Thermionic
11-11-2010, 15:42
During the mid 1970’s the Decca Recording Company designed and developed for its own use the first successful digital audio recording and post production system. Building on a history of genuine innovation and technical achievement, the Decca team at West Hampstead developed a digital mastering system that was for a time probably the best stereo recording system on the planet.
Seeking tangible improvements in the many production processes between the microphones in the concert hall and the final LP or CD, Decca’s development team lead by Tony Griffith took the experience they had gained in developing an early form of digital video disk system known as 'Teldec' and built what was also to become probably the longest lived digital audio post-production system.
Remarkably, as newer commercial alternatives were introduced such as the 3M multitrack system, the U-Matic based 1620, the Sony F1 combination or DASH linear reel-to-reel recorders, Decca were never sufficiently impressed enough to change their ways of working. Either the sound quality was inferior, or the industrial video-cassette based machines were too cumbersome and slow in use, or combinations of both. (Those of us old enough to recall linear video editing, will appreciate how time-consuming it was while waiting for the videocassette machines to clunk their way through a tape.)
Decca started using their digital system in the late 1970's, and I was lucky enough to witness it still in use in November 1997. This was a twenty year plus working lifespan, which for any piece of pioneering technology seems rather impressive.
Michael Mailes worked at Decca for 45 years and 1 day, so I imagine that during this time he must have seen and done it all. I hope he won’t mind me including here his comments I posted on my web site some years ago, as I am sure they will be of general interest to members of this forum:
"Having been a balance engineer/technical engineer I was involved a lot with Tony and his group in the early days. We had to put them right in what was really needed by the boys in the field as opposed to people working in nice "BBC" type situations! All classical recording sessions were done in outside locations, and in the case of the USA/Canada under extremely expensive time restrictions.
We at the recording studios in West Hampstead, (later to become The English National Opera's building) had little to do with the Video department, approx. 4 miles away. Only when the videodisc was abandoned did we start to have contact with Tony Griffiths and his team. We knew very little about Digital Recording, only what we had read in magazines. Soundstream was getting some publicity, but to Decca who did everything 'in house', having editing done in the middle of the USA was out of the Question.
Interesting perhaps to note that in 1980 John Williams became musical director of The Boston Pops orchestra. I had made recordings with Arthur Fiedler/Boston Pops and had experience working in Symphony Hall. Philips made a contract to record J.W and found that they didn't have equipment or crew in America at the time of the 1st proposed recording date. Decca had recording equipment in the States having made contracts with several orchestras. Logical conclusion! Ask Decca!
Together with Stanley Goodall (Recording engineer) we made J.W's first recording; 'Pops In Space' recorded on the Soundstream system.
Having got the Decca system up and running, with a fair thought for what was likely to be needed by the recording department, demonstration time arrived. Recordings at this time were being made on Studer B62 machines (2 for safety, tapes being flown back from USA & Canada on 2 separate planes). Tapes ran for 1Hr. Machines and the operator were in the control room; timings were by mechanical counter. IVC Machines would have to be in another room due to the noise of motors and fans, therefore required,
remote control, simple box with transport controls? How to locate take start points? Time codes could not be read when tape was spooling. On analogue recordings a 50hz tone was put at the end of each take which by pushing the tape near the heads while spooling resulted in a higher pitch so takes could be located by counting the blips. Using one of the analogue tracks on the IVC and bringing the output back to headphones solved this problem. Tape speed was giving 1 Hr, later extended to 75 mins by reducing tape speed. Having to spool tapes back on to the feed reel was time consuming and could be a problem in America where minutes counted on recording sessions. The 75 minutes made planning tape changes a lot easier. Video practice was to use time codes continuously running, no good to us with machines in another room. Sessions ran on a stop/start process, time being allowed for play back to artists. We would have no idea how much tape was left on a reel. This was easily fixed, machine stops, TC stops!
Analogue machines having record and replay heads meant that rare tape faults were heard at the time, another requirement for a digital system. The IVC machine of course had this, tales of other record companies having to use parts of analogue recordings edited into digital recordings filtered through to us! Access time for playbacks was just like analogue, none of the clunking and clicking of U-Matics for us.
Another perhaps interesting note, For some years Decca had been making many of the Readers Digest recordings.
RCA doing the production/mastering. They chose to use JVC as a recording system, on sessions the producer liked us to use our system as well in order to make play back to artists faster!
The Decca principal for recording was to get it right on the session (apart from editing) so post production was a no no! wherever possible. However the 'time factor' on opera recordings meant that it was not always possible to do a retake if a singer was drowned out by a too loud orchestra or chorus. Along with our very reliable B62's we used Studer J37 (4 tracks), later 8 track A80. Next project make us a 4 track digital recorder. This was done, almost overnight, by adding extra cards (and modifying others) to a 2 track processor, extending its length some 8 inches. With the aid of the Decca Digital Mixer 2 track masters could be made with out going back to analogue. Very little was changed during the life of the recording system. Better remote facilities came along using a simple computer mother board and VDU. This allowed us to write take data onto the tape and allowed for faster playbacks, locating different takes on both machines simultaneously.
Not a perfect system, nothing was, we suffered a bit with with bad batches of tape, rare faults with the processors, like most electronics if it works for a month it will probably go on forever. IVC recorders to us (compared to the Studer machines, beautifully built easy to maintain and set up) were perhaps more semi-professional. They didn't always travel well.
In the early days of Digital recording analogue machines were used in parallel, however I think it true to say that every digital recording issued was made totally digitally.
I believe this to be true: post war, Decca together with EMI were given captured German tape recorders, EMI as we know went into the recorder market. Recorders and loudspeakers were the only part of the recording chain Decca didn't make. Cutting lathes, cutting heads, amplifiers, some microphones, mixers, all made in house. The reputation of Decca with its Full Frequency Range Recording, came about during WW2 when they developed a full frequency range cutter whilst doing 'secret' work for the navy.
Decca, having its own development engineers (and workshops both mechanical and electrical) was always looking for ways to improve audio. Cutters were capable of FFRR/FFSS but tape was the weak link in the chain. I believe that by the time Digital Recording came along analogue tape was about as good as it was going to get. Being mostly involved with classical recording with its wide dynamic range tape hiss was always our biggest problem. Putting high audio levels onto tape, sometimes higher than was recommended! produced distortion and compression at high frequencies. Recording to the recommended peak level meant that tape hiss was more obvious. Pre-echo was another problem with analogue tape, loud music 'bled' through at a lower volume to a previous section of tape on the feed reel. Unlike some companies that copied their 'Master' tapes and used the copies for mastering Decca worked from the original (edited) tapes. Two machines being used on sessions to give a second chance in case of damage when editing, physically cutting tapes and joining with sticky tape was a shade primitive. With the need to make copy tapes for export to licensee's abroad, master tapes for cassette production, radio stations etc an extra amount of hiss/distortion was added. This also meant that for compilation tapes for records like: The worlds best tenors, Your favourite overtures etc, a copy tape had to be made. Building our own amplifiers for tape machines and tweaking others we had probably got as good as we were going to. The only recordings that didn't suffer from tape hiss were Organ recordings; the instruments always produced their own background noise!
So now to Digital. Signal to noise/distortion, only the A-D's & D-A's to worry about since you got back from the tape exactly what you put on it. Our converters were 'home designed and built' so we were again able to learn from the horse's mouth how to keep them aligned for best performance. First converters 16 bit, later 20 bit. Tony Griffiths and his team were adamant the sampling rate should be 48Khz, Sony plumping for 44Khz. This meant that when the CD came along sample rate converters had to be made in order to produce master tapes for the CD production process as the factories mastered from U Matic machines. When Mitsubishi produced their 32-track recorder it was capable of 44, 48 and 50K, also the earlier 3M 32 track machines could be used at 48K.
Record buyers heard only the sound through their own audio equipment; they did not have the opportunity to hear the sound direct from the recording console. We as engineers could compare the console output to the off-tape sound. Comparing off-tape digital sound/off-tape analogue sound and direct from the console sound it was obvious that Digital won hands down. Of course the first Digital records were issued on vinyl. Vinyl versus Compact Disc, another controversial point for discussion! There were many critical comments made about 'Digital' mostly totally invalid. If people preferred analogue sound it was not generally because of technical reasons it was just 'The sound' that they were used to.
I remember after a playback on one of the early digital recording sessions in Chicago, (Solti/Chicago Symphony Orchestra) one of the woodwind players asking, "was that a digital playback?" on being told it was, he said "I thought so, the woodwinds and upper strings sound much cleaner". There was a fairly large anti digital faction but on many occasions we put on carefully controlled demonstrations and usually convinced people that digital recording were truer to the sound the engineers produced on recording sessions. It was said I believe at an early demonstration of the wax cylinder that it was difficult to tell the difference between the recording and the live sound!!!!!!
Any hardcore AOS members using wax cylinders out there?:eyebrows:
The system shown here is one of the acquision digital tape recorders used on location. These machines were regularly shipped all around the world. Decca’s preferred method was to use a pair of these recorders locked together in tandem to capturing the performance. Back at the Recording Centre in Belsize Road, London the tapes would be edited in the digital domain and then transferred to analogue tape for the disc cutting suite or (post the introduction of CD) to the Sony 16 bit U-Matic PCM16xx standard for transfer to CD.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0003.jpg
This (hopefully) still operational recorder is I believe one of only a few Decca machines now left in the UK, and it consists of (above):
1) An 18 bit digital to analogue converter (top 1U rack unit) using a BBC asic, an Ultra Analog 2040 convertor and Apogee low dispersion filters. This was configured to process SPDIF and AES as well as the Decca signal, and was typically used in the CD transfer suite to quality check CD playback. (I use this unit almost on a daily basis and it sounds at least as good as my Marantz DA12.)
2) A 20 bit analogue to digital converter(lower 1U rack unit). This takes a studio line level signal and converts it to a digital data stream in the Decca format. The sample rate is 48 kHz. Over time as convertor technology improved they were able to increase the resolution, this late example has DBX chips. (It seems that post their noise reduction system DBX became involved with digital conversion technology.)
3) A signal processing unit (euro card frame) that takes the digital data stream and converts it into a signal that a modified video transport can record (such as adding the various video sync pulses). It also generates timecode and error correction data. Working simultaneously in record and playback it also displays off-tape record level via a peak hold PPM. It reads the off-tape timecode and monitors the overall quality of the recording (by counting lost samples). It also has a limited amount of system self-diagnostics.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-1.jpg
4) A modified IVC800 series 1-Inch helical video tape transport (above). In the early 1970's these American machines came in a number of versions and were well regarded as professional non-broadcast video recorders. The 1 Inch tape was a fairly robust medium and the single head alpha wrap system gave good resolution. This type of machine would have been a better choice than perhaps an equivalent 1/2 Inch video tape recorder. This version is capable of off-tape monitoring (having 3 rotary video heads; erase, record and play). A special ‘Decca’ record modulator is fitted as well as a microprocessor controlled jog/shuttle edit control system using the video industry standard RS422 interface (I think).
Decca's works cost for one of these recording systems was about £13,000 (in 1970's money), which perhaps represented at the time a quantity of electronics and expense similar to that of a contemporary mini computer. They used an industrial video tape transport because digital audio had a much higher data through-put than any computer memory or tape drive of the time could cope with. In the long term Decca planned to use a type of optical disk to record the audio. (How far-sighted was that?)
Of course the use of modified video recorders of various formats was later to become the accepted way of storing audio digits for some time (as with Sony's U-Matic based 1600/10/30 equipment, and their later F1 and R-DAT systems).
Previously to this digital audio system Decca and Telefunken had collaborated on an ill-fated video disc playback system designed for consumer use, a sort of record player for TV (the prototype used a Garrard Lab80 I recall). Heading up this team was the ex-BBC Mr Tony Griffiths, and while not succeeding with this venture his team were however fresh from wrangling with digital video processing, slow speed telecine machines and floppy (vinyl!) disk digital recording. So transferring digital audio to a reliable
workhorse of an IVC video recorder was not thought to be difficult. In Mr Griffths words, 'apart from the a/d development, the digital processing posed no problems'.
This is of course a question of perspectives...
My own involvement with Decca came about by in interest in old television technology. To cut a long story short; over the years I had developed some low key contact with Decca as they used IVC machines similar to mine. Eventually they asked if I might like to acquire some of their IVC equipment as they would be closing the Recording Centre and would be getting rid of most of it.
In the end I made two visits, and at the time of our first visit the facility was still very much alive. It was remarkable to see the extent to which Decca had developed their by then very venerable digital recording chain and how this system was interfaced to their industry standard Sony 1630 CD mastering equipment.
When we eventually found the place, Decca's famous and historic facility looked from the outside to be a rather anonymous and ageing warehouse. With Abbey Road just down the road this equally famous 'Recording Centre' was a bit of a contrast. The ground floor seemed to be made up of storage facilities for their location recording equipment. Within this area, their extensive collection of equipment was housed in many 'well-used' flight cases, piled up on rows of shelving.
Having ascended a large service lift (in pitch darkness actually as the light had blown), we came out into a bright modern looking 'hospitality' area. I recall seeing a billiard or Ping-Pong table, sundry bits of seating and various tea and coffee making facilities. This was like a contemporary office reception area or indeed what recording studios seem to look like these days; lots of light wood and Grey painted surfaces.
On a wall was a clock with its face fashioned from an LP with hour batons marked 'DECCADIGITAL'. From this recreation area we were taken through a couple of double doors, past Decca modified Studer A80 recorders skulking in the corridors, into what proved to be a sort of editing and post-production area. Adjacent to this room were 2 or 3 sound proof editing suites. To our right on the other side of a glass partition wall were offices. To our left were three substantial 'watertight doors' (like in submarines but bigger) separated from each other by well stuffed (buttons, displays and winking lights) six foot equipment bays. Spread out in front of us were a collection of what at first sight looked like old Mahogany radiogram cabinets. Each of these rather odd-looking and randomly parked objects contained an IVC videotape transport under a smoked Perspex lid, and were in fact acoustic enclosures.
While we were chatting we could see that some of these machines were run-up and 'rocking and rolling' as one of the editors worked away behind one of the 'watertight' doors. As far as we could see, mostly everything seemed to have been exclusively Decca in origin. The equipment racks contained in-house processing, time code and editing electronics. I had assumed that bits of PC lurked inside, but this equipment was designed before IBM processors were fast enough to do the job, and were actually Decca designed computers based on AMD 68000 processors.
The editing rooms themselves were quite small, and were perhaps some 10 or 12 feet square. They were lined with dark-wood effect acoustic treatment. Roughly in the middle of the room was a control desk on which was placed a computer monitor, a PC type keyboard, a commercial video-editing controller, and a basic Decca digital audio mixer.
Digital peak level metering was done via the computer monitor, though for making analogue copy tapes for artist's approval a pair of old fashioned PPM's were used. We were told that Decca actually invented the Peak Programme Meter and not the BBC.
All the sound mixing and editing was done in the digital domain. Each of these desks used dither to remove quantising distortion, and had two channels of digital equalisation. They also had a programmable notch filter that could synthesise up to 3 notches of varying width and depth. This was used to 'remove anything from Hum to TV line frequency whistle'.
About four feet in front of the desk and some six feet apart, were a pair of stand-mounted B&W 801 monitors. They were set facing inwards towards the operator and were driven by a large HH Mosfet power amplifier (500 Watts?). I have to say that the sound quality could have been better, it was loud enough though. But this monitoring system was not for balance or ultimate quality evaluation, just for making sure that the edits were good and there were no obvious problems with the recording.
The computer monitor displayed the music's 'waveform' together with a time cursor. Standard stuff now but this was all their own. Decca were the first users of video editing software for audio post-production and they developed their own ways of doing this which were years ahead of anyone else.
As with video-editing, the various edit points are stored in the editing computer’s memory and can be changed before actually making the edit. With this equipment, the edits which are in fact very rapid cross fades between two playback machines onto one recording machine are made with high accuracy, far more precise than with a razor blade (that everyone else used).
This was all done before PCs were fast enough so they built their own editing computers and memories. Ram was also used to store a limited amount of the digitised audio itself. This was to speed up the editing process while the mechanical tape machines caught up. This was far quicker in use than Sony's U-Matic based commercial digital audio system.
It was quite a revelation to realise the amount of editing done to something as seemingly 'pure' as a classical music recording. It was not at all a question of just setting up some mikes in the right places and hitting the record button. For perhaps several dozen and possibly hundreds of edits were carried out on each work.
‘The sound was always that at the performance but the musical accuracy was much better. This results in a play back free from the annoying defects, which become recognisable points of anticipation after a few playbacks.'
While loudspeakers were used most of the time during the editing process, the operator actually checked the quality and integrity of each 'splice' with a pair of headphones.
Almost all of the Decca system was of their own design and manufacture, at first with their own soldering irons and 'Veroboard', though later outside subcontractors were used. But most of the final assembly and software programming remained in house.
Mr Griffiths advised me that Denon in Japan and Tom Stockhom in America (Soundstream?) also had developed early digital audio systems. But Denon's system was 12 bit (again using video recorders), and Mr Stockhom's used a multi-channel instrumentation recorder. Stockhom also attempted recording audio direct to a mainframe computer, but this apparently took 'most of the night to upload'.
Decca were well aware of these other systems, but their recording teams said "If it won't do what we normally do, then we won't use it".
During our first visit we also saw a couple of other mixing suites that contained more advanced digital facilities, including EQ and digital time alignment for microphones (another Decca first). One lone special digital mixing desk sitting in solitary splendour was also fitted with advanced 'endless belt' type digital faders. Their 8 channel mixers used 64 bit arithmetic, with dedicated and home brewed TTL processing for high speed calculation. (This was in the days of the Cray-1which used similar circuitry.) Both DG and Phillips were supplied with versions of this mixer for their own mastering systems. Recording memory for the various editing suites was provided by the various IVC transports laying around in the central post-production machine area as described.
Sadly this was all to end, and in December 1997 this facility was closed down. Ten of perhaps thirty five digital recorders shipped were to Polygram in Holland for archive transcription purposes. The rest were scrapped or found their way to one or two new homes (and I have Polygram's paid invoice!). It had been decided to use normal commercially sourced equipment in the future and not to continue with development or use of the Decca equipment. This also included their next generation of a newly developed magneto-optical disc system as well (see image and info below). In fact some ex employees had already departed to set up Gennex who manufactured the GX series of optical recorders, which were built upon Decca technology.
Some further images:
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-2.jpg
20 bit A-D convertor with cover removed.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-3.jpg
18 bit D-A convertor with cover removed. The D-A board is on the left and the large black rectangle is the Audio Analog hybrid D-A circuit. Next is a 'Decca' PLL board and next to that an input selector switch board, and on the Right is a switched mode PSU. (The front mounted phono sockets are my addition.)
And to conclude I thought it was worth going to the trouble of presenting here an account of what little I know about this minor footnote of British audio history. Decca kept their cards close to their chests because they thought patenting was not cost effective, and I suppose some mystery did not harm 'the brand'. But their undoubted significant technical achievement should not be overshadowed by the more well known commercial products from Sony et all. Simply put, Decca were there doing the job with a better system long before all those other guys.
What might have been... This (below) is probably the only surviving example of a new optical transport that Decca were working on when they were shut down. It uses a Maxoptic 5 Inch magneto-optical disc to record four audio channels and has a much improved user interface. This was to be used with a laptop PC and a separate set of digital convertors via a MADI interface. Unfortunately this example doesn't appear to be operational.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-4.jpg
Most of this was an edited version of the Decca section of my web site, and now I need a large :cool:
Seeking tangible improvements in the many production processes between the microphones in the concert hall and the final LP or CD, Decca’s development team lead by Tony Griffith took the experience they had gained in developing an early form of digital video disk system known as 'Teldec' and built what was also to become probably the longest lived digital audio post-production system.
Remarkably, as newer commercial alternatives were introduced such as the 3M multitrack system, the U-Matic based 1620, the Sony F1 combination or DASH linear reel-to-reel recorders, Decca were never sufficiently impressed enough to change their ways of working. Either the sound quality was inferior, or the industrial video-cassette based machines were too cumbersome and slow in use, or combinations of both. (Those of us old enough to recall linear video editing, will appreciate how time-consuming it was while waiting for the videocassette machines to clunk their way through a tape.)
Decca started using their digital system in the late 1970's, and I was lucky enough to witness it still in use in November 1997. This was a twenty year plus working lifespan, which for any piece of pioneering technology seems rather impressive.
Michael Mailes worked at Decca for 45 years and 1 day, so I imagine that during this time he must have seen and done it all. I hope he won’t mind me including here his comments I posted on my web site some years ago, as I am sure they will be of general interest to members of this forum:
"Having been a balance engineer/technical engineer I was involved a lot with Tony and his group in the early days. We had to put them right in what was really needed by the boys in the field as opposed to people working in nice "BBC" type situations! All classical recording sessions were done in outside locations, and in the case of the USA/Canada under extremely expensive time restrictions.
We at the recording studios in West Hampstead, (later to become The English National Opera's building) had little to do with the Video department, approx. 4 miles away. Only when the videodisc was abandoned did we start to have contact with Tony Griffiths and his team. We knew very little about Digital Recording, only what we had read in magazines. Soundstream was getting some publicity, but to Decca who did everything 'in house', having editing done in the middle of the USA was out of the Question.
Interesting perhaps to note that in 1980 John Williams became musical director of The Boston Pops orchestra. I had made recordings with Arthur Fiedler/Boston Pops and had experience working in Symphony Hall. Philips made a contract to record J.W and found that they didn't have equipment or crew in America at the time of the 1st proposed recording date. Decca had recording equipment in the States having made contracts with several orchestras. Logical conclusion! Ask Decca!
Together with Stanley Goodall (Recording engineer) we made J.W's first recording; 'Pops In Space' recorded on the Soundstream system.
Having got the Decca system up and running, with a fair thought for what was likely to be needed by the recording department, demonstration time arrived. Recordings at this time were being made on Studer B62 machines (2 for safety, tapes being flown back from USA & Canada on 2 separate planes). Tapes ran for 1Hr. Machines and the operator were in the control room; timings were by mechanical counter. IVC Machines would have to be in another room due to the noise of motors and fans, therefore required,
remote control, simple box with transport controls? How to locate take start points? Time codes could not be read when tape was spooling. On analogue recordings a 50hz tone was put at the end of each take which by pushing the tape near the heads while spooling resulted in a higher pitch so takes could be located by counting the blips. Using one of the analogue tracks on the IVC and bringing the output back to headphones solved this problem. Tape speed was giving 1 Hr, later extended to 75 mins by reducing tape speed. Having to spool tapes back on to the feed reel was time consuming and could be a problem in America where minutes counted on recording sessions. The 75 minutes made planning tape changes a lot easier. Video practice was to use time codes continuously running, no good to us with machines in another room. Sessions ran on a stop/start process, time being allowed for play back to artists. We would have no idea how much tape was left on a reel. This was easily fixed, machine stops, TC stops!
Analogue machines having record and replay heads meant that rare tape faults were heard at the time, another requirement for a digital system. The IVC machine of course had this, tales of other record companies having to use parts of analogue recordings edited into digital recordings filtered through to us! Access time for playbacks was just like analogue, none of the clunking and clicking of U-Matics for us.
Another perhaps interesting note, For some years Decca had been making many of the Readers Digest recordings.
RCA doing the production/mastering. They chose to use JVC as a recording system, on sessions the producer liked us to use our system as well in order to make play back to artists faster!
The Decca principal for recording was to get it right on the session (apart from editing) so post production was a no no! wherever possible. However the 'time factor' on opera recordings meant that it was not always possible to do a retake if a singer was drowned out by a too loud orchestra or chorus. Along with our very reliable B62's we used Studer J37 (4 tracks), later 8 track A80. Next project make us a 4 track digital recorder. This was done, almost overnight, by adding extra cards (and modifying others) to a 2 track processor, extending its length some 8 inches. With the aid of the Decca Digital Mixer 2 track masters could be made with out going back to analogue. Very little was changed during the life of the recording system. Better remote facilities came along using a simple computer mother board and VDU. This allowed us to write take data onto the tape and allowed for faster playbacks, locating different takes on both machines simultaneously.
Not a perfect system, nothing was, we suffered a bit with with bad batches of tape, rare faults with the processors, like most electronics if it works for a month it will probably go on forever. IVC recorders to us (compared to the Studer machines, beautifully built easy to maintain and set up) were perhaps more semi-professional. They didn't always travel well.
In the early days of Digital recording analogue machines were used in parallel, however I think it true to say that every digital recording issued was made totally digitally.
I believe this to be true: post war, Decca together with EMI were given captured German tape recorders, EMI as we know went into the recorder market. Recorders and loudspeakers were the only part of the recording chain Decca didn't make. Cutting lathes, cutting heads, amplifiers, some microphones, mixers, all made in house. The reputation of Decca with its Full Frequency Range Recording, came about during WW2 when they developed a full frequency range cutter whilst doing 'secret' work for the navy.
Decca, having its own development engineers (and workshops both mechanical and electrical) was always looking for ways to improve audio. Cutters were capable of FFRR/FFSS but tape was the weak link in the chain. I believe that by the time Digital Recording came along analogue tape was about as good as it was going to get. Being mostly involved with classical recording with its wide dynamic range tape hiss was always our biggest problem. Putting high audio levels onto tape, sometimes higher than was recommended! produced distortion and compression at high frequencies. Recording to the recommended peak level meant that tape hiss was more obvious. Pre-echo was another problem with analogue tape, loud music 'bled' through at a lower volume to a previous section of tape on the feed reel. Unlike some companies that copied their 'Master' tapes and used the copies for mastering Decca worked from the original (edited) tapes. Two machines being used on sessions to give a second chance in case of damage when editing, physically cutting tapes and joining with sticky tape was a shade primitive. With the need to make copy tapes for export to licensee's abroad, master tapes for cassette production, radio stations etc an extra amount of hiss/distortion was added. This also meant that for compilation tapes for records like: The worlds best tenors, Your favourite overtures etc, a copy tape had to be made. Building our own amplifiers for tape machines and tweaking others we had probably got as good as we were going to. The only recordings that didn't suffer from tape hiss were Organ recordings; the instruments always produced their own background noise!
So now to Digital. Signal to noise/distortion, only the A-D's & D-A's to worry about since you got back from the tape exactly what you put on it. Our converters were 'home designed and built' so we were again able to learn from the horse's mouth how to keep them aligned for best performance. First converters 16 bit, later 20 bit. Tony Griffiths and his team were adamant the sampling rate should be 48Khz, Sony plumping for 44Khz. This meant that when the CD came along sample rate converters had to be made in order to produce master tapes for the CD production process as the factories mastered from U Matic machines. When Mitsubishi produced their 32-track recorder it was capable of 44, 48 and 50K, also the earlier 3M 32 track machines could be used at 48K.
Record buyers heard only the sound through their own audio equipment; they did not have the opportunity to hear the sound direct from the recording console. We as engineers could compare the console output to the off-tape sound. Comparing off-tape digital sound/off-tape analogue sound and direct from the console sound it was obvious that Digital won hands down. Of course the first Digital records were issued on vinyl. Vinyl versus Compact Disc, another controversial point for discussion! There were many critical comments made about 'Digital' mostly totally invalid. If people preferred analogue sound it was not generally because of technical reasons it was just 'The sound' that they were used to.
I remember after a playback on one of the early digital recording sessions in Chicago, (Solti/Chicago Symphony Orchestra) one of the woodwind players asking, "was that a digital playback?" on being told it was, he said "I thought so, the woodwinds and upper strings sound much cleaner". There was a fairly large anti digital faction but on many occasions we put on carefully controlled demonstrations and usually convinced people that digital recording were truer to the sound the engineers produced on recording sessions. It was said I believe at an early demonstration of the wax cylinder that it was difficult to tell the difference between the recording and the live sound!!!!!!
Any hardcore AOS members using wax cylinders out there?:eyebrows:
The system shown here is one of the acquision digital tape recorders used on location. These machines were regularly shipped all around the world. Decca’s preferred method was to use a pair of these recorders locked together in tandem to capturing the performance. Back at the Recording Centre in Belsize Road, London the tapes would be edited in the digital domain and then transferred to analogue tape for the disc cutting suite or (post the introduction of CD) to the Sony 16 bit U-Matic PCM16xx standard for transfer to CD.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0003.jpg
This (hopefully) still operational recorder is I believe one of only a few Decca machines now left in the UK, and it consists of (above):
1) An 18 bit digital to analogue converter (top 1U rack unit) using a BBC asic, an Ultra Analog 2040 convertor and Apogee low dispersion filters. This was configured to process SPDIF and AES as well as the Decca signal, and was typically used in the CD transfer suite to quality check CD playback. (I use this unit almost on a daily basis and it sounds at least as good as my Marantz DA12.)
2) A 20 bit analogue to digital converter(lower 1U rack unit). This takes a studio line level signal and converts it to a digital data stream in the Decca format. The sample rate is 48 kHz. Over time as convertor technology improved they were able to increase the resolution, this late example has DBX chips. (It seems that post their noise reduction system DBX became involved with digital conversion technology.)
3) A signal processing unit (euro card frame) that takes the digital data stream and converts it into a signal that a modified video transport can record (such as adding the various video sync pulses). It also generates timecode and error correction data. Working simultaneously in record and playback it also displays off-tape record level via a peak hold PPM. It reads the off-tape timecode and monitors the overall quality of the recording (by counting lost samples). It also has a limited amount of system self-diagnostics.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-1.jpg
4) A modified IVC800 series 1-Inch helical video tape transport (above). In the early 1970's these American machines came in a number of versions and were well regarded as professional non-broadcast video recorders. The 1 Inch tape was a fairly robust medium and the single head alpha wrap system gave good resolution. This type of machine would have been a better choice than perhaps an equivalent 1/2 Inch video tape recorder. This version is capable of off-tape monitoring (having 3 rotary video heads; erase, record and play). A special ‘Decca’ record modulator is fitted as well as a microprocessor controlled jog/shuttle edit control system using the video industry standard RS422 interface (I think).
Decca's works cost for one of these recording systems was about £13,000 (in 1970's money), which perhaps represented at the time a quantity of electronics and expense similar to that of a contemporary mini computer. They used an industrial video tape transport because digital audio had a much higher data through-put than any computer memory or tape drive of the time could cope with. In the long term Decca planned to use a type of optical disk to record the audio. (How far-sighted was that?)
Of course the use of modified video recorders of various formats was later to become the accepted way of storing audio digits for some time (as with Sony's U-Matic based 1600/10/30 equipment, and their later F1 and R-DAT systems).
Previously to this digital audio system Decca and Telefunken had collaborated on an ill-fated video disc playback system designed for consumer use, a sort of record player for TV (the prototype used a Garrard Lab80 I recall). Heading up this team was the ex-BBC Mr Tony Griffiths, and while not succeeding with this venture his team were however fresh from wrangling with digital video processing, slow speed telecine machines and floppy (vinyl!) disk digital recording. So transferring digital audio to a reliable
workhorse of an IVC video recorder was not thought to be difficult. In Mr Griffths words, 'apart from the a/d development, the digital processing posed no problems'.
This is of course a question of perspectives...
My own involvement with Decca came about by in interest in old television technology. To cut a long story short; over the years I had developed some low key contact with Decca as they used IVC machines similar to mine. Eventually they asked if I might like to acquire some of their IVC equipment as they would be closing the Recording Centre and would be getting rid of most of it.
In the end I made two visits, and at the time of our first visit the facility was still very much alive. It was remarkable to see the extent to which Decca had developed their by then very venerable digital recording chain and how this system was interfaced to their industry standard Sony 1630 CD mastering equipment.
When we eventually found the place, Decca's famous and historic facility looked from the outside to be a rather anonymous and ageing warehouse. With Abbey Road just down the road this equally famous 'Recording Centre' was a bit of a contrast. The ground floor seemed to be made up of storage facilities for their location recording equipment. Within this area, their extensive collection of equipment was housed in many 'well-used' flight cases, piled up on rows of shelving.
Having ascended a large service lift (in pitch darkness actually as the light had blown), we came out into a bright modern looking 'hospitality' area. I recall seeing a billiard or Ping-Pong table, sundry bits of seating and various tea and coffee making facilities. This was like a contemporary office reception area or indeed what recording studios seem to look like these days; lots of light wood and Grey painted surfaces.
On a wall was a clock with its face fashioned from an LP with hour batons marked 'DECCADIGITAL'. From this recreation area we were taken through a couple of double doors, past Decca modified Studer A80 recorders skulking in the corridors, into what proved to be a sort of editing and post-production area. Adjacent to this room were 2 or 3 sound proof editing suites. To our right on the other side of a glass partition wall were offices. To our left were three substantial 'watertight doors' (like in submarines but bigger) separated from each other by well stuffed (buttons, displays and winking lights) six foot equipment bays. Spread out in front of us were a collection of what at first sight looked like old Mahogany radiogram cabinets. Each of these rather odd-looking and randomly parked objects contained an IVC videotape transport under a smoked Perspex lid, and were in fact acoustic enclosures.
While we were chatting we could see that some of these machines were run-up and 'rocking and rolling' as one of the editors worked away behind one of the 'watertight' doors. As far as we could see, mostly everything seemed to have been exclusively Decca in origin. The equipment racks contained in-house processing, time code and editing electronics. I had assumed that bits of PC lurked inside, but this equipment was designed before IBM processors were fast enough to do the job, and were actually Decca designed computers based on AMD 68000 processors.
The editing rooms themselves were quite small, and were perhaps some 10 or 12 feet square. They were lined with dark-wood effect acoustic treatment. Roughly in the middle of the room was a control desk on which was placed a computer monitor, a PC type keyboard, a commercial video-editing controller, and a basic Decca digital audio mixer.
Digital peak level metering was done via the computer monitor, though for making analogue copy tapes for artist's approval a pair of old fashioned PPM's were used. We were told that Decca actually invented the Peak Programme Meter and not the BBC.
All the sound mixing and editing was done in the digital domain. Each of these desks used dither to remove quantising distortion, and had two channels of digital equalisation. They also had a programmable notch filter that could synthesise up to 3 notches of varying width and depth. This was used to 'remove anything from Hum to TV line frequency whistle'.
About four feet in front of the desk and some six feet apart, were a pair of stand-mounted B&W 801 monitors. They were set facing inwards towards the operator and were driven by a large HH Mosfet power amplifier (500 Watts?). I have to say that the sound quality could have been better, it was loud enough though. But this monitoring system was not for balance or ultimate quality evaluation, just for making sure that the edits were good and there were no obvious problems with the recording.
The computer monitor displayed the music's 'waveform' together with a time cursor. Standard stuff now but this was all their own. Decca were the first users of video editing software for audio post-production and they developed their own ways of doing this which were years ahead of anyone else.
As with video-editing, the various edit points are stored in the editing computer’s memory and can be changed before actually making the edit. With this equipment, the edits which are in fact very rapid cross fades between two playback machines onto one recording machine are made with high accuracy, far more precise than with a razor blade (that everyone else used).
This was all done before PCs were fast enough so they built their own editing computers and memories. Ram was also used to store a limited amount of the digitised audio itself. This was to speed up the editing process while the mechanical tape machines caught up. This was far quicker in use than Sony's U-Matic based commercial digital audio system.
It was quite a revelation to realise the amount of editing done to something as seemingly 'pure' as a classical music recording. It was not at all a question of just setting up some mikes in the right places and hitting the record button. For perhaps several dozen and possibly hundreds of edits were carried out on each work.
‘The sound was always that at the performance but the musical accuracy was much better. This results in a play back free from the annoying defects, which become recognisable points of anticipation after a few playbacks.'
While loudspeakers were used most of the time during the editing process, the operator actually checked the quality and integrity of each 'splice' with a pair of headphones.
Almost all of the Decca system was of their own design and manufacture, at first with their own soldering irons and 'Veroboard', though later outside subcontractors were used. But most of the final assembly and software programming remained in house.
Mr Griffiths advised me that Denon in Japan and Tom Stockhom in America (Soundstream?) also had developed early digital audio systems. But Denon's system was 12 bit (again using video recorders), and Mr Stockhom's used a multi-channel instrumentation recorder. Stockhom also attempted recording audio direct to a mainframe computer, but this apparently took 'most of the night to upload'.
Decca were well aware of these other systems, but their recording teams said "If it won't do what we normally do, then we won't use it".
During our first visit we also saw a couple of other mixing suites that contained more advanced digital facilities, including EQ and digital time alignment for microphones (another Decca first). One lone special digital mixing desk sitting in solitary splendour was also fitted with advanced 'endless belt' type digital faders. Their 8 channel mixers used 64 bit arithmetic, with dedicated and home brewed TTL processing for high speed calculation. (This was in the days of the Cray-1which used similar circuitry.) Both DG and Phillips were supplied with versions of this mixer for their own mastering systems. Recording memory for the various editing suites was provided by the various IVC transports laying around in the central post-production machine area as described.
Sadly this was all to end, and in December 1997 this facility was closed down. Ten of perhaps thirty five digital recorders shipped were to Polygram in Holland for archive transcription purposes. The rest were scrapped or found their way to one or two new homes (and I have Polygram's paid invoice!). It had been decided to use normal commercially sourced equipment in the future and not to continue with development or use of the Decca equipment. This also included their next generation of a newly developed magneto-optical disc system as well (see image and info below). In fact some ex employees had already departed to set up Gennex who manufactured the GX series of optical recorders, which were built upon Decca technology.
Some further images:
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-2.jpg
20 bit A-D convertor with cover removed.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-3.jpg
18 bit D-A convertor with cover removed. The D-A board is on the left and the large black rectangle is the Audio Analog hybrid D-A circuit. Next is a 'Decca' PLL board and next to that an input selector switch board, and on the Right is a switched mode PSU. (The front mounted phono sockets are my addition.)
And to conclude I thought it was worth going to the trouble of presenting here an account of what little I know about this minor footnote of British audio history. Decca kept their cards close to their chests because they thought patenting was not cost effective, and I suppose some mystery did not harm 'the brand'. But their undoubted significant technical achievement should not be overshadowed by the more well known commercial products from Sony et all. Simply put, Decca were there doing the job with a better system long before all those other guys.
What might have been... This (below) is probably the only surviving example of a new optical transport that Decca were working on when they were shut down. It uses a Maxoptic 5 Inch magneto-optical disc to record four audio channels and has a much improved user interface. This was to be used with a laptop PC and a separate set of digital convertors via a MADI interface. Unfortunately this example doesn't appear to be operational.
http://i161.photobucket.com/albums/t229/VPR5/IMAG0000-4.jpg
Most of this was an edited version of the Decca section of my web site, and now I need a large :cool: