SOUND DUBBING: The history of Video Sound Post Production for television in the UK: Part One – ‘Analogue’.

This is a ‘personal look’ based on my experience of the early days of TV sound dubbing and relates to working for ITV’s London Weekend TV, along with my interest in the methods employed by the other major broadcasters in London; the BBC and Thames TV. So excuse me if I’m initially concentrating on them. I would however, be interested to hear of practices at the other ITV companies and in the regional BBC studios in the years before ‘digital’ changed everything.

THE CONTENTS:

1| THE ARRIVAL OF THE PROFESSIONAL VIDEO TAPE RECORDER
2| PROBLEMS FOR SOUND THAT VIDEO NEEDED TO OVERCOME
3| THE EXISTING FILM SOUND DUBBING TECHNIQUE
4| AN EXAMPLE OF HOW DIFFICULT IT WAS TO DUB ADDITIONAL SOUND ON A VIDEO
5| 1970| THE THAMES TELEVISION MEDWAY SYSTEM
6| 1970| HOW SOME OF THE OTHERS WERE DUBBING
7| 1972| THE ARRIVAL OF TIMECODE AUDIO SYNCHRONISERS
8| 1974| THE BBC SYPHER SYSTEM
9| 1980| THE AUDIO KINETICS Q-LOCK
10| FADER AUTOMATION ARRIVES
11| MORE DUBBING SUITES and MORE TAPE TRACKS
12| WHAT WERE WE DUBBING?
13| THE INDEPENDENT SOUND DUBBING FACILITIES

1| THE ARRIVAL OF THE PROFESSIONAL VIDEO TAPE RECORDER

Prior to the arrival of the video tape recorder, television companies had to record the programmes they made ‘live’ with video cameras, onto film if they wished to be able to ‘time-shift’ them in anyway. This was a particular problem for the US Networks with the time-zones.
The first successful commercial video recorder for broadcast use came from AMPEX and was the VR-1000 Quadruplex (Quad) VTR, which after a number of earlier versions commenced production in 1956. It used 2-inch tape and overcame the high frequencies required by using a fairly low tape speed of just over 15ips, but byrapidly spinning four heads in a common headstack, sequentially transversing across the tape, it vastly increased the apparent tape speed to enable the frequencies required for video.

Wikipedia:
“The advertised price for the Ampex Videotape Recorder in late 1956 was $45,000 (equivalent to $504,000 in 2023)”.

It was followed by models from RCA and TV companies bought Quad VTR’s in ever increasing quantities in the following years. These initially were used to delay transmission of programmes that were still being made ‘as live’ but eventually the need to edit tapes had to be tackled.

‘QUAD HISTORY’ – QUAD VIDEOTAPE GROUP:
“The September 29, 1958 issue of Broadcasting Magazine explains more, in a feature about ad agencies moving live commercials to tape.The first edited videotape programme is reputed to be a drama ‘The Red Mill’ at CBS in April 1958 although it appears this was to overcome a strike by technicians:
“BBDO ad agency executive Al Cantwell told Broadcasting that the production was much too complicated to be handled live by network executives who suddenly found themselves manning cameras, booms and lights.  So it was taped in pieces and then edited together, a “thing never done before or since,” according to Cantwell.
Each point where sequences were spliced together had seven seconds of black. The alternate to the black was “roll over” on home screens. Although ways were found to shorten the black somewhat, it was decided to eliminate the black and take the roll over.”
http://www.quadvideotapegroup.com/category/quad-history/

Initially the technique that was empliyed to cut the 2-inch Quad videotapes was a pretty major surgical operation!
Editing requiring finding the edit point on the 2-inch tape approximately by playing the tape to the desired point, and then removing the roughly marked up 2″ tape from the spinning ‘Quadruplex’ head-stack and loading it into an editing block. Editing video was much harder than audio as it was vital to maintain the continuity of the ‘control track’ that the VTR relied on for picture stability.
Commercial ‘editing blocks’ became available such as the Smith Splicing block shown below.
The chosen area was then painted with a solution of suspended iron filings called EdiVue directly onto the tape, and as it was necessary to keep the control-track pulses evenly spaced and intact, the editor waited a few moments until the solution dried and revealed these pulses.
The VT editor then made a butt splice on the tape clamped in an editing block through a ‘surgical technique’ using an editing microscope device attached to a guillotine.

Edits could be done within a programme but it wasn’t normally encouraged at all, firstly because the soundtrack was half a second in front of the vision, so a cut made there would lose a bit of the incoming sound. The ‘hole’ needed filling in, which editors could do by relaying it with the sound copied off a 1/4″ machine, however that took time and despite the efforts of the VT Editor, the edit could still make the picture ‘frame roll’ if not perfect …. so it was best to do it after a fade to black or before the start of the next part.
2″ videotape was also extremely expensive…so the cost would also be the third reason not to edit.

This YouTube video explains the complexities of the 2-inch Quad VT in some detail, including the amount of ‘tweaking’ required just to get it to record well……VT line-up used to take ages I remember!

Use the ‘back’ button to return to the text here:

The 2-inch VTR’s survived in operation from 1956 to well into the 1980’s when the 1-inch machines, usually the C-Format, took over in most countries.

For sometime sound on the 2-inch VTR couldn’t be post-produced without copying it to another machine, thus degrading the programme in both sound and vision by another generation, which in the analogue era was unfortunate and so by the end of the ’60s the 2-inch QUAD VTR’s had evolved an editing system using 2 machines, one playing back and the other ‘assemble editing’ from it, initially via somewhat crude manual ‘synching’, so that sections from the Master tape on the ‘Play’ machine were re-recorded in sequence to make a finished version. Both pictures and sound were then 2nd generation. This two-machine editing was automated with the EECO 900 system and more commonly found was the 2nd generation Ampex VR2000, like the one in the last YouTube video which brought ‘Editec’, an automated way of punching into an existing tape to ‘assemble edit’.

For much more information on the 2-inch Quad VTR, other VT equipment and its use in the BBC, visit the ‘VT Old Boys website’:
https://www.vtoldboys.com/donk01.htm

****************************************************************************************

2| PROBLEMS FOR SOUND THAT VIDEO NEEDED TO OVERCOME

With only one audio track, and a poor quality ‘cue track’, 2-inch machines weren’t capable of any real additional sound work on their own and there was no ‘still frame’ ability so one of the difficulties faced in using video was that it could only be viewed when in ‘play’. When you stopped playing the video, and the picture disappeared.
Likewise, you could only record onto even 1/4-inch audio tape when it was running at the correct speed, so putting say the previously mentioned ‘explosion’ onto video required it also the be already up to correct run speed, resulting in difficulties in placing sound accurately.
Multiple audio ‘tracks’ were not yet available in the video world and although multitrack tape recorders were arriving during the late ’60s there was no way of synchronising them with the video tape machine yet.
Because constant use of a 2-inch Quad VTR was so expensive, and produced wear on the master tape, a cheap lower quality VT machine was needed for rehearsing during sound work.

****************************************************************************************

3| THE EXISTING FILM SOUND DUBBING TECHNIQUE

What did work however was the existing film dubbing technique, which had been developing since the first days of the ‘talkies’, and by the 1960s an audio recorder, usually a mono Nagra using the ‘pilotton’ system, recorded the ‘production audio’ at the sametime as recording a 50-Hz (in Europe) or 60-Hz (in the US) pulse produced by an accurate internal crystal onto the tape. This pulse used a record/playback head aligned 90-degrees out of phase, and the pulse was therefore cancelled out on the main audio track.
During subsequent playbacks, the Nagra was connected to a ‘Sync Lock Oscillator’ resolver unit and the recorded 50/60Hz tone was referenced to a accurate ‘master 50/60-Hz’ signal (fed to its input) and it then drove the Nagra’s servo motor system to remove the original slight speed fluctuations inherent in any tape recording . The sound was therefore totally free of the ‘drift’ that occurred on any normal tape recorder, even the best professional ones.
By transferring off a Nagra/SLO unit onto sprocketed 16mm or 35mm ‘mag tape’ controlled by yet another servo system, usually using a ‘Selsyn motor’, allowed multiple of these sprocketed ‘Mag Recorders’ to be run in complete synch, to provide multiple audio sources for sound dubbing.

The various different bits of dialogue along with the many added sound effects and the music would be split to separate pieces of sound mag film and a series of ‘pre-mixes’ could be made to reduce those tracks down, allowing the film sound dubbing mixer to have an easily controllable number of inputs to his mixer.
Film was wonderfully ‘manuable’ because you could view the individual frames easily, so for instance, adding an explosion to a frame of 16mm or 35mm film, just required the sound editor to stop his ‘Steenbeck’ editing table at the required picture frame and slip the sound of the explosion on a suitable bit of ‘mag track’ into sync with it. He could now ‘wind on’ keeping the picture and its new added sound effect in perfect sync.
Dubbing a film for television was capable of producing complicate sound tracks assembled in a methodical manner, even if it took considerable copying of the analogue audio and an amount of ‘pre-mixing’ of ‘stems’ along the way.
Even after video sound dubbing got well underway, some director’s still preferred to copy to film so as to use the ‘manuable’ film track laying and dubbing for their video programmes.

****************************************************************************************

4| AN EXAMPLE OF HOW DIFFICULT IT WAS TO DUB ADDITIONAL SOUND ON A VIDEO

In 1969/1970 London Weekend TV were making a wartime drama, in colour for the first time, called ‘Manhunt’. Although shot in the studio ‘as live’, much of this series was also made using the heavy EMI 2001 cameras and 2-inch Quad recorders ‘on location’, using an ‘outside broadcast’ unit. All dialogue was record in either the studio sessions or the location ‘outside broadcast’. The ‘location’ parts were then usually played through the studio control rooms to be assembled onto the master tape in the correct programme order along with the studio sequences.

One episode in ‘Manhunt’ ‘Intent to Steal’ transmitted 29th May 1970, was about an attempt by the French Resistance to steal part of a secret rocket motor from a factory. A large section of the script in the second and third parts had almost no dialogue inside the factory and the sequence, taking place during a daytime raid by RAF Mosquitos, was a mixture of video and film and comprising recorded gunshots, mainly from the German’s Schemeizer and the Resistance’s Thompson machine-guns using ‘blanks’, along with ‘flash-bang’ dummy explosions which although very loud, weren’t effectively the correct sounds that the programme sound track required. Most of the film sequences were for some reason mute.

Paul Faraday, a ‘grams operator’ at LWT was given the task by Sound Supervisor Mike Fairman, who was mixing the programme, to assemble a working sound-track for it.
Paul sat in the sound prep room at the LWT Wembley studios and timed every explosion whilst watching a playback from the master 2-inch video tape. He then very meticously assembled suitable gunshots and explosions on a number of reels of 1/4-inch mono tape, using the TR90 and BTR1 tape decks. Each sound effect was separated by a piece of white ‘leader tape’ to achieve the necessary spacings, as determined by his ‘stop-watch’!
Because of the lack of any ‘Nagra type’ servo motor control, the TR90 and BTR1 tape decks however could only maintain a close enough ‘sync’ for very short periods, perhaps 4 minutes at best. Thus numerous stops and restarts were required to play out a section with close enough picture sync.

Paul used a variety of visual clues to trigger each section, and he maintained that he could hit ‘a visual cue’, of say when a certain actor’s arm touched the edge of frame, to trigger those tapes within perhaps 1/100th second. I never doubted that he was right about his ability to do that.

Two 2-inch VTR’s were in use, with the edited master tape playing off, whilst a second VT recorded the original sound plus Paul and Malcolm’s added new effects. The ‘commercial breaks’ in an ITV programme dictated the sections available before stopping the VTR’s, as stopping and re-synching them was still a slow operation. This required Paul and his helper Malcolm Treen, hitting those ‘cues’ consistently during each whole part of the programme, whilst Mike Fairman was at the mixer adjusting the levels to suit the programme as it was being recorded back to the second VTR.

This was the first time I saw a comprehensively ‘sound dubbed’ programme being made ‘on video’ and was a great feat of manual ‘tape operating’, that illustrated the reason why a proper ‘sound dubbing’ system for video was required to bring it to the level easily achieved in film!

****************************************************************************************

5| 1970| THE THAMES TELEVISION MEDWAY SYSTEM

Whilst we at LWT were struggling to start a method of doing post-production sound on our programmes, over at Thames TV in Teddington the were developing the first use of a servo-locking video dubbing system using a ‘digital code’ in the UK.

NEWS RELEASE FROM THAMES TV -1970 IBC EXHIBITION:
Digital Servo System for the Synchronization of Non-sprocketed Tape Recorders:
“This system is being described fully at IBC70. It has been designed to overcome a number of defects encountered when using currently available equipment by using helical scan VTR machines wherever possible, and by using multi-track audio recorders to provide greater operational flexibility and by providing a synchronizing system to allow any of the three types of recorder (that is quadruplex VTR, helical scan VTR, and multi-track audio recorder) to be locked to any other. The original programme sound is copied on to one track of the multi-track audio recorder and the vision on to a helical scan machine. These are operated in synchronism whilst the new sound track is prepared using multi-track techniques. To eliminate’ wow’ problems the helical scan VTR  is used as the controlled element in the synchronizing system. The new sound track is transferred from the audio tape back to the quadruplex VTR by use of the same synchronizing system, using the VTR as the controlled element to eliminate the power amplifier which would be needed to drive the audio machine. The synchronizing system records, using normal bias, a numerical code on the cue track of the quadruplex VTR, the sound track of the  helical scan VTR and one track of the multi-track audio recorder, whilst copying the edited programme material. The present code is a ten-bit, bipolar return-to-zero code counting television fields in pure binary. A space is left between each group of ten bits for synchronizing the decoding circuits. The ten-bit code repeats at roughly 20-second intervals, implying that the synchronizing system can tolerate an error of a tape length equivalent to ± 10 seconds. The system was designed with this fairly limited range to enable the code to lie entirely within audio band limits. The company is investigating, as mentioned elsewhere, a standard digital code to be used both for VTR editing and for the location, process control and automated editing of film sequences.”

 Thames introduced this digital code synchronisation method for video sound dubbing that used a cheap helical-scan VTR and a multitrack audio machine during the ‘track laying’ and also allowed a synchronised ‘layback’ to the master 2-inch Quad VTR.
Tom Sloley, a Thames Projects engineer, who had worked at tape machine manufacturer Leevers-Rich, was responsible for designing a synchroniser using a digital 10-bit ‘address code’ that was simultaneously recorded on the ‘audio cue track’ of the source 2″ Quad VTR master tape, a ‘lo-band’ helical-scan VTR ‘viewing’ copy, and onto a 6-track Telefunken audio multitrack. The original mono soundtrack from the 2-inch VTR was also transferred to the audio multitrack and this was ‘enhanced’ with music and effects before being mixed down for the ‘layback’, when it was relayed back onto the master 2-inch Quad master audio track, replacing the original. 
This early 10-bit digital code though repeated often and therefore required the operators to park the machines together within a 30-second window, within which it could servo control the multi-track into ‘lock’ with either of the video decks.

MIKE PONTIN -THAMES TV SOUND SUPERVISOR:
“ABC TV / Thames TV had a strong Programme Sales department. They were always keen to sell our considerable drama output to overseas buyers. It was thought that if we could supply M&E tracks this would help sales.
By the early seventies Thames engineering department had designed a basic timecode (Teddington Lock) and a synchronising device. This offered the possibility to transfer much of the gram op function to post-production. A system for post-production dubbing was developed with the hope of producing more sophisticated audio for the UK market and M&E tracks for overseas sales. It was christened MEDWAY (Music, Effects, Dialogue).
The system comprised:
An Ikegami ¾” helical scan video machine.
A Telefunken 6 track 1” audio recorder
An in-house time code generator which produced a repeating code every 20 seconds.
An in-house synchroniser controlling the Telefunken capstan motor.
Remote tape counter displays for Ikegami and Telefunken placed under the picture monitor.
Remote tape deck controls for both machines mounted in a console.
A 12-channel ‘free grouping’ Neve mixing console. (This was one of the ‘add-on’ sub-mixers from the main studio Neve consoles.)
A standard gram op’s complement of tape decks etc.

VTR LAY OFF: 2-inch VTR, Helical-scan VTR and audio multitrack

“This was an operation carried out by VTR dept. The edited master video tape was played back and video was transferred to the Ikegami. VTR audio was routed to the Telefunken. During the transfer all three machines recorded the digital address code.”

AUDIO TRACK LAYING: Helical-scan VTR and audio multitrack

“This was the first process in the dubbing suite. The sound supervisor responsible for the studio sound took charge in the dubbing suite and the director normally attended also. Usually they would elect to view a substantial section and discuss what needed to be done. The Sound Supervisor would then decide how he was going to deploy tracks etc.
Whilst track laying, the designated mixing console faders were set at unity and monitored off the Telefunken sync head.
The machine counters were manually zeroed at minus 30secs on the VTR clock. From this point the machines would lock and tracklaying could commence. To roll back, the video machine was parked at 30 secs before the point of interest as seen on the helical scan video and then the audio machine was rolled back so that the tape counters matched. Both machines were started together on a common start button and normally locked quite quickly. The supervisor was responsible for dropping in and out of record using remote controls whilst the gram op concentrated on “cueing in” Fx or music correctly. Music and effects were laid at an appropriate level with monitor faders at unity gain in order to minimise the adjustment required on mix down. The sound supervisor noted all additions on a track sheet. He also noted necessary manipulations of the dialogue track.

The following track lay out was used:
Track 1 Dialogue (play only)
Track 2 Music
Track 3 Fx
Track 4 Mix
Track 5 Misc  (very low level timecode break-through but useful in practice).
Track 6 Timecode

AUDIO MIX DOWN: Helical-scan VTR and audio multitrack

“When track laying was complete the tapes were rolled back to the VT clock and the Telefunken heads were switched to replay. The supervisor ran the tapes to sync and dropped into record on the mix track. The complexity of mixing varied enormously. Sometimes it was a very simple one pass affair and sometimes quite short sections would be performed by frequently rolling back and dropping in and out of record.”

VTR LAYBACK: 2-inch VTR and audio multitrack

“The 1” audio tape was delivered to the VTR department and the mix track was transferred using the synchroniser with an offset to the master VT tape. There was no sound department involvement in this process and there was regular concern about the quality of the final track which was fourth generation at best. A great deal of care needed to be taken over machine line up and batch testing of tapes. No noise reduction processes were used but the Telefunken six track noise and track separation specs were quite good.”

Music & Effects Tracks for Programme Sales:

“Very few M&E tracks were produced and of course they were woefully short of the accepted international standard. At best they were a starting point for some unfortunate overseas film dubbing outfit. The material was provided by transferring an M&E mix from the Telefunken to a  1/4” Nagra.
However the facility was very popular with directors and soon most of Teddington’s drama and LE output was processed through the dubbing suite.” [2]

I’ve no doubt such a system would be seized upon by the Drama and LE Depts, as that’s where the need to ‘enhance’ the studio sound was most acute.

****************************************************************************************

6| 1970| HOW SOME OF THE OTHERS WERE DUBBING

Prior to being able to view programmes on cheap helical-scan video decks, everyone was using expensive 2-inch QUAD VT time and adding new soundtracks to them in various differing ways:

Yorkshire Television – VTR Cue Track Dubbing:

“The method which we have adopted at present simplifies the dubbing procedure and eliminates loss of sync risks completely.
The cue track performance of the RCA TR 70B VTR’s which are in use in Yorkshire Television is sufficiently good to permit of its use as a second sound track.
Minor modifications only were required to enable high quality transfer of audio signals from the main to the audio track. Thus, when we wish to enhance the sound track of a programme recorded on video tape we need only replay one VTR machine, route the second channel via a sound dubbing theatre where the extra material is added, and simultaneously record on the same machine’s cue track. In order to avoid unnecessary transfers the programme is then transmitted using the cue track rather than further rerecording back to the main audio track.
In order to further simplify the operation and reduce the number of occasions when the 2 in VTR must be played, ½ in helical scan video tape copy is made immediately the edited video master is produced.”

I imagine the 2-inch Quad VTR’s ‘Cue track’, which was less than half the width of the normal audio track and near the noisy ‘control track’ and gave very inconsistant results. We tried it with RCA TR70s and the audio usually had poor HF responses when played back on another machine from the original recorder.

BBC before ‘SYPHER’:

In 1977 John Eden-Eadon wrote a summary of ‘SYPHER’ in BBC Engineering, and he detailed how the BBC worked prior to the new helical-scan/multitrack system:

So 10 hours of 2-inch VTR time were estimated to dub a 50-minute ‘Onedin Line’ drama episode, with all the effects and music being played in and levels balanced ‘as live’.

London Weekend TV – ‘Tim and Syncs’:

For sometime we used record the dub onto a mono Nagra III with 50-Hz pulse derived from the Main ‘Station Sync Pulse Generator’. The technique required the 2-inch VTR to be playing back through the dubbing desk, as we added sound effects and music ‘as live’. The finished tape was layed back to the Master VTR using a recorded ”spoken count’ or ‘sync-claps’ that the VTR operator had added to prior to the clock on the VTR and was copied to the Nagra.
Our first accurate dubbing system was created by Maintenance Engineer Dave Lancaster, who took a Studer A80 8-Track which had been lanquishing unused, and ‘Nagra Sync’ modified so that its motor was servo controlled using the Nagra ‘Sync Lock Oscillator’ unit, and during the ‘layoff’ we recorded the 50-Hz reference from the main ‘SPG’ onto Track 8.
Just like on the Nagra, this provided the necessary accuracy of speed required, however correct ‘positional information’ was necessary so we still used the method already mentioned from earlier days when the 2-inch VTR operator recorded a ‘sync-clap’ (or just counted!) which now was put onto the cue-track of the Main VTR, the helical scan VTR and multitrack A80 during the ‘line-up clock’ period on the tapes before the programme commenced. The appropriate combination of VTR and audio deck were brought into sync totally ‘manually’ though, as explained later.

The next ‘big step’ was to replace this with a continuous recording of ‘The Speaking Clock’, which could be found by dialling it up on any phone! We had it wired to our audio patchbays:

 THE UK ‘SPEAKING CLOCK’:
“At set intervals of ten seconds a voice announces, for example “At the third stroke, the time will be twelve forty-six and ten seconds……”, with three beeps following.”

As it was wasteful to use yet another audio track on the 8-Track just for recording the continuous ‘Speaking Clock’, I suggested it was mixed with the 50-Hz pulse, via a high-pass filter…..and so it was named ‘Tim and Syncs’, and was with us for sometime before Maglink arrived.
The audio operation with this was truly ‘manual syncing’ and required the operator listening on headphones to chase the appropriate video deck’s recorded ‘Tim and Syncs’, using a ‘black box’ vary-speed unit; toggling faster or slower as required. Most operators could soon hit sync in a very few seconds, so the Sound Mixer didn’t have to wait too long each time.
We often produced 1/4-inch audio mixdown’s of shows and concerts and locked them back to the edited pictures using just 50-Hz for stability, and in 1976 Gram Op Graham Hix used ‘Tim and Syncs’ to lock a stereo mixdown of a Jethro Tull Concert with a 2-inch VT replay to send to Capital Radio as UK Commercial Radio’s first ‘stereo simulcast’, with pictures being shown on ITV.

****************************************************************************************

7| 1972| THE ARRIVAL OF TIMECODE AUDIO SYNCHRONISERS

The Thames ‘Medway’ system used their own 10-bit digital code, but by end the ’60s the video world was discovering ‘Timecode’.

“The development of SMPTE Time Code goes back to 1967, when, to facilitate electronic videotape editing, the California-based company EECO developed a method of encoding video frames with an absolute time and frame number, based on a method used by NASA to synchronize telemetry tapes from its Apollo missions.”
EECO, Inc., The Time Code Book, 1983

“US Society of Motion Picture and Television Engineers ST-12 was proposed by SMPTE in 1970 as a universal labeling system for frames on video tape. In this standard, each frame would be counted at a specific rate: frames would make up seconds, seconds would make up minutes, and minutes would make up hours. This, of course, would then be displayed to look like a clock: HH:MM:SS:FF.”

Video editing systems were soon develo[ped that used SMPTE Timecode to accurately address every video frame, and of course provide the perfect positional information that was so needed.

The audio world also caught on and in 1972 EECO, 3M and others soon were selling SMPTE Timecode synchronisers.

Ad in ‘Recording Engineer and Producer’ magazine for 3M 79 Series in August 1972.

My own company LWT decided to use the MagLink synchroniser.

Maglink had ‘enough buttons’ to provide control and show the positional information for up to 6 ‘slaves’ and from about 1974 LWT used it for dubbing with an 8-Track Studer and an Ikegami Helical-scan VTR. It worked well enough for LWT to use it for a number of years.
Maglink didn’t use the recognised ‘SMPTE Timecode’ though and had its own proprietry version of it, so a converter unit translated the VTR ‘Timecode’ into Maglink code when working with the 2-inch VTRs.

Pictured in 1978/9, LWT’s first Sound Dubbing Suite at the South Bank Studios. Note the 8-Track A80 Studer with the last track marked ‘Maglink’. On top of the 8-Track machine is a Nagra ‘SLO’ synchroniser, and the ‘black box’ vary-speed controller. Gerry Ticehurst and Peter Mealing are obviously searching for some ‘sync’ audio on one of the 2-track A80’s. These also had the ‘Nagra sync’ facility mentioned below, and so we often still used with 50-Hz pulsed audio tapes from music recording sessions. The IVC helical-scan VTR was hidden away in a cupboard but having the 8-Track in the dubbing suite was deemed to be important for the amount of ‘manual manipulation’ it received!

****************************************************************************************

8| 1974| THE BBC SYPHER SYSTEM

In 1973 the BBC order a Neve console for their new servo controlled multitrack system for video dubbing, and was labelled by BBC’s Head-Of-Sound John Eden-Eadon as ‘SYPHER’ for ‘SYnchronised Post-dub Helical-scan and Eight-track Recorder’. It may have taken sometime to get the BBC designed synchroniser ‘up to speed’, and completed Sypher Suite was operational in 1974.

Like Thames TV, The BBC let the VT Dept do the transfers to the helical-scan and 8-Track, and also layback the finished dubbed tape. This was much disliked I imagine by Sound Supervisors as you need to complete the layback and check it ‘off-tape’ to really be able to sign it off! It also meant extra analogue tape generations and playbacks for the ‘dialogue’:
1: VTR 2-inch Original Recording
2: VTR 2-inch Editing copy
3: Multitrack Layoff (VTR Dept multitrack)
4: Mixdown – bounced within multirack (Sound Dept multitrack)
5: VTR Layback – Playback off VTR multitrack
6: Record to Master VTR 2-inch audio track.
Television wasn’t as carefully about ‘lining up’ audio tracks at every session, as commercial recording studios did, so slight freg. response errors in both playback and recording were going to be all to common!

****************************************************************************************

9| 1980| THE AUDIO KINETICS Q-LOCK

In 1979 I was manning LWT’s ‘Sound Transcription Unit’, waiting for my next commentary recording session, when I took a call from an ex-LWT Video editor Tim Whiffin. Tim had left LWT and was now working with Ian Southern, also ex-LWT, in Ian’s company Audio Kinetics. Having already produced a tape-machine locator unit, Tim was wanting to try out the ‘breadboard circuits’ of his new ‘synchroniser’ on our Studer A80.
Tim had been an outstanding VT Editor with considerable electronic skills and had amongst other things developed a fast-acting servo control for the RCA 2-inch VTs, that took the usual ‘pre-roll’ for a VTR from 10-seconds down to 3-seconds. That change was much loved by the studio ‘PA’s who had to accurately ‘count-in’ the arrival of each VTR playback insert! In 1970, I’d watch Tim use his own modified Revox A77 that used a 50-Hz pulse on audio track 2, to servo control its playback speed when laying back audio, well before we were able to do the same thing with A80 Studers.
In 1980 we got one of the first Audio Kinetic Q-Lock 2:10 synchronisers and realised it was far ahead of the somewhat ‘touchy’ Maglink we used, and it was soon in service in LWT’s dubbing suite and replaced by the 3-machine Q-Lock 3:10 pretty soon. This allowed us to have an A80 ‘timecoded 1/4-inch ‘slave’, as well as the VTR Helical machine and the Studer A80 multitrack.

****************************************************************************************

10| FADER AUTOMATION ARRIVES

Passing programmes through a sound dubbing suite after editing became the norm on television programmes during the 70’s, and every company opened at least one more suite, and upgraded to 16 and then 24-Tracks during the early ’80s. The outstanding invention was ‘Fader Automation’ which finally enabled complex sound dubs using so many tape tracks.

The first reference to a fader automation system built into a console that I could find was in one of the ads produced by the Canadian console manufacturer Olive in 1971. They designed an advanced console with a ‘recall’ system for many of the desk parameters, including the faders. Olive consoles though had a difficult time and the few that were built were unreliable, such as the one bought by Caribou Ranch in 1973 that was soon replaced by a Neve.

Others, like Allison Research in 1972 produced their Model 256 Programming Unit that could include the 940 Automated Fader, but it was a system looking for a ‘host mixing console’. Quad-Eight produced Compumix; a complete set of faders…..also looking to be hooked up to another console:

The ‘Line-level’ Compumix that plugged into your existing console, giving 24 automated faders and 6 groups. Like others at this time it didn’t use timecode but recorded the data to a track on the multitrack and then bounced that, plus new information onto a second track. Not in the photo is the fairly large rack of electronics it required.

1976| NECAM – real moving faders

GEOFF TANNER:
“The first moving fader automation was NECAM (meaning Neve computer assisted mixing) of 1976. NECAM used a toothed belt to carry the motion from the motor, so moving the fader spun the motor and made it feel quite heavy compared to a regular fader but it was the first and looked pretty damn impressive when first shown.”

In August 1976, Neve were undertaking an extensive conversion of one of AIR London’s existing studio desks to take NECAM.

We’ll see more of NECAM in sound dubbing desks later:

****************************************************************************************

11| MORE DUBBING SUITES and MORE TAPE TRACKS

As the UK recording studios quickly moved into using 16 and then 24-track machines, so did the TV broadcasters and soon all the major companies had one or more larger capacity dubbing suites, usually whilst still hanging onto their first ones though, as here with LWT’s:

LONDON WEEKEND TELEVISION, SOUTH BANK :

Above is the now very worn looking first Sound Dubbing Suite at LWT circa 1983; the same room as in the earlier picture. The original Neve BCM10 mixer is still in situ, as its 2 Groups still allowed for a ‘Monitor’ and a ‘Record’ Group. There are more powerful speakers; a pair of H&H powered KEF LS5-1AC’s and the synchroniser is now an AK Q-Lock. The original big ‘PEG’ (Programme Effects Generator) ‘sound-effect machine’, seen on the left is still running its short FX loops, and the grams area hidden on the right now has a couple of NAB Cart Decks for more FX. The multitrack is still the original A80 8-Track though. A Studer A80 16-Track arrived when Sound Dubbing 2, seen below, was commissioned.

1982| LWT Dubbing Suite 2:

The second LWT suite was constructed in 1982 and fell under the influence of the Eastlake’s ‘LEDE’ philosophy, that’s ‘Live-End/Dead-End’ where a hard stone surface at the front of the room is countered by a highly absorbant rear. True ‘Eastlake’ construction would have considerable bass traps, probably in the ceiling, which was certainly not an option here.

You don’t need a full set of Groups to feed the multitrack in a dubbing suite and here 4 Groups can be used to feed the 16-Track A80. The ‘need for more level’ when mixing music brought about the Tannoy SRM-15XB’s speakers seen in the above ad. They were not universally admired though.

circa 1980| Thames TV, Teddington :

Thames at Teddington had a new suite installed with a 16-Track Neve:

The upper photo shows Teddington’s updated suite after comissioning. They are using a Studer TLS2000 synchroniser and Tannoy SRM15’s which probably dates this to circa 1980, plus a U-Matic VCR. Thames had ‘free-grouping’ Neve’s in their studios, and the lack of group faders here indicates a similar approach on this multitrack desk, but there is no NECAM fitted.

The 2nd dubbing suite, with its ‘grams area’ off to the right.

BBC SYPHER SUITES:

1980 | SYPHER 2, BBC Television Centre:

The BBC at Television Centre built a second Sypher Suite with a 28-channel 16-Group Neve, which must have been one of the first with equipped with NECAM moving faders. The BBC continued using 8-Track A80s throughout, as all final mixes ended up on the 8-Track as these were fitted to the VTR layoff/layback suite, but Sypher 2 also had a 16-Track A80 to provide more tracks.

Here’s a great insight into Sypher 2’s capabilities, the BBC Traing video, showing Mike McCarthy dubbing ‘The Hitchhikers Guide to the Galaxy’ in 1980:

Also shown in that video was ‘Little SYPHER’, which was a simple sound suite for voice-overs and prep work.

1984| SYPHER and Film Dubbing, BBC Pebble Mill, Birmingham

The next SYPHER Suite built went to BBC Pebble Mill in Birmingham in 1984, when the existing film dubbing area was converted into a Sypher Suite for the ‘video’ work and a compatible film suite, both with Neve desks based on the new 51 Series and fitted with NECAM 96.

The BBC had always built their own synchronisers and surprisingly, the 3rd-generation ‘Maxim’ was still only ‘Master and one Slave’ capable and the two of the Studer A810 1/4-inch machines had their own synchronisers. As the ‘Sound of Sypher’ video showed, much of the triggering of FX sources was through a ‘GPI’ panel of relays using timecode.

Late 1980’s| SYPHER 3 and 4, BBC Television Centre:

At the end of the ’80s, the BBC moved to the buy the then very ‘fashionable’ Solid State Logic consoles, getting the ‘Broadcast’ version, the SL5000 for their 3rd and 4th Sypher Suites. The SSL’s had the capability of recall of all the desk settings, a boon for post-production mixing, but that recall of all the desk EQ and Aux controls was very labour intensive, as you had to work through every single channel matching the position of everyone the controls that had been used to those on the computer display. Unlike the Neve NECAM moving-fader system, the usual SSL fader automation used VCA faders and required you to match the fader position with the screen display to return to a previous level, and the faders themselves did not move, however the BBC went about changing this!

The BBC had for years used the Painton Quadrant fader (see ‘here’ for more on quadrants) but BBC Radio phased those out for flat faders once it started buying Neve consoles, although they kept the usual Beeb practice of ‘pull the fader knob towards you to fade-up’. Amazingly though the BBC TV mixers still preferred the curved top quadrant design, and despite the existing TV Neve’s also having flat P&G faders, they now decided to return to having ‘quadrant faders’ fitted to the SSL’s and they made SSL fit them as motorised ‘moving faders’! That must have cost a fortune in R&D work as it hadn’t been done before; all other SSL’s having flat VCA’s.

GARY CLARK:
“I remember being told each suite cost 250K when commissioned. Thus was considerably more than my 3 bed semi cost at the time.
Data was stored on Bernoulli 8 inch floppy drives. Audio file data was stored using a PCM video recorder.
Much time in Sypher was spent waiting for sound maintenance to come and fix the system though some truly great projects were mixed in the suites.
The quadrant faders had massive problems with the motors having to move the fader over a curve. They would frequently burn out. I had one catch fire on my while mixing. This caused no end of jokes about my hot mix”.

The ‘clout’ of getting a contract from the BBC, particularly a manufacturers ‘first one’, was regarded as so important that companies were always willing to take on unrealistic amounts of development costs. Surely developing a motoristed quadant fader would never have been compensated for by the BBC’s fee. The same costs problem had happened to Neve and Calrec with BBC work already.

In a review of the sound facilities at Television Centre published in 2003 the BBC had the same SYPHER Suites, but more changes had come:

LARRY GOODSON – HEAD-OF-SOUND:
“Although we don’t use the one inch video any more we have
still retained the eight track recorder, even though we use many more tracks
for track laying. We may use AudioFile and twenty-four track machines but because
the videotape and sound sections here are separate we have to have a standard
interchange medium with them, which is always an eight track audio tape and
a U-Matic video cassette.
The console in SYPHER 1 is a Neve 66. In SYPHER 2 there is an older Neve and
in 3 and 4 there are SSL 5000s. SYPHER 1 and 2 both have Necam 96 automation.
”It’s a system we like very much and we have it also in our music
studio. The programmable routing on the 66 is not a facility that’s used
very much in the post production area but the other facilities on the desk suit
us very well. We still use Necam 96 because that’s what we specified for
our music studio and what we retrofitted to the Neve in SYPHER 2. Because we
already had two systems when it came to SYPHER 1’s refurbishment, we specified
96 again because we knew the system and we don’t like swinging too many
changes on operators because it doesn’t actually help them any. So we said,
‘We think Flying Faders is wonderful but will you make us a 96 because
that’s what we have’, and they did and it was wonderful. I’m
not sure they would do it now, but at that time they had only just discontinued
96. If we were to carry on now I think we would have to change to Flying Faders
but that’s not a problem for the time being.
Also in SYPHER 1 we have grams and tapes, CD players, DAT machines and
an Akai sampler. In SYPHER areas, all tape recorders, quarter-inch, one-inch
and twenty-four track are timecoded and synchronisable. You can run anything
in the suite to picture, and because of additions we have made to the Akai samplers
we can run these locked as well, and of course AudioFile. Most of the synchronisers
are now Studer TLS4000s although because of the requirements of NECAM there
are some Adams Smith synchronisers as well. Long ago when we began this game
there weren’t any synchronisers around at all and the BBC built its own.
We looked at commercial synchronisers when they started coming out from other
manufacturers and although they worked, we were still building synchronisers
which were significantly faster on our kind of operation, largely because our
synchronisers relied on reading timecode while the machine is in spool. Most
manufacturers don’t even now do that. They tend to give them an address
to head for and sniff the timecode from time to time, which if you are doing
a rapid roll back and forward, roll back and forward repeatedly, is very slow
compared with reading timecode all the time. They did that because they believed
that reading timecode while spooling tape would wear the heads very quickly.
It doesn’t because an air cushion builds up. Some of our own synchronisers
still are faster, but most that we have now are commercial ones”.
From an article by David Mellor on Record-Producer.com

****************************************************************************************

12| WHAT WERE WE DUBBING?

Portable video cameras had not really arrived into common use until the early 1980’s, and therefore prior to that TV documentaries, which always require a considerable amount of audio post work were still on film.
Most TV dubbing work was ‘tarting up’ programmes that had past through the studios, although these sometimes had location OB and film sequences in them.

DRAMAS: Usually required the most sound work and something like LWT’s ‘Upstairs Downstairs’ period drama, would require interior and exterior FX to be added, if that hadn’t been done at the recording stage, plus occasionally music. The extensive use of ‘mood setting’ music scores was not common at this time. A sequence in a ball-room though, might be an interesting item for sound to cope with, as it was important to not mask any of the recorded dialogue and to allow for any subsequent picture editing, without any possibility of sound ‘jump cuts’. It was a typical procedure to play in a few bars of the music to help the actors, and then cut it for the rest of the scene. This had to be realistically replaced of course, with the required ‘perspectives’.

SIT-COMSs: These were the staple diet of both the BBC and ITV’s schedules and although recorded with real audiences present, it was expected that the audience reaction might be cut off during any picture editing. The sound dub would then ‘paper over’ the join with a new laugh, usually culled from elsewhere in the same show, or even recorded during the period when someone ‘warmed-up’ the audience before recording started. The golden rule in mixing the laughter from a ‘live audience’ was for the soundmixer to always have one hand on the main ‘Audience (Group) Fader’. The moment you stopped doing that……was when you ‘missed’ a laugh!
More contentious was adding additional laughter over sections. Producer’s often justified this by believing that the ‘joke’ required a better reaction than it got on the night. However when quizzed about this critiscm of ‘canned laughter tracks’, most Producer’s denied it ever happened!
There were though also comedy shows that were never recorded in front of an audience at all, as it was too complicated to record without many breaks. Usually these ended up being ‘played’ back to an audience to gather an ‘audience track’, that could then be used, even if altered a little.
A common problem with this was that the audience laughter could obscure an actor who ‘butted in’. Judging the gap to leave for audience reaction was a perennial problem actors had in recorded inserts.
ATV recorded ‘The Muppetts’ without any audience and used a clever 8-track cart system to trigger instant laughs that varied in level as you moved the fader through the carts ‘tracks’.

GAMESHOWS: Although most ‘bells and buzzers’ would happen during recording, sometimes these plus ‘move along’ music were added later. TV soundmixers were very adapt at handling a number of contestants speech mics at once of course. The technique of dropping those not talking by say 10 dB to ‘cleanup the spill’ was always used.

LIGHT ENTERTAINMENT and MUSIC SHOWS: Shows with ‘live house bands’ weren’t multitracked until the later years, but edits were frequent and music was often recorded in the ‘band room’ or outside recording studios and these were often relayed back into the final dub from mix tapes. Audience reaction was often ‘tarted up’ and most series of LE shows produced a ‘laugh tape’ of pre-edit laughter taken from the live recording for use in the dubs.

****************************************************************************************

13| THE INDEPENDENT SOUND DUBBING FACILITIES

With the arrival of Channel 4 in November 1982, using programmes made by independant production companies, brought about a need for both independent video editing suites and sound dubbing facilities. I’d left LWT in 1983 so got to use some of the new independent sound suites as the years rolled by, often for dubs to sit-coms, music shows and some ‘Mr Bean’s’, along with and a few video dramas.

1982| Complete Video, Slingsby Place, Covent Garden:

TV Producer Paul Smith had visited some US West Coast editing facilities and was impressed with their level of service and decided to bring the same to London. In 1982 he employed sound mixer Wouter van Herwerden to build a dubbing suite at Complete Video’s existing editing house in Covent Garden. Ir was equipped with a 32-channel SSL and used Q-Lock. The greatest joy of working there was ‘The Complete Diner’ – free cooked food! David Hamilton-Smith was usually at the desk when I did dubs at Complete.

1983| Videosonics, Hawley Place, Camden:

American recording Dennis Weinrich set up his own facility opposite TV-AM in Camden in 1983 and benefited from the new C4 work, so Dennis soon had a couple of rooms with Neve’s. Being a ‘sound-only’ facility, Dennis put in some of the new 1-inch C-Format VTR’s alongside his multi-tracks in his central machine room. Dennis made it a happy place to work.

Late 1970’s| Molinare, Fouberts Place, W1:

Australian sound engineer and (and cameraman) Stephan Sargent started ‘Moli’ in 1973 and it grew to encompass a couple of small TV studios, editing suites, and radio and TV sound studios. It moved to Fouberts Place in 1978…..and is still there. Certainly for sometime Moli used home built mixers and was never as ‘plush’ as the other two I’ve mentioned above.

1983| Limehouse Studios, Canary Wharf, Docklands:

Limehouse is the place I joined after LWT and is an example of the way ‘in-house’ dubbing was often undertaken as the number of UK TV shows made by independent producers continued growing
Limehouse was a two studio facility, built by converting an old ‘banana warehouse’ in London Docklands into a well equipped and beautiful looking facility in 1983. The studios made a variety of prorammes for Channel 4 and any sound dubbing required was carried out in the existing TV sound control rooms.

With a cheap video deck and a suitable synchroniser linked to your multitrack – you could be ‘dubbing’ in any suitable sound control-room. Limehouse hitched up their Otari- MTR-90 24-Tracks to the Q-Lock 3:10, seen in the desk racking, and dubbed to a U-matic’s pictures.
By the way, it wasn’t just the BBC who had their ‘faders backward’ where you pull the fader toward you to fade-up…..so did Thames TV, Southern TV and Limehouse, which was also started by ex-Southern TV people.
We worked on the drama, sit-coms and music shows produced at the studios, and sometimes used a digital Sony 24-Track as well. A notable music show of 1985 was ‘Blue Suede Shoes’ that was broadcast as another simulcast with Capital Radio with an all-star cast of Carl Perkins, George Harrison, Eric Clapton and Ringo Starr, should have been well worth watching and listening to.
Alas, the ‘stereo sound’ turned out to be virtually ‘mono’:

BARRY FOX – STUDIO SOUND MAY 1986:
“The reason was beyond the control of Channel 4 and Capital. Dave Edmunds
mixed the sound from a Limehouse multitrack and took a deliberate, policy
decision not to go for wide stereo. He wanted a sound in keeping with the era
of the original music. There was talk of a re -mix for the simulcast but it was
vetoed. So what went out was as near as makes no difference mono.
Under the circumstances it was probably a good thing that Channel 4 failed
miserably to advertise the simulcast in advance together they will be able to find
something with a bit of width”.

***********************************************************************************************************

In the next part we’ll look at how ‘digital’ transformed the sound post-production world.

CREDITS and REFERENCES:

[1] Lots more historic photos from both ABC and Thames TV can be viewed at: arts-tv.org.uk, the ‘Association of Retired Thames Staff’ website, a great resource.
[2] Many thanks to Mike Pontin, ex-Thames TV’s Head-Of-Sound, for talking to me and writing up his memories of the MEDWAY system.
[3] The SSL computer certainly revolutionised the task of mixing many audio tracks, but if you want to see how slow it was ‘back then’, there’s an extremely good look at the SSL’s automation on a video made by Sound-On-Sound magazine, with Karen Down explaining all the details on an SSL6000E:

****************************************************************************************