This is only a preview of the May 2021 issue of Silicon Chip.
You can view 41 of the 112 pages in the full issue, including the advertisments.
Articles in this series:
Items relevant to "Programmable Hybrid Lab Supply with WiFi – Part 1":
Items relevant to "Digital FX (Effects) Pedal – Part 2":
Items relevant to "Arduino-based MIDI Soundboard – Part 2":
Items relevant to "Variac-based Mains Voltage Regulation":
Articles in this series:
Purchase a printed copy of this issue for $10.00.
The History of Videotape – part 3 Cassette Systems By Ian Batty, Andre Switzer & Rod Humphris The Bulletin, Volume 96, Number 4903, April 27 1974, pages 72-73: http://nla.gov.au/nla.obj-1617182059 The previous two articles described the electronic and tape interface systems for video recording and playback, up to the development of VHS & Betamax. While professional/broadcast systems overwhelmingly used reel-to-reel tape, for domestic use, cassettes are much easier to handle. And even at a TV station, when dealing with hundreds of thousands of tapes, cassettes made life a whole lot easier. R eel-to-reel videotape recorders used similar tape speeds to audio recorders. The popular Electronics Industry Association of Japan (EIAJ) standard accepted the audiotape speed of 19.05cm/s (7.5 inches/s [ips]) for NTSC and 16.32cm/s (6.426ips) for CCIR/PAL. Standard 7-inch reels could therefore hold an hour of standard tape or 90 minutes of long play tape, with 5-inch reels offering only 30/45 minutes of play time. While these high speeds gave good audio response, the audio industry’s previous adoption of the compact cassette showed the way forward. 86 Silicon Chip Try as they might, designers of reelto-reel were limited in how far they could miniaturise their offerings. Using smaller tape reels allowed for a smaller deck, but even Akai’s standout VT-100, overlapping its reels to save space, was limited to 30 minutes due to its high tape speed of 21.8cm/s (8.6ips). Sony’s AV-3400, running at 19.05cm/s, also managed only 30 minutes. Nobody was going to consider a home VTR with these running times – you’d need more than four reels to watch the 1956 version of War and Peace, and you would only get it in black and white. As with the final developments of Australia’s electronics magazine portable quadruplex VTRs, machine electronics were shrinking to the point where the tape reels, video head drum and the transport dictated the final size of the design. U-matic Already prominent in the open-reel video recorder market, Sony took the plunge and led the development of VCR systems. Needing a cassette of acceptable size, Sony designers settled on dimensions of 219 x 136 x 38mm. The width and depth were dictated by the sizes of the two reels; the thickness, by the use of 3/4-inch (19.05mm) tape. siliconchip.com.au Fig.33: the U-loading principle used by the Sony U-matic system. It is elegant but mechanically very complex. It was the resulting unreliability that led it to fall from favour. A 60-minute record/play time demanded a slower tape speed for the reel size and length of tape available, and the audio speed of 3.75ips or 9.5cm/s was chosen. While the EIAJ system had been developed for colour recording/playback, a half-inch/12.5mm tape width lacked sufficient head-to-tape speed for acceptable performance at the reduced speed dictated by the smaller reel size in the proposed video cassette housings. The 19mm tape width gave longer video tracks. Run at a tape speed of 9.5cm/s around a 110mm head drum, the U-matic achieved a head-to-tape speed 853cm/s (336ips). The U-matic’s electronic and head drum design was an evolution, but tape handling would need a revolution. The tape would somehow need to be drawn out of the cassette shell, wrapped 180° around the head drum, engage with the stationary erase, consiliconchip.com.au trol track and audio heads, and be sandwiched between the transport capstan and pinch roller. This mechanism would be the precedent for all subsequent VCR systems. The solution was the loading ring (see Fig.33). The U-matic cassette was loaded by pushing it into the carrier, then dropping it over the loading mechanism. A cutout in the cassette shell allowed the main extraction guide to sit behind the tape inside the cassette. At the same time, the cassette door flipped open. On loading, the loading ring rotated clockwise, with the main extraction guide pulling the tape out of the cassette and drawing the capstan and up to six path guide pins behind it. The tape presents its oxide surface outwards, so the loading mechanism wraps the tape oxide against the erase head, video head drum and heads, audio and control heads and the capstan. Australia’s electronics magazine The pinch roller contacts the back of the tape (unlike in audio compact cassettes), so the likelihood of the tape sticking to the pinch roller is greatly reduced. Once the tape is fully loaded, the capstan and head drum both spin up to operating speed. On playback, a solenoid closes the pinch roller against the capstan (you can see a video of this at https://youtu.be/AFu7FhBDCrA). Contact with all heads (erase, video, control and audio) is by tape tension alone. There are no pressure pads. Three adjustable guides (master entry, video entry, video exit) position the tape precisely; it must be aligned vertically to micrometre accuracy so that video tracks on the tape will exactly match the path of the video heads. All non-video heads are aligned manually to match the positioning determined by the three adjustable guides. The audio exit guide is a simple pin with no vertical adjustment. Audio alignment relies on the video exit/audio entry guide, perfect alignment of the capstan spindle and vertical/azimuth adjustment screws for the audio and control track head mounting platform. The cassette reels rotate in opposite directions. While this seems odd, it means that the inner circumferences are going in the same direction, and this allows the tape from the fuller reel to intrude into the space vacated by the emptier reel, it also helped keep the tape tensioned. There is not enough space inside the cassette for two full reels! It’s an engineering marvel. The head of U-matic development, Sony’s Nobutoshi Kihara, urged his principal engineers Akinao Horiuchi and Yoshimi Watanabe to produce “Nothing too complex, try to find a simple and reasonable design. Remember that it must be easy for people to use.” Horiuchi and Watanabe did produce a machine that was a snap to use: insert a cassette, wait a few seconds, hit play. Internally, it’s a mechanical jungle. Fig.34 shows an exploded view of just the loading ring, giving some idea of the mechanism’s complexity. The initial design only extracted and threaded the tape in play or record, with fast forward or rewind seeing the tape withdrawn into the cassette. This reduced tape wear, but could only rely on an inaccurate, uncalibrated mechanical tape counter. May 2021 87 But the control track contained a highly accurate 25 pulse-per-second signal, one ‘pip’ for each recorded television signal frame. A revised tape mechanism used two arms to draw the tape part-way out of the cassette and engage it over the control track head as soon as the cassette was inserted. This ‘half-load’ allowed the control track circuitry to pick up the control track signals and to drive an electronic tape counter in rewind and fast forward. Play and record would still need full tape loading, and the tape counter would work in both these modes as well. Each end of the tape was spliced to a short length of transparent leader. Optical sensors were triggered by the change in opacity to signal the end of the tape and to stop any current play, record, rewind or fast forward. Some models also offered an auto-rewind feature. Recording format Aside from housing the tape in a cassette, U-matic is pretty similar to formats that preceded it. The slanted video tracks occupy almost 80% of the tape’s width, with the linear control track at the top edge, and two linear audio tracks at the bottom edge. Each linear track has an unrecorded strip on either side (a guard band) to prevent pickup from adjacent tracks. Stereo audio recorders do the same thing to provide separation between the left and right channels. The video tracks also use guard bands. Being only 85µm wide, severe demands are placed on the mechanical and electronic alignment of the VCR’s mechanism and transport. So U-matic’s designers allowed a 52µm guard band between the video tracks. This works just fine in practice, but it’s giving up almost 40% of the total tape real estate. Guard bands would become a target for the next generation of VCR designs, as engineers tried to pack as much signal information as possible on smaller, and slower, tape systems. The width of all tracks, and their spacings, have been exaggerated for clarity in Fig.35. In reality, there are some 110-plus video tracks across the width of the almost 15mm allowed for video recording. Notice that the head gaps are perpendicular to the video tracks. This is unremarkable, as it’s how audio and video systems commonly work. Indeed, any off-perpendicular azimuth error causes significant loss of high-frequency playback both in audio and video systems. Fig.34: and here you can see just how complicated the U-matic loading ring was. We would hate to have to pull it apart to replace worn components! 88 Silicon Chip Australia’s electronics magazine siliconchip.com.au Sony’s initial release Sony’s 1971 release of the VO-1600 model U-matic (Fig.36) offered a builtin tuner and TV signal output, and was aimed at low-cost markets, including domestic consumers. While it succeeded in education and industry, its cost, size and one-hour runtime saw it fail to take off in the domestic market. The VO-1600 also lacked a timer. Although Sony offered an external timer/ tuner at extra cost, the VO-1600 failed to meet all the criteria for a home VCR that anyone could just put in the stereo shelves and use with no ‘sidecar’ equipment. Readers are probably more familiar with the VO-1800, which lacked the inbuilt tuner, and the VP-1000, which was a player only. Meanwhile, in Europe… Philips released the 1-inch EL3400 in 1964, and entered the domestic Fig.35: the layout of the tracks on U-matic tape. The guard bands were necessary to prevent cross-track interference but took up quite a bit of space. open-reel market with half-inch VTRs beginning with their 1969 release of the desktop LDL-1000. Although easy to use, it lacked a tuner, forcing users to have existing TV receivers modified to supply video and audio signals for the VTR. Such modified sets were known as receiver monitors. The LDL-1000 achieved some success, but recalling the success of their audio Compact Cassette system (July 2018; siliconchip.com.au/ Article/11136), Philips began devel- Fig.36: a Sony VO-1600 VTR which used the U-matic system. It also had a built-in TV tuner and TV signal output. Source: www.ebay.com/itm/163608576903 siliconchip.com.au Australia’s electronics magazine opment of a cassette system for video recording. Their N1500 (Fig.37), released in 1972 (just one year after Sony’s U-matic), offered an integrated design. Containing a tuner and a timer and able to supply a standard television signal output, the N1500 hit the spot with consumers, except for the problem of tape length. The N1500 can claim to be the world’s first domestic VCR (video cassette recorder). Philips’ VCR system mechanism, like their compact cassette mechanism, was offered royalty-free to manufacturers who agreed to maintain the design standard and use the VCR logo. You can see a video of a VCR tape loading at https://youtu.be/9-Bw8m65mVY The VCR cassette stacked the supply and reels above each other in a coaxial design. At only 125 x 145 x 40mm, it was much more compact than the standard U-matic cassette. Its width (under 60% that of U-matic) helped moderate the size of the entire tape drive mechanism. While this elegant solution offered a genuinely compact medium, the complexity of its threading mechanism meant that its reliability was only fair. Using a half-inch tape with a conventional 180° degree omega wrap (Fig.38), the Philips VCR was able to offer 60-minute record/play times May 2021 89 Fig.37: the Philips N1500 VCR had an integrated tuner and timer, making it the first VTR suitable for use in the home. But the maximum recording length of one hour meant that as soon as Betamax and VHS came along, it was obsolete. Courtesy of Greatbear Audio & Video Digitising: www.thegreatbear.net/ project/philips-n1500-n1700/ at the CCIR/PAL speed of 14.29cm/s (5.63ips). Philips attempted to market to the United States in mid-1977, but NTSC’s higher field rate (60Hz vs CCIR/PAL’s 50Hz) forced an increase in tape speed to around 17.2cm/s (6.8ips), giving only 50 minutes for a cassette. A thinner tape, offering the full 60 minutes for NTSC, proved unreliable in use. Other compromises finally made their VCR unsuitable for the American and other NTSC markets, while the introduction of VHS in 1977 convinced Philips to abandon the US market. As a result, their VCR was only market- ed to the UK, Europe, Australia and South Africa. Philips tape loading is simpler than that of the U-matic (see Fig.39). Sony had put every interaction (transport, heads and guides) in the external tape path. Philips cleverly used two cassette doors: an upwards-hinging one at the front for tape extraction, and a sliding one at the right, allowing the audio/control track head and the pinch roller to intrude into the cassette. Video entry and exit guides, and the capstan, also intruded vertically into the cassette as it was loaded downwards, giving a much more compact tape transport than that of U-matic. The pinch roller and audio/control heads, mounted on a pivoted arm, were swung into place for playback and recording. Where the U-matic head drum was designed with slip-ring contacts from the heads to the VCR electronics, Philips used a rotary transformer design that had already been used in Ampex 1-inch open-reel VTRs. Although more difficult to design and manufacture, the rotary transformer overcame noise and signal loss caused by slip-ring corrosion or misalignment. It would become the design of choice in Beta, VHS and following formats. The N1500 was developed as far as the N1520 production model. Dispensing with the inbuilt tuner, the N1520 offered recording/playback and full electronic assembly/insert video and audio editing. Released in 1973, it beat Sony’s VO-2850 workalike U-matic editor to market by a full year. Fig.38 (below): the tape loading mechanism of the Philips VCR. It used a 180° omega-wrap which, combined with the half-inch tape, made it significantly more compact than the Sony U-matic system. Fig.39 (above): a direct size comparison between the Philips VCR system and Sony U-matic. 90 Silicon Chip Australia’s electronics magazine siliconchip.com.au Regrettably, the Philips VCR format suffered from unreliable tape loading/ handling, and that dreaded one-hour time limit. Philips did develop a long-play VCR, the N1700 series, by halving the tape speed. Not released until 1977, when the Sony-JVC/Beta-VHS melee was well underway, the Philips VCR lapsed into obscurity. The follow-on Video 2000 suffered a similar fate (see https://youtu.be/SeSz6MoX00Q). Panasonic Video Cartridge Wanting to join the race, National/Panasonic came out with the NV5120 video cartridge (Fig.40). Based on their reel-to-reel half-inch EIAJ colour VTRs, these machines used a video cartridge containing a single tape reel of 30 or 60 minutes duration. The format was properly known as EIAJ2 or EIAJ-M. For loading, the tape was driven out with a stiff transparent leader. This was captured by the transport and driven along a slot that encircled the head drum. The leader would catch onto the internal takeup reel, and normal playback/recording would be available once the leader had been taken up, and the videotape proper followed. The tape was permanently engaged, so fast-forward and rewind offered picture search. While convenient (about the same size as a Philips VCR cassette), the Video Cartridge could not be developed beyond a 60-minute playing time. Also, you were forced to completely rewind the tape before ejection. Panasonic’s Video Cartridge had one unique ability: it was possible to do high-speed copying. Conventional tape mechanisms had to pass the tape over the video head drum for recording, and it was impossible to do this at any higher than standard play speed, as the video tracks would not be laid down correctly. But the tape from a Video Cartridge could be extracted and laid oxide-to-oxide against a master tape, and wound onto a takeup reel. The master/copy tape pack was then subjected to an intense, high-frequency magnetic field that transferred the magnetisation from master to copy. While this would usually erase a tape, the master’s formulation had such high coercivity that its recorded patterns were unaffected. Copying a 60-minute tape took around three minutes. Ironically, this was mostly the time taken to transfer the audio track using conventional high-speed techniques. Betamax vs VHS Leveraging off the success of U-matic, Sony’s Betamax, released in Japan in 1975, should have dominated the domestic marketplace. It had an all-in-one design, inbuilt tuner, RF output for direct connection to a standard TV set, conveniently-sized cassette and colour recording and playback. The second ‘format war’ saw Sony’s Beta face off against JVC’s Video Home System (VHS). Betamax was not just named after the second letter of the Greek alphabet. Rather, it’s an Anglicised version of the Japanese term used to describe the way signals were recorded onto tape and the letter β resembled the tape path through the loading system. The cassette size (155 x 95 x 25mm) appears to follow Masaru Ibuka’s declaration that it should be “the size of a Sony diary”. One wonders whether any brave individual thought of saying “I am most sorry, Ibuka-san, but you just can’t get enough tape into that size for a decent playing time”. It seems no-one did, and, and so the seeds of Beta’s downfall were sown. Sony retained the proven “U” loading principle, reversing the loading direction (see Fig.41 and https://youtu. be/1i_xirpJ550). Some describe this as the “B” loading system. Like U-matic, Beta suffered from slow loading/unloading times. Apart from size, Beta’s mechanism differed from U-matic in several ways. First, the tape was left fully threaded for all modes: play, record, fast-forward, rewind and pause. This allowed users to step between modes much more rapidly than with U-matic, which either wholly or partly unloaded for rewind and fast-forward. Beta also used two extraction guides rather than U-matic’s initial single guide. The master entry guide is mounted on a swing arm and draws tape to the left over the erase head. The main extraction guide is mounted on the loading ring with the pinch roller and other guides. Rotating anti-clockwise, it loads the tape to the right and wraps tape around the head drum and over the control/audio heads. Beta also swapped the positions of supply and takeup reels within the cassette, with both reels rotating in the same direction. Some later models reversed the loading direction, reverting to that of U-matic (see https:// youtu.be/1aFtDRtzKA0). Third, Beta used conventional sideby-side reels, rather than the overlapping design of U-matic. Finally, Beta used metallic leaders on each end of the tape. Pickup coils at each end of the tape path are driven by oscillator circuits. When a metallic leader passes by, the oscillator’s Fig.40: a Panasonic “Video Cartridge” VTR. As with the U-matic and Philips systems, its maximum onehour recording time was the final nail in its coffin. Source: www. labguysworld. com siliconchip.com.au Australia’s electronics magazine May 2021 91 Fig.41: the Betamax tape path. While Beta video quality was somewhat superior to VHS, once again, it was the maximum recording duration (initially one hour) that doomed it. VHS was also arguably a more elegant mechanical solution. Fig.42: when the playback azimuth differs from the recording azimuth by just a few degrees, high-frequency signals are severely attenuated. This was taken advantage of to prevent track-to-track crosstalk, by recording adjacent tracks using heads set at different azimuths. activity changes sufficiently to signal the end of the tape to the VCR’s system control circuitry and the tape is stopped. Azimuth recording U-matic was an oddball format, us92 Silicon Chip ing ¾-inch tape in a cassette that allowed one reel’s tape pack to overlap the other reel’s vacated area for both to fit in. Beta went back to the proven tape width of half-inch, with conventional side-by-side tape spools. Due to the low tape speed necessitated by Australia’s electronics magazine the small cartridge, steps had to be taken to pack the video in as much as possible. The first economy was to dump the guard bands used all the way from Quadruplex to U-matic, reclaiming up to 40% of the available tape area. But now, it would be impossible to prevent a video head from picking up some signal from the tracks adjacent to its intended track signal. Sony’s solution was azimuth recording. As noted above, tape recording formats (of all kinds) commonly align the head gap to be precisely perpendicular to the tape. Fig.42 shows the effect of azimuth errors. In the top diagram, a perfectly vertical tape head gap scans identicallymagnetised areas across the width of the tape, and a unique signal (the originally recorded one) is recovered perfectly. The lower diagram shows that if the head gap is off-vertical, the gap will scan differently-magnetised areas across the tape. Multiple signals are recovered, and the effect is to ‘smudge’ the amplitude of high-frequency signals. So if a playback tape head is off-azimuth to the original recording, there’s a severe loss of high frequencies during replay. This effect is exploited in azimuth recording. Each head’s gap is offset from the other; Beta uses angles of +6° and -6°. Beta’s FM luminance signal uses frequencies between 3.8MHz and 5.2MHz, and the 12° difference between the even field track and the odd field track pretty well eliminates crosstalk. This means that, even if the odd field track’s head happens to overlap onto the even field track, it cannot detect the even field signal due to its azimuth error. Minor tracking errors will not affect FM luminance playback. Betamax release The SL-7200 (Fig.43) was released in 1976. It featured inbuilt VHF/UHF tuners, but needed an external clock for timer recording, and you couldn’t automatically record more than one show at a time. But Beta’s biggest problem was the short recording/playback time of only 60 minutes. Sony seems not to have learned from their own experience with U-matic’s limited tape time, or to have noticed the same issue with Philips’ VCR format. siliconchip.com.au Fig.43: a Sony SL-7200 Betamax VCR. Source: http://takizawa.gr.jp/uk9o-tkzw/tv/SL-6300.pdf While U-matic’s one-hour duration had been acceptable for industry and education, how was anyone expected to record, for example, an American Football game that would often run for three hours? Yes, you could pause the tape every time there was a stoppage of play or a commercial break. But then you might as well just watch the game. And what about your favourite movies? Hardly anything is going to fit on just one cassette. Video rental shops would get behind a format that could put an entire movie on just one handy cassette: VHS. And why, oh, why, use a cassette top that only showed the supply reel (Fig.44)? Yes, you could tell when a tape was fully played/fast-forwarded, but how do you know much you’ve used once you start? Some two-window cassettes were made (Fig.45), seemingly trying to catch up with the more informative design of VHS cassettes. JVC’s Video Home System VHS seems a bit of a Betamax copycat. Sony had consulted with JVC and Matsushita (National) in the early 1970s, aiming to unify a new design based on the U-loading format. Sony engineers were dismayed to find that JVC’s advertising of a ‘new’ video format used elements of Beta’s design: azimuth recording and rotating-phase heterodyne colour (described in more detail below). The success of VHS follows from such a simple idea that you wonder how Sony missed it: enough tape to Fig.45: a later Betamax cassette which added the much-needed second window. But it was too late; VHS was already winning the format war. Fig.44: a standard Betamax cassette. The single window was also a strange design decision as it made it difficult to judge just how much of the tape you had used up. siliconchip.com.au run for two hours without needing long play and its compromises. VHS’s longer tape length, and consequently larger tape reels, required a cassette 187 x 108 x 25mm in size (Fig.46). But VHS is not a simple copycat. JVC probably considered the “U” loading system, but adopted the quicker, simpler “M” load. This uses two arms that extract the tape and draw it out to either side of the head drum (see Fig.47 and https://youtu.be/MPYrKtmuQ41). There are arguments that M loading increases tape tension and wear, but its loading speed, more compact size and its lack of tape-hanging-in-mid-air paths combined to make it the technology of choice for VCRs. However, note that there was an oddball Grundig VS-340 that used B-loading. Given that all the loading mechanism has to do is get the tape onto the drum, it obviously worked well enough, and the user would never know the difference. VHS cassettes used transparent tape leaders. A small lamp on a post intruded into the cassette, and two optical sensors (one on the supply side, one on the takeup side) viewed the lamp via small ‘tunnels’ in the cassette body. Normally, the opaque tape would block the sensors’ view of the lamp, but the leader would allow light through and signal start/end-of-tape. This lamp was vital to proper tape handling, so the VCR’s control system would test the lamp for continuity before allowing operation. Service techs were frequently reminded, for a VHS set with “no operation”, to check the tape sensor lamp first. Following JVC’s release of the HR3300 in 1976 (Fig.48), National Panasonic came on board. Video hire companies endorsed the much longer playing time that VHS offered in standard play and VHS would come to dominate home video recording. Track layout Fig.46: the now-familiar (to anyone over 35, anyway) VHS cassette. The track layout for VHS is shown in Fig.49. VHS uses ±7° azimuth offsets between the video heads/tracks, but otherwise works just like Betamax. While the offset azimuth works fine for luminance frequencies above 3MHz, it is ineffective for the down-converted ~627kHz (626.953kHz) chroma signal. Lower frequencies are less affected by azimuth errors, so some other means of eliminating chroma crosstalk was needed. Australia’s electronics magazine May 2021 93 Fig.47: the VHS tape path. It uses M-loading, where the tape is pulled onto the head drum by two sets of moving guide wheels. This makes for a more compact mechanism. The solution was to take the chroma signal and progressively rotate one track’s phase by 90° for each scan (let’s call it the B track). The other (A) track was recorded ‘as is’. On playback, a two-line delay would give cancellation of the undesired chroma signal. It’s a bit complicated, so let’s just leave it at that – you can check the references below if you’d like to delve more deeply. Sound quality With a tape speed less than that of the Compact Cassette, audio qual- ity was going to suffer. It had only been just adequate with the Philips VCR system, with a bandwidth of 100Hz~12kHz. Beta managed to get 50Hz~10kHz at standard play and 50Hz~7kHz for long play. VHS managed 50Hz~10kHz standard play, but, depending on the model, only up to 4kHz for long play; barely better than telephone quality. Engineers had already packed a good part of the video signal’s bandwidth onto half-inch tape with an ingenious combination of FM and AM recording. Given that FM broadcast Fig.48: an early JVC HR-3300 VHS VCR from around 1976. Source: https://en.wikipedia.org/wiki/File:JVC-HR-3300U.jpg 94 Silicon Chip Australia’s electronics magazine radio could give a high-quality stereo performance, why not employ FM for the audio channel? That would also provide the option of stereo audio. And that’s what they did. Hifi audio recording fed program audio to frequency modulators and then onto the tape. While the electronic design was already available (FM transmitters, FM receivers), the problem was where to put the signal within the available tape bandwidth. Colour Betamax VTRs split off their luminance and chrominance signals, using frequency modulation for the luminance at frequencies above 3MHz. This had left a band centred around 620~650kHz for the amplitudemodulated chrominance signal, and it only extended to around 1MHz. So why not put the FM sound in at about 1.5MHz? Going to 1.5MHz FM audio meant that the audio signal would be recorded in the video section of the tape, and would have to be recorded by the rotating video heads along with the video signal. That’s where the available spectrum existed, and it would have been quite impossible to record any frequency higher than about 10kHz on a linear track, let alone the approximately 1.5MHz FM audio signal. Sony shifted the luminance signal up the spectrum by 400kHz to make extra space available, then used four FM signals: Head A 1.38MHz (left) and 1.68MHz (right), and head B at 1.53MHz (left) and 1.83MHz (right). This allowed Sony to continue using just two video heads, and, in some models, to provide for an external hifi audio processor. For VHS, though, there wasn’t enough spare spectrum, so VHS hifi used depth multiplexing (Fig.50 shows the complete VHS hifi recording spectrum). The FM signal would penetrate the tape’s oxide layer to a depth of around 1µm, while the higher-frequency video signal would only penetrate some 0.3µm. This saw the VHS audio FM signal recorded by a separate pair of record heads, placed some 60° in front of the video heads. The audio heads needed to record first, as the audio signal’s greater depth penetration of around 1µm would have erased the shallower 0.3µm video, had the video been laid down first. While the existing audio signal was partly erased by the following video, the erasure was only shallow. The siliconchip.com.au remaining audio magnetisation was strong enough to be successfully recovered, with the benefit that, being frequency-modulated, any minor tape imperfections would not affect sound quality. Unlike Beta, VHS hifi could not be added to existing two-head VCRs. VHS used either two-head linear audio or four-head hifi/two-head linear. In common with broadcast FM, Beta/VHS hifi used preemphasis at the upper end of the audio band to improve signal-to-noise ratios. This preemphasis was removed by a deemphasis circuit during replay. Also, a companding (compressing-expanding) system compressed the dynamic range during recording from 80dB to 40dB. Left uncorrected, such compression would sound most unnatural, with quiet sounds made unnaturally loud. On playback, the off-tape 40dB dynamic range was expanded back to the original 80dB. With a few other tricks, VCR hifi audio managed a signal-to-noise ratio of 80dB, frequency response of 20Hz~20kHz, with wow and flutter (speed variation) of just 0.005%. And it met these specs at standard, long and triple play. The resulting audio quality was pretty much indistinguishable from Compact Disc. There were even some hardy souls who used hifi VCRs as high-quality audio recorders. Fig.49: unlike U-matic tape (shown in Fig.35), VHS has no guard bands between the video tracks, allowing for higher density and thus longer playback/ recording times. To prevent crosstalk between tracks, they are recorded with alternating azimuth offsets of ±7°. rior colour performance. A side-by-side replay of standard colour bars shows better definition and less noise/artifacts in the colour signal than for VHS. A pity about the one-hour cassette. We’ll look at Super-Beta and S-VHS in the next (and final) article in this series. Part four will also describe how manufacturers responded to the demand for ever smaller and lighter VCRs. We’ll also have a short bit on LaserDisc for those who thought we had forgotten about it. References • Video Cassette Recorders, Humphris, Rod, 1998, TAFE Course Notes • U-matic development by Sony: www. sony.net/SonyInfo/CorporateInfo/ History/SonyHistory/2-01.html • The Impossible Feat inside Your VCR, from Technology Connections: youtu.be/KfuARMCyTvg • The VHS cassette was more clever than Beta: youtu.be/hWl9Wux7iVY • Also check out the rest of his YouTube channel: youtube.com/channel/ UCy0tKL1T7wFoYcxCe0xjN6Q • The history of VTRs before Beta and VHS: www.labguysworld.com • An extensive picture gallery of Philips VCR, Beta and VHS: www. oldtechnology.net • Special thanks to Rewind Museum for the use of various images: www. SC rewindmuseum.com Was Beta Better? Arguably, yes. Beta’s wider FM bandwidth offers somewhat superior luminance definition. Specifications put Beta (luminance resolution 260 lines) a little ahead of VHS (240 lines) at standard play. Beta’s use of a high-amplitude pilot burst for colour correction gives supesiliconchip.com.au Fig.50: the spectrum of hifi VHS recorded onto tape. It’s essentially the same as standard VHS but with the addition of two audio channels frequency-modulated onto 1.4MHz and 1.8MHz carriers. Australia’s electronics magazine May 2021 95