This period saw the wide acceptance of MIDI and a shift away from established BBC Engineering standards. In November, 1986, Jonathan Gibbs left the Workshop to become Manager Operations Groups 2 ＆ 3 (Radio) and it was decided that he would be replaced by an ‘ordinary organiser’. In 1987, Richard Attree, the last composer to be recruited by the Workshop, joined the department. Richard’s work, although not to everyone’s taste, was very slick, polished and well-presented. Unlike previous composers, he was new to the BBC and unfamiliar with the Corporation’s strange ways.
Dick Mills, for reasons only known to himself, now called the author ‘Biggles’, as confirmed by a sticker attached to the engineers’ door. From March until July of 1987 the author worked in the post of Senior Engineer, Maida Vale, whilst Mark Wilson took his place. At the end of this period Mark was given an attachment as a ‘sort of organiser’, but at a reduced grade.
In 1988, Brian prepared the 30th anniversary brochure for the Workshop and also celebrated his 50th birthday with a Mac-shaped cake. On March 28, the Workshop held another party, this time to celebrate its thirtieth anniversary, the last such event. Meanwhile, rumours were circulating that Mark Wilson had been offered a two-year contract as the department’s Development Coordinator.
In July of the same year, an extra composer was attached to the Workshop. This was Steve Marshall, an ‘also-ran’ from the interview board that selected Richard Attree. He was highly talented and didn’t hesitate to use the tape loops and loop stands that had been discarded by most composers. His attachment was extended whilst Malcolm Clarke was away with his back in traction. Steve’s departure in September was celebrated at the Frog and Firkin, located near the Portobello Road. The Firkin brew-pubs, sadly now extinct, were established by David Bruce, originally under the Bruce’s Brewery banner. Their strongest real-ale brew was Dogbolter, a name that described how two dogs reacted when their drunken owners had fallen into a northern beck. Steve himself described how he once provided music for a ‘magician’ who was to escape from a trunk using a key that he’d secreted on himself. Unfortunately, he wasn’t actually given the key, so Steve kept hammering away on his keyboard until they eventually slid the trunk off the stage. Presumably, he was eventually released.
By this time, Dick Mills had been offered early retirement but was understandably reluctant to take it. The uneven workload on the composers led one composer to contend that the department had ‘two passengers’. In reality, the kind of material expected of the Workshop had also changed significantly.
As well as a BBC salary, all of the composers received a ‘performing rights’ payment whenever their music was used anywhere in the world. These extra funds supplemented their income, sometimes over many years. Mark Wilson, in the meantime, was looking at a house in Brixton, about which he said ‘the only good thing is the walls’.
During this period three hundred items of equipment were marked with security paint and labels!
By 1986, two further studios were updated, both with Soundcraft 2400 consoles. These desks were similar to the earlier 1624 mixers but had 24 group outputs. This meant they were ready to accept a pair of 24-track (two-inch) machines that would eventually arrive from BBC Scotland.
The new Studio C was constructed on the original site, but expanded eastwards to include a room formerly used by Tommy Hardcastle. For many years Tommy had cared for the musical instruments used by the BBC orchestras. Although rarely seen, he was known to exist by virtue of the rattle of metal coat hangers heard through the walls each morning and evening!
Initially, the large size of the expanded studio shocked some people. However, the end nearest the main door was available for use as an acoustic recording area. It also could act as an ‘escape’ area where the composer could relax or have discussions with clients.
Studio E (Room 14) and Studio F (Room 13) were converted into an office for Orchestral Management whilst the new Studio E took form in the former Film Unit Viewing Theatre. Since this area was very wide, an elongated lobby was constructed between the studio and the corridor. This provided a much-needed storage area and gave extra sound insulation to the studio itself.
The new studios differed from earlier installations. For example, each tape machine or gramophone was connected via a 56-way Varicon connector: much cheaper and more convenient than a Hypertac. Also, the multitrack pattress had three Varicons, each carrying eight send and return circuits. To accommodate this new standard, the original 8-track Bel noise reduction units were modified: the original XLR connectors at the rear were replaced by a pair of Varicon connectors.
Previously, all studio wiring was in BBC PIN/10 or PIN/20 cable, both developed from Post Office varieties. These contained twisted-pairs of single-stranded wire that were screened and insulated. Unfortunately, they were also very stiff to handle and the single strands had a nasty habit of falling off. Instead, the new studios had multi-stranded Belden cable throughout, with crimp connections at the Varicons and ‘D’ connectors. This made for rapid wiring and high reliability.
In January of 1986, Studio E was in service and a temporary Studio F was in use within the old Film Unit Projection Room (V7). Initially, the latter had an existing low-cost Seck mixer and BBC DK2/21 monitoring unit, but these were replaced in the same year by a Soundcraft 200B 8-channel desk.
In September 1987, plans were afoot for a new Studio H, to be built in Room 19, further along the corridor. This was for Richard Attree and would replace the old Studio H (Room 11) with its Series 2 desk. Once again, a new Soundcraft 2400 desk was chosen, the last to be installed in the Workshop. Although the room contained a structural pillar, the desk could be jiggled into a comfortable position. The studio was to have a black ceiling and a red ‘snake’ lighting system, both chosen by Richard.
Other innovations were introduced: for example, all MIDI tielines and other non-critical circuits were assembled using ‘round and flat’ ribbon cable that was wired to insulation displacement connector (IDC) ‘D’ sockets. This made for very fast construction. Unfortunately, in 1990, some of the original ribbon wiring had to be replaced to prevent timecode from being induced onto the VCR audio outputs.
The equipment rack had ‘clip-on’ boxes for IEC mains distribution panels, MIDI tielines and audio connections. The latter were arranged in blocks of eight send and return circuits and were wired to Varicon connectors according to the Soundcraft standard, in exactly the same way as a multitrack.
The BBC’s Director General, Michael Checkland, opened the new studio on April 19, 1997, with the aid of a cake shaped as keyboard. Brian did a little speech. The DG replied that he didn’t understand all his words and that he’d only come for the cake: ‘I declare this cake well and truly cut’. To quote Brian Hodgson: ‘the cake was a mess, as Guy [the chef] had hit the LSD the day before and it was thrown together that morning. To make matters worse, it was cut with a knife that had been used earlier for chopping onion and had not been washed, so its taste was as bizarre as its appearance’.
As previously mentioned, Syncwriter was fitted with the Musical Instrument Digital Interface (MIDI), a system for transferring musical information between synthesisers and computers. This revolutionary interface would eventually allow composers to create and edit musical performances with ease. By modern standards, its transfer rate of 31.25 bits per second (bits/s) was very slow. However, this was originally chosen as a sub-multiple of processor speeds of the time, typically 1 MHz.
In addition, MIDI inputs used an opto-isolator of limited speed. Each receiver included such a device to avoid electrical interference with audio circuits. In practice, balanced circuits such as RS-422 probably would have worked just as well. In the event, MIDI didn’t suffer from speed problems: the handling of data was cleverly arranged to take full advantage of the available bandwidth.
Opto-isolators also limited the theoretical cable length to 15 metres. The symptoms associated with excessive cabling were erratic, sometimes appearing as ‘sticking’ notes. Tests conducted by the author in November of 1987 indicated that it was possible to send MIDI along 50 metres of cable, assuming that the cable had a capacitance of 30 picofarads per metre (pF/m), resulting in an overall capacitance of 1.5 nanofarads (nF). In practice, shorter distances were always used to ensure reliable operation.
To convey MIDI over intermediate distances, several MIDI to XLR adaptor cables were manufactured. These were wired with pin 4 of the DIN plug joined to pin 2 of the XLR and with pin 5 of the DIN plug to joined pin 3 of the XLR. As it happened, these adaptors were rarely used.
MIDI was initially viewed with suspicion and scepticism. Hence, in June of 1986, Jonathan Gibbs presented a ‘MIDI Seminar’ that would gently guide composers into this new world of music.
The first MIDI sequencer was the Yamaha QX1, which provided eight MIDI outputs. Its small display strained the eyes but it worked well and soon received updated ROMs. The big breakthrough, however, came in May of 1986, when the first Apple Macintosh computer was demonstrated, running Total Music software with a MIDI adaptor. This machine was slow, but two months later a Mac Plus model arrived on loan, complete with Mark of the Unicorn’s Professional Performer sequencing application and Professional Composer, a program for creating musical scores.
The diagram below shows how a simple MIDI system could be connected:-
As you can see, the output of the MIDI keyboard (usually a MIDI synthesiser in the early days) was plugged to the input of the MIDI adaptor (also known as an interface), whilst instruments and other MIDI devices were connected to its output. As shown in this example, a ‘daisy-chain’ was often used on the output of the interface, allowing up to three devices to be linked via their ‘Thru’ sockets. This meant that a typical adaptor with three output connectors could feed a total of nine devices.
Looking back at those early Macs, it seems impossible that anything could be done with one megabyte of RAM, a Motorola 68000 processor running at 10 MHz and a 700 K floppy disk drive. In fact, with their tiny screens and peculiar screens, these machines looked like toys. But behind the unprepossessing frontage there lurked some very impressive and efficient system software.
It wasn’t long before that ‘old serpent’ could be seen slithering through the Maida Vale undergrowth: once he got into the Radiophonic office, Brian soon indulged in an Apple or two! In August 1996, a Mac Plus with internal Micahdrive hard disk drive, large screen and screen adaptor had arrived. And by December, Brian was having fun experimenting with MacDraw, a simple drawing application.
In January 1987, Opcode’s DX/TX Librarian application arrived, allowing a composer to edit the sounds within a Yamaha synthesiser via a standard MIDI interface. The MacNifty SoundCap audio digitiser also appeared. Unfortunately, the Mac Plus had a newer serial port that lacked the +5 volt supply for the digitiser (this problem also occurred with the original Opcode MidiMac MIDI adaptor). For the SoundCap, the author created a 19-way ‘D’ plug from a cut-down 25-way plug and inserted this into the external floppy disk socket. A clever idea, but this didn’t work either!
In February, ten further Macs appeared in studios, offices and in the engineering workshop. The studio machines were to have internal Micahdrive drives, but the company went into liquidation. Therefore each machine had a 20 MB Apple HD-20 external drive, connected to the Mac via SCSI.
Each studio machine was also fitted with a Pro-Plus MIDI interface. Although satisfactory, this only worked via one of the Mac’s serial ports. The Workshop therefore investigated Rod Argent’s MIDI Communicator interface. Unfortunately, this objected to the presence of AppleTalk, Apple’s networking system. The company’s representative, a Mister Birch, often visited the Workshop and became known as The Man From Argent, although even he couldn’t do the impossible. By August of 1987, the MIDI Communicators had been abandoned and were replaced by another Opcode model.
In August, the Mac in Studio H acquired a large Radius screen and adaptor card. This useful monitor could view different material to that shown by the built-in screen on the computer. It could also be rotated into ‘portrait’ or ‘landscape’ positions, the software automatically switching itself accordingly.
At this stage, the Mac Operating System (Mac OS), then known simply as the System, didn’t include the modern desktop environment known as MultiFinder. This meant that only one application could be used at a time. To improve the situation, Peter Howell employed Switcher, a special utility that neatly swapped applications by sliding the entire screen image sideways. In October 1987, Mark Wilson tried to use Apple’s new MultiFinder instead of Switcher but this failed to work.
In the middle of 1988, six Mac II computers were installed in all studios. Soon after this, the MIDI connections in most studios were controlled by a Sycologic M16 MIDI matrix. This gave 16 MIDI inputs and outputs, connections being made or broken by pressing buttons on the control unit.
The Yamaha DX7 was one of the first synthesisers to come with MIDI. This instrument took the musical world by storm, mainly because its synthetic sounds contained intense but natural harmonics. It used a process known as frequency modulation (FM), in which two or more ‘virtual oscillators’, known as ‘operators’, interacted with each other. The technique, implemented entirely in the digital domain, created very pure sounds. Initially, some composers didn’t like the way the DX7 ‘stood out’ from other material, but it was simply showing up the inadequacies of older synthesisers.
More Yamaha FM synthesisers followed hot on the heels of the DX7. The first was the TX816, a large rack-mounted device with eight XLR outputs. In effect, it contained eight DX7 ‘boxes’, but without a musical keyboard. It had MIDI inputs for each module as well as a common input. In later years, the TX816 was sometimes connected to a studio’s main mixing system via a Yamaha KM802 submixer.
In August 1987, the Workshop received a TX81Z and DX7 II FD. The TX81Z was smaller than the TX816 and had a similar polyphonic ability, but without eight discrete outputs. The DX7 II FD was an improved DX7 that incorporated a floppy disk drive. By 1988, a similar device appeared in rack-mounted form as the TX802. This had eight outputs and accommodated microtonal tuning.
Other Yamaha products included the DX1 and KX88 MIDI keyboards, as well as the RX11 and RX5 drum machines. The MEP4 MIDI Processor was useful since it could modify a MIDI performance, often with dramatic results and could be employed to filtering out unwanted data from a MIDI circuit.
Roland products also arrived in large numbers, including the S50, a keyboard-based sampler that stored sounds on floppy disk. For ease of use, this was normally connected to a Rhombus Eizo or similar VDU. The Workshop also purchased at least one D50, a synthesiser that matched the S50. By February of 1988, most composers preferred using a single musical keyboard, putting older keyboard-based synths out of fashion and the newer rack-mounted S550 and D550 devices in favour.
The Roland TR707 and TR727 Rhythm Composers were low-cost drum machines, the latter having Latin American sounds instead of standard percussion instruments. Since the MIDI connections to these devices were often looped to other devices, the MIDI Out sockets were converted into MIDI Thru connections. The Roland SRV2000, an excellent electronic reverberation device was also used.
Yamaha and Roland didn’t have it all their way. For example, Alesis produced the low-cost Midiverb and Midifx processors whilst Valley People manufactured the superb Gatex noise gates. Alesis was also responsible for the Dominator, a wonderfully effective sound-enhancing device.
Another innovative device was the Twister, an eight-channel MIDI-controlled amplifier, in which the gain of each channel was set via a single MIDI input.
Inevitably, all this new equipment displaced many older devices, some of which went into storage. Victims included the PPG Wave 2.2, Godwin String Concert, Roland 100M, Synthex, Oscar and Wasp, in addition to an earlier Akai sampler and KR55 rhythm machine. Meanwhile, the Workshop donated the Yamaha CS40 and Oberheim OBX8 to a school, whilst the venerable CS80 was sent to London’s Guildhall School of Music. The once-cherished Fairlight CMI was apparently sold to a French customer for £7,500, although this deal eventually caused financial complications!
Other despatched items included the Great British Springs and the EMT Gold Foil reverberation unit. Although Mark Wilson tried to use a Wavemaker VCF with a Roland MIDI to CV Converter, these units, original intended for the Phoenix, never became popular and were soon sent into oblivion.
Macs were used for administration and engineering work from the very start. In February 1987, Brian Hodgson even installed Microsoft’s Word on the engineering computer, although the author was hesitant to use it. By March of the same year, an AppleTalk (now known as LocalTalk) network was in use between the offices and engineering areas, although each machine initially needed a power transformer to avoid joining the ‘technical’ and ‘general service’ earthing systems. Brian now eagerly used TopMail to send and receive messages over the system. In July, another e-mail application known as InterMail conflicted with Performer and had to be removed from all studio machines.
Originally, Ray Riley had used VIEWSHEET on a BBC Micro for the spares and equipment lists. Jonathan Gibbs soon arrived with all the software necessary to transfer this data to Excel on the Mac. Unfortunately, the 2 MB partitions on the Mac’s hard drive were too small to take the Excel application. Despite this, transfers were eventually made at 1200 bits/s over the serial ports. The data was then saved onto the main ‘network’ machine in the office and backed-up onto floppy disk.
The Mac didn’t entirely eclipse the BBC Micro, which continued to be used for special engineering work. In August 1986, it was possible to use a modem with the BBC machine, running at 300 baud, and painfully slow by modern standards. In fact, 1200 baud couldn’t be accommodated by the BBC’s Monarch PBX telephone system, to which the Workshop’s Diavox phone system was connected.
By October 1987, the author was using MacDraw for technical drawings and soon progressed to the more advanced MacDraft application. By July of 1988, MacWrite 5 had replaced the clumsy Word and Works software, even though MacWrite 5 could only open one document at a time.
In November 1987, the author made a circuit to connect the Mac’s loudspeaker output jack to the speech circuits of the Diavox telephone system. This allowed an application such as MacDialler to dial-up memorised phone numbers. One month later, this device, consisting of a transformer, resistor and capacitor in a box, had been installed in all areas. In 1991, the phone system was again replaced, this time by a Minc system installed by Britannic Telecom. Unfortunately, this was entirely digital and didn’t accept the analogue signals necessary for tone-dialling, so the boxes were then removed.
In May 1988, a device known as Thunderscan arrived, enabling an Apple ImageWriter printer to be used as an image scanner. A special sensor was inserted in place of the standard printer head, paper containing the original image was loaded into the printer and special software did the rest.
The Workshop always suffered with multitrack formats. Initially, the BBC used 8-track (one-inch) tapes. With the first Studio E, the Corporation advocated 16-track (two-inch) but then changed its mind and adopted 24-track (two-inch). So, the Workshop ended up with machines in all three formats.
Worse still, the main music studios used Dolby A noise reduction, whilst the Workshop used Bel, or nothing at all. As a consequence, multitrack machines were often moved around, accompanied by a host of adaptor cables and Dolby boxes. Inevitably, the line-up was wrong, which required all the tracks to be adjusted for both record and replay. Although a 24-track machine could play tapes created on a 16-track machine, the ‘wrong’ tracks had to be avoided and levels would always go awry.
During the summer of 1987, two 24-track machines arrived from Scotland and were installed in Studios C and E. In Studio C, the first eight channels of the wall-mounted 16-channel Bel noise reduction unit had been used for the 8-track machine, with the others connected to two A80 stereo machines. To accommodate 24 tracks, an 8-channel unit was added to the rack, whilst stereo Bel noise reduction units were installed in the trolleys of the stereo machines, as in other areas.
All multitrack machines had a combined remote control and autolocator. This gave the composer control over the tape transport and incorporated a digital clock that showed the elapsed time in hours, minute and seconds. Most important of all, it allowed the composer to return to the ‘zero’ point on the clock. The Survival autolocator was commonly used, although often modified. For example, most Studer A80 multitrack machines could be switched to run at 7½ or 15 inches per second (in/s or IPS). However, one particular A80 that only worked at 15 or 30 in/s couldn’t be easily modified. To get around this problem, Ray Riley had to install a divider circuit within the associated autolocator.
The Survival autolocator in Studio A was modified so that Elizabeth Parker could ‘prime’ any track on her 16-track A80 to go into record mode, a state known as ‘record ready’. Without this feature, she would have to walk over to her machine every time she wanted to record a new track. Meanwhile, Studio D was without a Survival autolocator, so Ray Riley created his own device by boxing-up a spare A80 ‘remote’ clock. Funnily enough, this worked better than a Survival autolocator since it didn’t ‘hunt’ back and forth around the ‘zero’ point. Indeed, by the summer of 1988, all the Survival autolocators had been modified to use the internal ‘zerolocator’ circuitry contained in each A80.
The Soundcraft SCM381-8 8-track didn’t use the Studer remote control system. However, by the autumn of 1987, the machine in Studio H had been modified to accept a ‘tape start’ from Syncwriter.
Of course, the days of conventional multitrack were numbered. In June of 1988, the Workshop evaluated an Akai 12-track recorder that used standard Video 8 cassettes. Although this didn’t catch on, it was a precursor to ADAT, the Alesis 8-track format. Further evidence of the demise of multitrack came from Richard Attree in Studio H. He asked for 20 dB gain switches to be installed in the desk’s monitor return circuits, allowing them to be used as extra channels for synthesisers.
By now, tape machines were mainly used for recording, not for sound manipulation. In the summer of 1987, two Lyrec TR55 stereo tape recorders were rescued from the stores at Broadcasting House. Nine months later the author despatched them to Redundant Plant: although these machines appeared rather smart, their speed stability left a lot to be desired. The three old Revox A700 quarter-inch recorders were also scrapped, along with the Studer A80 8-track machine from Studio D. In addition, the reduced use of film in television saw the demise of the four-plate Steenbeck film viewing machine, although the six-plate model was retained in the Workshop’s library area.
The use of Rotary-head Digital Audio Tape (R-DAT or DAT) was considered in the spring of 1987, although the first machines weren’t evaluated by the department until December of the same year. There was some hesitation in adopting this system, mainly because it was based on traditional video tape technology. Despite such reservations, the format was was highly reliable and a pair of domestic Sony DTC1000 machines were brought into service. At least one of these was modified for remote control of both the pause function and triggering of the Automatic Music Search (AMS) system.
By March of 1988, discussions were underway as to whether to Workshop should use quarter-inch tape machines with timecode or DAT. The Sony PCM2500 DAT machines then arrived, each made up of two boxes: the transport mechanism itself and an extra unit that provided digital audio connections via an AES/EBU interface, Sony/Philips Digital Interface (S/PDIF) and Sony Interface (SDIF2).
These machines introduced a problem. The DTC1000 used input and output signals at -10 dB, with the record control normally set at ‘3’, giving 20 dB of headroom. The PCM2500, on the other hand, operated at 0 dB, with the record control at ‘8’ and 30 dB of headroom. This disparity remained until 1991, when the DTC1000 models were fitted with an amplifier card, giving a standard output level. A 2.2 kohm resistor was wired across the input control to give a lineup at ‘4’ with 12 dB ‘in hand’.
Late in 1988, the new ‘circular’ Studios B and F received two PCM2500 DAT machines with ‘wired’ remote controls whilst the two DTC1000 machines were made available to ‘roam’ the other studios.
The relocation of Studio E caused problems because the area contained all the Workshop’s audio tielines. Therefore, a new ‘central bay’ had to be created in the corridor between Studios C and D. The bay’s jackfield was connected to the original cables, as relocated or extended within the roof space. Each studio had ten balanced circuits, the last three of which could be used for an echo plate. Twenty further circuits connected the bay to the main Maida Vale ‘plate room’. Even with these links, it wasn’t possible to connect anything more than an 8-track machine to any other studio in the building.
Ray Riley also provided the bay with a test oscillator that generated 1 kHz tone at the standard level of 0 dB. In addition, the jackfield had access to numerous audio transformers (in the BBC, erroneously called ‘repeating coils’ or ‘rep coils’) as well as several ‘phase reversing’ pairs of jack sockets.
The bay also contained a new technical earth busbar (a thick strip of copper with holes tapped for 0 BA bolts), wired to the special incoming earth circuit. Each studio had a separate earth wire, allowing the earthing integrity of the studio to be checked by ‘lifting’ the wire from the bus. Although standard 100 amp earthing wire was originally used, later installations employed a multi-strand ‘PSF’ cable. The latter ensured a low impedance earth, even for radio frequency (RF) signals.
In addition, insulation displacement connection (IDC) ‘blocks’ were installed for non-critical circuits. Blocks of this type allowed each wire from the studio to be ‘broken away’ from the associated bay wiring by the insertion of a special tool. A series of such blocks, each connected to a PSN/20 cable from each studio, were used for special circuits. These included the blue light, red light and cue light circuits, also the cue and studio loudspeaker (SLS) audio feeds. A 'virtual earth' buffer amplifier combined the audio circuits from all the studios. This allowed any studio to feed signals to the Piano Room, then an acoustic recording area fitted out with portable talkback boxes and headphone units.
Further blocks, connected to a PSN/40 cable from each studio, conveyed the stereo ringmain. This allowed composers to listen to the BBC’s radio networks and the outputs of the main Maida Vale studios. Originally, the ringmain was in mono, wired via PSN/20 cable, but by August 1987 it had been upgraded to stereo. Each office then had a stereo ringmain box that contained a rotary selector switch and two mono jack sockets, one for each channel of the stereo signal. In later years, a ‘Radio Frequency (RF) Ringmain’ replaced these connections. This innovative new system superimposed the BBC’s internal circuits onto ‘carriers’ via a normal VHF FM radio aerial feed. All the studios then had FM tuners, allowing composers to listen to commercial radio networks as well as BBC programmes.
The bay also included a radio frequency (RF) amplifier. This boosted the incoming UHF TV and VHF FM radio aerial feed and distributed it to the studios. The RF outlet in each studio was isolated using low-value capacitors in both the signal and screen circuits, so avoiding any possible earthing problems. Each studio also had a video tieline that appeared on the video patch panel on the bay. In theory, this allowed cameras and monitors to be used for recording sessions but was hardly ever used.
Dave Young’s intercom system was long gone, but the modern equivalent didn’t have its ‘listen-in’ feature. Instead, it made a ‘call’ sound, disturbing the composers and their recording sessions. In some areas the author replaced the three phones by a single instrument and a switch box. This worked, and although such ‘tampering’ broke the Post Office rules, no one seemed to notice. Despite these improvements, the line rentals for the phone circuits were now costing almost £4,000 a year. In a very short time, a new £8,000 Diavox system from Thorn-Ericsson was installed adjacent to the bay.
Following BBC traditions, the Workshop fitted most devices with ‘professional’ XLR connectors. This usually involved installing audio transformers to ‘balance’ the circuits. Although such transformers avoided ‘hum loops’ and other earthing problems, they also compromised the audio quality. Initial experiments indicated that removing these transformers didn’t have any adverse effects. Following these tests, audio transformers were extracted from nearly all the devices in the Workshop.
The Workshop also suffered from complications regarding the wiring of XLR connectors. Britain and most European countries used a convention that defined pin 1 as the screen, pin 2 as the line (L) or ‘hot’ wire, and pin 3 as the return (R) or ‘cold’ wire. Unfortunately, the American standard, as used in department’s Soundcraft desks, swapped pins 2 and 3, using pin 3 for the ‘hot’ circuit.
For unbalanced circuits, reversals in the wiring of XLRs wasn’t important, although audible effects could be heard on very low-frequency or impulsive sounds. However, when it came to unbalanced signals the connection could be lost altogether. Since several mixing desks and Bel units were already wired with pin 3 as ‘hot’, the Workshop decided to adopt the American standard.
Headphone connections also gave problems. Traditionally, BBC mono headphones came with a Post Office quarter-inch tip-ring-sleeve (TRS) jack plug. This PO316 plug had a smaller tip than the modern stereo version and mated with the sockets on a standard BBC jackfield. In addition, to minimise circuit loading, the headphone inserts were wired in series across the plug’s ring and tip.
Later stereo headphone sockets had the left-hand channel wired to the tip, a ‘phase-reversed’ right-hand connection on the ring and a common earth on the sleeve. Stereo headphones were wired to match, allowing mono and stereo headphones to be used anywhere. Unfortunately, everyone else in the world was using standard stereo jacks with simple left and right-hand connections.
This meant that all BBC equipment had to be modified to work with this system. The author, whilst at Radio Projects during 1990, tried to dissuade the Corporation from spending thousands of pounds of licence-payers money on such outdated practices. Ray Riley sensibly suggested that standard headphone plugs be used and that an adaptor be introduced to connect into studio jackfields.
By June of 1986, the Workshop decided to adopt commercial headphone standards. Existing PO316 sockets on most tape machines were perfectly happy to receive a standard stereo plug. Fortunately, the phase reversal wasn’t audible on headphones and further modifications weren't required. Some modern plastic sockets had to be replaced, or modified with the aid of a quarter-inch drill!
The Workshop also standardised its loudspeakers. The old BBC LS5/5 units in the Piano Room and the small LS3/5A speakers (with BBC AM8/12 amplifiers) in the temporary Studio F, were replaced by Roger’s LS5/8 units. Each loudspeaker had a single stereo Quad amplifier that the BBC had fitted with an active ‘crossover’ circuit. The amplifier then gave two outputs, one for the ‘woofer’ and one for the ‘tweeter’. There was also a need for alternative loudspeakers that would simulate a domestic hi-fi system. The Fostex model 6301B was used, a small unit with inbuilt amplifier and volume control. Some composers used these in stereo pairs, others singly to check mono compatibility.
By the end of 1987, the department experimented with removing the input transformers on the LS5/8 loudspeaker amplifiers. This gave a 3 dB improvement in performance at 15 kHz, which was quite audible. Also, in Studios A, B, C and D, the loudspeaker amplifiers were plugged directly into the consoles, rather than via the original 4-pole F＆E (EP) pattress sockets. In Studio E, however, the EP connectors were replaced by XLRs. All the amplifiers were then fitted with short ‘gender bender’ cables that converted the audio inputs to the commercial ‘pins point in the direction of signal’ standard.
In the early days, the Workshop didn’t have any safety standards: most devices were powered from unreliable distribution boards with 3-pin 13 amp ‘D ＆S’ sockets. These accepted a matching plug whose live pin actually contained a fuse. Such non-standard connectors allowed the ‘technical’ equipment to be segregated from those devices fed from the ‘general service’ supply. In addition, the staff were unlikely to steal these peculiar plugs for use at home! At Television Centre they adopted an alternative form of 13 amp plug that had the normal square pins but all rotated by ninety degrees.
By 1986, the Workshop was a safe place to work, with all areas fitted with standard 13 amp sockets and high-quality mains wiring. In fact, the BBC had a ‘rule’ that all equipment should have a three-core mains lead. This ensured that there was a safety earth for conducting away fault currents, preventing such currents from flowing through the staff! However, many newer devices arriving at a Workshop came with two-core cable, supposedly conforming to ‘double-insulation’ standards.
Fitting a three-core lead to every double-insulated device wasn’t entirely desirable. Therefore, each piece of equipment was assessed on its own merits. Usually the insulation resistance was satisfactory and the construction sufficiently robust to prevent the mains supply getting into contact with any external metalwork. Unfortunately, some devices had capacitors between the chassis and live or neutral wire, presumably in an attempt to reduce radio frequency (RF) interference. Removing these components didn’t cause any problems and eliminated this potential safety hazard.
The mains cable at each device was either captive or connected via a detachable ‘inlet’ connector. The BBC’s usual choice was the XLR LNE mains connector: although similar in appearance to an audio XLR connector, this had small gold-plated pins rated at 3 amps. Perhaps because it was smaller, a socket was always fitted to the equipment itself. This meant that it was theoretically possible to make contact with the mains on the plug pins, although these were reasonably shrouded.
Ideally, the earth pin of a mains connector should be the first to make contact when a plug’s inserted and the last to disengage when it’s removed. In practice, this didn’t always happen. When Malcolm Clarke went to unplug a pair of ‘inline’ XLR LNE connectors he received a slight shock! Following this, the LNE connector went out of favour. In several instances the latch button on the plug was cut away so that it couldn’t be removed from the device, effectively creating a captive mains cable.
Many commercial products used a three-pole IEC inlet, rated at 6 to 10 amps, although some double-insulated equipment had a 2-pole version. Portable devices often came with a much smaller connector, also in 2-pole form. All these connectors were electrically satisfactory, although the 2-pole versions weren’t ideal since they allowed reversal of the live and neutral wires. Unfortunately, some variations also had a nasty habit of falling out at the least provocation!
As equipment proliferated, it became clear that the standard 13 amp connector was too cumbersome. Instead, the Workshop decided to adopt IEC outlets. These were normally in the form of a ‘strip’ of five connectors with an attached cable and 13 amp plug or a standard IEC inlet connector. Each strip contained a fuse, usually rated at 5 amps. The first of these strips appeared in Studios B and E in February 1987, but by the end of the year they were installed in all areas.
In the summer of 1987 Studio B was updated to the Studio H ‘standard’, complete with new effects rack and updated 16-track wiring. Room 11 (Studio H) was dismantled and plans were afoot for updating the existing engineers area in Room 12 whilst adopting Room 11 as an engineering office.
In the autumn, Studio B had a problem with background ‘buzz’. In February, Technical Investigations discovered a neutral-earth ‘short’ in the mains wiring at the other end of the building. This was associated with the alternating current (AC) emergency lights in Maida Vale 1, the studio used by the BBC Symphony Orchestra. Incredibly, this fault created an earth current of around four amps!
Further changes were in progress: new video switches were fitted to the monitors used for sampling machines and the old ‘services trolley’ for acoustic recordings in the Piano Room was dismantled.
©Ray White 2001, 2004.