to broadcast studio feeds. As technology progressed and transistors took over from valves, TBCs shrank, and their capabilities improved. Source: www.lionlamb.us/quad/ ► An Ampex VR-3000 “portable” ► VTR. These were popular with reporters as the tape could be re-used many times, as opposed to film, which could be used only once and then discarded. Source: wikimedia user Gunnar Maas 50 Silicon Chip Australia’s electronics magazine siliconchip.com.au mation as a phase-modulated signal. Any phase errors during replay would create visible shifts in hue; reds might become greenish, giving a deathly cast to the faces of actors and newsreaders. By the time colour television was introduced, advances in timebase correction were able to cope with VTR phase errors, giving faithful reproduction within the fundamental limitations of NTSC. The most advanced quad machine was Ampex’s VR-3000 (shown at lower left). Its ability to record and play back video, and a wide range of other signals, saw it used by the US military as an aid to vehicle- and aircraft-mounted surveillance systems, as well as its peaceful use in replacing movie film as the reporter’s medium of record. Its portability demanded the usual circuit rethinking and redesign. By then, solid-state electronics was well-established as the technology of choice, allowing compact electronics such that it was mainly the mechanical transport which dictated the equipment’s final size. But there was one last challenge. Large quad machines used vacuum or air-pressure systems to bring the tape into proper contact with the headwheel. This was impractical with the VR-3000, so an elaborate and highlyprecise tape guide/shoe mechanism was required. Most quad machines have gone to scrap. Some remain in the hands of dedicated collectors and museums. The few in working order are used to recover archival tapes for digitisation and preservation, or in live demonstrations of this ingenious technology. The operator recalls Randall Hodges was one of the earliest operators of VTR technology. He recalled his experiences for this article. Before videotape, news gathering and other outside-the-studio material was shot on film, or came in by a remote relay. Film had been around since 1923. It had matured by the 1950s – everyone knew how to use it and equipment was plentiful. Film worked fine, but it needed expensive developing equipment and chemicals, and it could only be used once. Processing easily took 45 minutes to an hour. Film copying used specialist equipment and was costly and time-consuming. And if the camera operator missed a shot, if it was siliconchip.com.au Fig.8: the basic principle of analog timebase correction. The correction needs to be continuously variable over a range of 0-64µs. Since it was too difficult to do this in a single stage at the time, a 0-300ns continuously variable delay was combined with a series of switchable 125ns delay lines. out of focus or poorly framed, no-one could tell for sure until the film had been developed and run. So videotape recorders (VTRs) were genuinely revolutionary. You could record and play back instantly, and the audio track could be recorded simultaneously, or separately in post-production to match the vision. You could also copy videotape easily, cheaply and almost instantly. Although a reel of tape was expensive, good-quality tape was OK for perhaps a hundred re-uses, thus making it economical compared to single-use movie film. The VTR made it practical to record shows for repeat transmission, or to pick out segments for inclusion in other shows. Yes, it was possible to film a television monitor (called a “kine” or “kinny”), but the quality was never very good, and duplication of film stock is expensive. ing back a tape recorded on a different VTR: RCA to Ampex, or vice-versa. High-frequency playback equalisation varied between machines, so we would record colour bars at the start of every tape. For an interchanged tape, we would play back the colour bar section and adjust equalisation for each of the four heads. Head wear could also lead to one (or more) tracks being recorded at lower amplitude compared to the others. This would demand adjustment regardless of where the tape had originated. Tape problems Early formulations used “brown tape” (ferric oxide), which was quite noisy and shed oxide like dandruff. This grade of tape would cause head clogs that could wipe out the signal from one head (or all four) completely. Common quad problems Head-banding could be a problem with the early machines playing back their own tapes. Since each video track was only 16 lines, it was vital that each head played back with exactly the same signal strength. It became more common when playAn Ampex quadruplex VTR (video tape recorder) in use. There were various different configurations over the history of the machines; in this case, the controls are next to the tape reels with monitoring equipment overhead, but other machines were narrower with a smaller side control panel and more rack-mounted equipment above and below the tape deck. Australia’s electronics magazine March 2021  51 The improved “black tape” (chromium dioxide) was much better. Its signal-to-noise ratio was superior, and it shed much less oxide. With brown tape, we’d be on standby with a lint-free cloth and a spray can of Freon (later phased out in favour of isopropyl alcohol). The headwheel spins at over 10,000 RPM, and the video head tips are less than a millimetre wide. If you think this sounds like a highly precise circular saw, you’re right! The combination of the shoe curvature and the vacuum guiding system theoretically ensures that each head makes first contact with the tape a little way in from the extreme edge. This prevents the head from catching on the tape edge, and ensures that the tape runs smoothly. Tape damage can take many forms, but edge damage (scalloping) creates a “wavy” edge, and this can allow the video head to impact the extreme edge of the tape. And cut it in half! In the worst case of putting a program to air, we would have to rapidly pause the VTR, open the shoe, draw maybe half a metre of tape through the head stations and wrap it onto the takeup reel, then punch it into play and hope that the tape would make it to the end of the program. Those were fun days! Going to air Servos take some time to run up to speed and lock, with the first generation of quad machines needing eight seconds from pushing play to delivering guaranteed stable off-tape video in sync with the station. We called this the pre-roll or rollback time. But quad VTRs do not give an image in pause, and cannot be played in slow motion, so we couldn’t use any visual cues to set the pre-roll timing. What we would do is find the start of the required program material by rocking the tape backwards and forwards and listening for the start of the audio. We would then manually roll the tape back, counting the one-second cue pulses as we did. We’d hear a “whoop” each time we rolled past a cue pulse, so eight whoops back would give us the pre-roll timing. Because we were rocking the tape manually, it was pretty slow compared to its normal 15ips speed, so the cue pulses’ usual clean ‘pips’ came out spread over time, and at a much lower audible frequency. We’d leave the VTR in pause and wait for the producer’s cue. Let’s say the show’s presenter was going to do a cross to VTR. The producer would know pretty well when the presenter was eight seconds from the cross, and would call up the VTR. We’d hit play, and the VTR would start and lock within the eight-second window. As the announcer threw to the VTR, the producer (or the panel operator) would punch to tape, and the VTR program would go to air. We eventually moved to Ampex AVR-2000s. These had much better servos, reducing our pre-roll times to four seconds. Good as those were, quad technology still could not produce a still picture or slow motion. If you ever used the next generation of helical scan VTRs (“C” format, Umatic, Beta, VHS or Video8), you will probably know that the tape could run at any speed from still frame to picture search, and give a picture of some kind. But quad offered none of these conveniences. It was ‘play or nothing’. Cooked by the valves When I started, we didn’t offer today’s 24/7 service, so the VTRs were turned off after the last show finished. Later on, we just left everything running 24/7. The first generation of valve-equipped VTRs put out a lot of heat. Our first operating rooms had no air conditioning, so it was uncomfortable for us and less than appropriate for the VTRs. Videotape likes the same range of temperature and humidity that people do, and this may have contributed to the poor reputation of the oldfashioned “brown tape”. It was a great relief when we finally got proper air conditioning. Over time, valve technology was superseded by solid state, greatly reducing the amount of waste heat generated by the VTRs and making our lives more comfortable and the machines more reliable. Quad cartridges We would air many shorts; mostly station promos and advertisements. These were recorded on two-minute lengths of two-inch quad tape, held in cartridges loaded into a conveyor system. The idea was that you’d cue up the cart, then hit play and put it to air. But they could be unreliable; so much so that we’d occasionally just record the whole ad break to open-reel quad tape and run it from the VTR rather than the cart machine. Conclusion The authors would like to thank Randall Hodges for assistance in writing this article. Next month, in the second part of four in this series, we describe the helical scan VTR technology and the first round of videotape format wars. References & videos While it looks awkward and bulky by today’s standards, this sort of portable video recording system revolutionised how TV was recorded and broadcast; especially the news. Source: www.labguysworld.com 52 Silicon Chip Australia’s electronics magazine VERA: youtu.be/rWCstPCcuKk An excellent presentation on quad technology: youtu.be/fpBRuheelu4 Editing two-inch videotape: youtu.be/7YtmwB9Ds5Y Cartridge machines: youtu.be/wM_2upiGUO0 Footage of Alexander M. Poniatoff: archive.org/details/cst_00007 A thorough written history: www. labguysworld.com/VTR_TimeLine. htm SC siliconchip.com.au