There can be hardly more influential a piece of technology than the magnetic recorder. In its many forms, it supplies a fundamental resource for a huge industry, provides domestic and in-car entertainment, records television images and computer data and even logs flight deck conversations on aircraft. All these diverse activities depend upon the same basic process: the movement of a magnetisable medium in relation to a magnetising unit (in the recording mode) and the generation of an electric current as a result of a reversal of the process (in the playback mode). Like many such inventions, it has proved to have a breadth of application that would probably astound its original inventors and developers.

The basis of magnetic recording can be traced back to the work of Oberlin Smith in 1888 but it was not until 10 years later that Valdemar Poulsen patented a device which he called the Telegraphone. Prior to this a number of magnetic recorders had been developed, the earliest of which used a wire strung across a room. The recorder travelled along the wire, recording as it went. A later and more practical version wrapped wire helically around a drum to create the electromagnetic equivalent of the purely mechanical Edison Phonograph. The more developed Telegraphone used spools of steel wire (later steel tape) as the recording medium and was not initially intended to record sound per se. Rather it was to record the electrical impulses from a telegraph system for storage and retransmission. (It may be remembered that this same motivation led Edison to the development of the phonograph.)

Given that the later development of the telephone used broadly similar technology to the telegraph, it was soon apparent that a machine that could record telegraph signals had the potential to record speech or whatever from a telephone. This realisation led to the first instances of timeshifting and data compression. Since both the telegraph and telephone had become vital to commerce and cheap-rate calls were unavailable during business hours, it became common practice to record non-urgent messages and calls on the office Telegraphone and to then make the outgoing calls at night when lines were less crowded and tariffs lower. The sender would play the recordings back at double speed and the recipient would record them at a similar speed. When slowed back down, the messages could then be retrieved as normal having not only been sent in half the normal time but at a lower than normal charge rate. While adequate in quality for speech, the Telegraphone was unsuitable for music recording, indeed wire recorders remained so notwithstanding the considerable improvements that were made in subsequent years (most notably by Marvin Camras). The search was on to improve sound quality to the point where magnetic recording could compete with disk systems.

In 1924, a machine using 6 mm wide steel tape as the recording medium was developed by Stille in Germany. Although the subject of considerable interest, the machine was not made available until 1931 when it appeared in the UK as the “Blattnerphone”. Intensive development work followed which involved Stille, the Marconi company and the BBC. This led, in 1934, to the appearance of the Marconi-Stille Recorder which used 3mm wide steel tape 0.08 mm thick running at 90 metres per minute (equivalent to just over 60 ips). This required approximately 3 Km of tape to record a 30 minute programme! The massively heavy reels (55lbs) required the assistance of block and tackle to be changed and the machine was regarded as so dangerous that, when it was in motion, personnel were not allowed into the bomb-proof room which housed it since a tape break created thousands of slivers of sharp steel which travelled at considerable velocity.

There is some contention as to the exact chronology of the development of what was to become the first conventional tape recorder. The basis of a new and substantially less hazardous recording medium was laid in 1928 when Fritz Pfleumer of Germany patented a method for depositing magnetic material onto strips of paper or plastic film. His work led in 1932 to collaboration between BASF and AEG with a view to continuing the development of magnetic recording and the first product of this work, the AEG Magnetophon was publicly demonstrated at the Berlin Radio Fair in 1935. Using 1/2 ” tape and with a running time of up to 20 minutes per reel, the Magnetophon was capable of a recording quality vastly superior to any of its predecessors and made its first music recording on November 19 1936 (a concert conducted by Sir Thomas Beecham).
An alternative chronology has it that Pfleumer and AEG initiated an earlier collaboration which produced a working machine in 1931 although it was not until BASF became involved in 1934, that recording tape per se was made available. It is probable, therefore, that the 1931 machine may have used paper-based tape.

Meanwhile, development work on the wire recorder continued. Marvin Camras is generally credited with the discovery of high frequency biasing (although Weber and Von Branmuhl of AEG are sometimes credited) which made truly high quality recording possible for the first time. This curious process uses the superimposition of a signal well above the audio range onto the audio signal itself in order to enhance the recorded signal. Quite how this process worked was not fully understood at the time since the tape medium could not record the bias signal itself but it appeared to act as some sort of electronic catalyst, facilitating the recording process without itself being directly involved.

Together with radio, the Magnetophon was regarded as an essential propaganda tool by the Third Reich and was widely used for the dissemination of speeches by Hitler and others. Hitler, ever one to embrace technologies that could be exploited for propaganda, is quoted as saying “without the loudspeaker, we would never have won Germany”. The Magnetophon also found considerable use in music recording for radio where its clarity of sound and running time made it vastly superior to conventional disks. It was these qualities that drew the Magnetophon to the notice of John Mullin of the US Army Signal Corps who, stationed in England in 1944, was puzzled by high quality night-time transmissions of orchestral music from Germany. These had none of the short durations or surface noise to be expected from disks and showed no sign of being live broadcasts.

Following the invasion of Europe in 1944, Mullin was assigned to study the technological advances made by the Third Reich, with especial regard to electronics. He soon discovered small magnetic recorders (“Tonschreibers” - lit. “sound writers”) that had routinely been used for dictation by field officers but these were of relatively poor sound quality and unsuitable for music recording and reproduction. It was not until July 1945 when Mullin was sent to Frankfurt to investigate a German jamming device designed to stall aircraft engines that he first encountered a Magnetophon. A few miles outside Frankfurt was a wartime radio station that had been taken over by the American Armed Forces Radio Service and, as a result of a chance meeting, Mullin was able to hear the machine used in the night-time broadcasts in operation for the first time:

“When he put the tape on the machine, I really flipped; I couldn’t tell from the sound whether it was live or playback. There was simply no background noise “

The existence of the Magnetophon was unexpected and its significance was largely missed except by Mullin who removed and dismantled four machines (two of them for the US government and two for himself) and a number of reels of tape. At much the same time, another member of the US Army Signal Corps, Richard Ranger, made the same discovery and he too removed several machines, dismantled them and had them shipped back to America. By 1946, both men had reassembled their respective machines and begun to demonstrate them in America. A particularly memorable demonstration was given by Mullin to the Institute of Radio Engineers in San Francisco in May that year attended by, amongst others, Alexander Poniatoff and Harold Lindsay of Ampex, a former aircraft component manufacturer seeking entree into the professional audio market.

One problem remained: tape. A captured Magnetophon had been used in a famously embarrassing incident in 1945 when it was used by Radio Luxembourg to record a speech made by General Eisenhower for broadcast to the people of occupied Germany. Tape was no longer being manufactured and it was necessary to reuse a reel of previously recorded tape. Like modern machines, the Magnetophon was equipped with erase circuitry to permit the reuse of tape but, on the machine in question, the circuitry was defective with the result that the words of Eisenhower were interspersed with a speech made previously by Hitler and recorded on the same tape. When broadcast, this caused enormous confusion and concern that Hitler was not in fact, as had been reported, dead, so much so that Eisenhower promptly ordered that captured tape was no longer to be used in order to avoid any future problems. Sensing nonetheless the importance of the technique, Eisenhower assigned one Major John Orr the task of producing new tape stock. Orr was able to contact the original inventor, Karl Pfleumer and, with his assistance, was able to produce usable tape within a few weeks. Orr went on to found his own tape manufacturing company which became a division of Ampex (later Quantegy) in 1959.

Musician and inventor of the eponymous guitar, Les Paul, recalls that he was approached by Richard Ranger in an attempt to interest him in his rebuilt Magnetophon, now called the Rangertone. Realising its potential, Paul arranged for Ranger to demonstrate the machine to Bing Crosby who was looking for a new recording medium. Crosby, suitably impressed, asked for 50 machines only to be told that Ranger could produce them at a maximum rate of one per year. Meanwhile, in June 1947, Mullin had demonstrated the Magnetophon to Murdo McKenzie of Bing Crosby Enterprises.Together with Mullin’s partner, William Palmer, they approached Poniatoff and Lindsay at Ampex who, in turn, approached Crosby for development funds. Crosby immediately underwrote the work with a loan of $50,000.

The worlds’ most famous crooner may seem at first glance to be an unlikely benefactor for the nascent tape recording industry but, in fact, Crosbys’ career had already been profoundly influenced by a number of technical matters beginning with the development of the condenser microphone in 1926. Singers such as Rudy Vallee had already discovered that the projected vocal style that had been prevalent prior to electric recording technology was no longer necessary and that a quieter, more intimate style was rendered effectively by the new microphones. The condenser microphone contributed additional warmth and helped Crosby to create a unique vocal style, based very much upon the capabilities of the technology. (Crosby was once described by William Paley of CBS as having a “phonogenic” voice).

Later in his career, Crosby became dissatisfied with the limitations of the transcription recording techniques then available. Transcription was, and remains today, a way of life in countries which, like the United States, operate within more than one time zone. A prime time radio show in New York would, if transmitted live, be received on the West Coast long before the end of the working day, missing a substantial part of its intended audience. Transcription recordings were a partial answer and one that, in different media, continues to be used up to the present time. Here, though, the timeshifting is practiced by the broadcaster rather than the recipient.

Crosby was not only unhappy with the audio quality of transcription recordings: his dissatisfaction was supposedly based, at least in part, on his penchant for golf. The requirement that he generate live radio shows on a regular basis allegedly constrained his golfing activities inasmuch as NBC, which broadcast his radio shows, would not permit the use of recordings (other than for transcription purposes) except of advertisements and he was therefore required not only to perform live but to do so in front of an audience. Much of the impetus for these restrictions came from unions, especially ASCAP (American Society of Composers and Performers) which sought to restrict the amount of airtime devoted to recordings although, in fairness, the major concern of the radio networks was based on the inferior audio quality provided by transcription discs

A change of allegiance to the ABC network allowed Crosby to provide a weekly prerecorded 1 hour show (from which he would garner $30,000 per week from the network plus $40,000 generated by selling broadcast rights to independent radio stations. The requirement was that, 2 days prior to transmission, Crosby should provide his show on 3 16″ aluminium disks. These disks, being standard transcription issue, were recorded at 33rpm and played for 10 minutes each. Unfortunately, these were derived, according to Schoenherr, from 4 minute, 78rpm originals, thereby creating serious continuity problems for the presenting broadcaster. Mullin also comments that the process of rerecording so compromised the sound quality that the sponsor (record player and radio manufacturer Philco) threatened at one point to cancel the show.

Other, more generous sources suggest that Crosby sought to improve the quality of his shows by allowing retakes and subsequent editing. In yet another curious parallel (to the later career of the Beatles), Crosby sought to eschew live performance in favour of the more sophisticated presentation offered by recording and subsequent editing. Instrumental in this transition was the post-Magnetophon technology offered by Mullin and Palmer in association with Ampex. Crosby recruited Mullin, his Magnetophons and the few reels of German tape that Mullin had secured from his expedition to Frankfurt.

Schoenherr reports Crosby’s view thus:

“By using tape. we could do a 35 or 45 minute show, then edit it down to the 26 or 27 minutes the program ran. In that way, we could take out jokes, gags or situations that didn’t play well and finish with only the prime meat of the show ; the solid stuff that played big. We could also take out the songs that didn’t sound good. It gave us a chance to try a recording of the songs in the afternoon without an audience, then another one in front of a studio audience. We’d dub the one that came off best into the final transcription. It gave us a chance to adlib as much as we wanted, knowing that excess adlibbing could be sliced from the final product. If I made a mistake in singing a song or in the script, I could have some fun with it, then retain any of the fun that sounded amusing.”

In a 1976 interview, Mullin casts a slightly different light on the proceedings. Referring to the recording of performances on 16″ disks, he says:

“Almost invariably, there was editing to be done. That meant copying some disks onto new ones, making adjustments as they went, maybe substituting a song that had gone better in rehearsal for the final take. Since they recorded everything in rehearsal as well as what took place before the audience, there were plenty of bits and pieces to work with.

Sometimes it was necessary to make what were called “predubs”. Say they wanted to use 3 cuts from 3 different disks, all within a matter of a few seconds. That didn’t allow enough time to get each one cued up during rerecording. So they would make little pretransfers or predubs, making copies until all the cuts were added. The final record, therefore, might be 2 or 3 generations removed from the original. “

This process automatically implied the loss of audio quality that had brought the show close to being dropped by its’ sponsor. In an attempt to rectify the problem, McKenzie retained Mullin and Palmer to operate their tape system alongside the disk one in order to see if it had the potential to overcome the problem of generational degradation. Mullin established his equipment in a small continuity studio and work began in 1947, using 2 recorders and 50 reels of captured German tape.

Tape editing as such did not exist at this time and Mullin was forced to evolve its techniques for himself. He initially sought to use the solvent based cut-and-stick techniques used in film editing but, when splices made thus proved to be painfully audible, he switched to ordinary adhesive tape which, though effective, tended to bleed adhesive, necessitating the liberal use of talcum powder prior to playback. Having dissected and reassembled reels of German tape in order to edit a show together, Mullin was confronted with its lack of consistency from one reel to the next. As a consequence of of the manufacturing technique employed by BASF, the output level would vary between reels which meant that, in the absence of a supply of fresh tape stock, Mullin was constrained to spend the night following a Crosby broadcast dismantling the edit pieces and reassembling them into the reels from which they were originally derived in an attempt to maintain some degree of consistency.

In 1948, Ampex delivered its first Crosby-financed machines and, as a result of Eisenhowers’ initiative, new tape became available in commercial quantities. This allowed Mullin and his associates to continue to develop recording and editing techniques - most notoriously the use of “canned laughter”.

Following his work with Crosby, Mullin went on to work with Ampex on the development of the video recorder in 1952. Unfortunately, the extended bandwidth of TV images required a linear tape speed between 100 and 360 ips (over 20 mph!) and it was not until 1956 when Charles Ginsburg established the helical scan principle used to this day that video tape recorders became a practical option. This principle remains in use, not only for video recording but in the all-but-universal DAT digital audio format. - Ampex and others continued to develop the magnetic tape recorder but, in the same year that they delivered their first machine, the rapidly recovering European audio industry entered the field as a result of the work of Willi Studer in Zurich.

Studer set up in business in 1948 to readapt American recorders to the needs of the European market but quickly realised the potential of a locally designed product. His first machine the “Dynavox” proved successful and was renamed the Revox T26, the first of a long line of products created under the Studer marque. Further models followed, aimed mainly at the semiprofessional market, culminating in the classic A77 of 1967. Meanwhile, Studer had pursued a policy of parallel development of industrial-grade machines, the best known of which - the 1″ 4 track model J37 - appeared in the early 1960s, becoming famous as the multitrack used by the Beatles in their work at Abbey Road. Development work continued, resulting in the uniquely modular A80 recorder which, in various configurations from cassette to 2″ 24 track established itself as the default studio workhorse throughout the 1970s and much of the 1980s with more than 11,000 machines installed worldwide. In later times, Studer maintained its position in the professional market by developing one of the first digital multitrack recorders to use the now-standard DASH (Digital Audio Stationary Head) format. The development of Studer machines paralleled that of Ampex and other manufacturers with a tendency towards ever increasing numbers of tracks and mechanical and electronic sophistication.

As tended to be the case, American manufacturers (and in later years, Japanese ones) led developments with European companies consolidating and refining their work somewhat later. The 8 track recorder first appeared in America in the mid 1950s but was not seen in Europe until the mid/late 1960s. The 8 track Studer appeared even later. This was due in part to few European users having any real idea of how such a machine might be used. It was left to innovative performers and producers such as Les Paul (owner of the first ever 8 track machine) and Tom Dowd of Atlantic Records (who took delivery of the third such machine) to evolve a new way of recording that could exploit its potential to the full.

In general, there has been a tendency to ever-increasing numbers of tracks on a given tape machine. Until the recent advent of digital multitrack recording, these increasing numbers posed an increasing problem in respect of track width and background noise levels. The format of 4 tracks on 1″ tape used by the Studer J37 provided excellent performance with almost 1/4″ of tape width available to each track. While this implied a relatively high level of innate tape noise, bandwidth and dynamic range were excellent and bleeding of information between tracks was minimal. As machines evolved, the packing density of tracks increased. First came 8 tracks on 1″, then 16 tracks on 2″ then 24 tracks on 2″ and even the sub-professional formats that crammed 8 tracks on 1/4″, 16 tracks on 1/2″ and 24 tracks on 1″ tape. With up to 24 tracks active at any one time, the cumulative level of tape and machine noise could very easily become unacceptable.

One solution often adopted by purists was to increase tape speed from the normal 15 ips to 30 ips which improved noise performance and dynamic range but the increased tape costs were often regarded as unacceptable. It was in this context that noise reduction systems were evolved. Many engineers acquired the habit of adding a little extra high frequency boost when recording and neutralising this with a corresponding cut on playback. Most unacceptable noise lay in the higher frequencies and it was reasoned that this approach of pre-emphasis/de-emphasis helped to minimise this problem. It was, however, a crude method that was quickly superseded by the marketing of the proprietary noise reduction system designed by Ray Dolby which rapidly became the industry standard. The Dolby system (now virtually a generic term for any type of noise reduction process) used a mixture of equalisation and dynamics processing both before and after recording and made possible the track bouncing and copying that was so much a feature of many works of the 1970s and 80s before multiple machine synchronisation was routinely available and prior to the advent of degradation-free digital copying in the mid 1980s.

The tape medium diversified into the domestic market with the introduction of 2 competing formats: the 8 track cartridge and the compact cassette. Both sought to make tape acceptable to a mass market by removing the inconvenience of threading reels. Supply and takeup were both enclosed in a plastic container (in fact the 8 track format was endless so only one reel was used. The mechanical complexity of the 8 track mechanism and relatively large cartridge size led to it being eclipsed by the technically inferior cassette which went on to command a substantial market share, competing effectively first with the LP and later with the CD.

Development of the professional tape machine has continued steadily with the introduction of numerous digital formats, mainly derived from helical scan video technology but, in very recent times, the advent of the hard disk recorder is beginning to pose a serious challenge. As processor speeds and disk capacity increase, the advantages of random access systems become more persuasive and it is entirely possible that the majority of professional uses of the tape medium will be superseded by random access technology within a decade or so. Even given the demise of its technology, the underlying principles of the tape recorder have, and will continue to, in many ways, dictate the way in which we work with audio. The best random access systems look and feel operationally similar to tape machines and still use the basic concept of magnetic recording in their hard disk stores: we may assume that it will be a long time before the tape recorder is forgotten.