The development of musical instruments has been influenced from the earliest times by the evolution of audio technology. Stone Age man had access to the knowledge and skills required to create a range of instruments from simple wind and percussion to early reeds. Many examples exist of fife-like instruments carved from hollow bones and contemporary non-western cultures exhibit instances of the use of naturally occurring objects such as shells or horns being used as sound generators.

The majority of instruments fall into a relatively small number of individual categories:

Percussion: Pitched or unpitched the percussion instrument is struck, usually by means of another object but, in many instances by the player’s hand. Such instruments are characterised by their resonant qualities which range from the relatively unpitched sound of a tensioned skin or other membrane to the exactly poised output of a sonorous metal.

Simple wind: Based on a length of tubing which  determines the pitch generated. Such instruments, in primitive forms, are capable of only a limited range and probably derive from chance voicings of hollow objects such as animal horns, hollow bones etc. They were later augmented by:

Complex wind: Such instruments use vibrating reeds or similar devices to create their pitch which is then consolidated and amplified by a structure based upon the simple wind instruments. Even early examples of both wind categories show an awareness that tube length determines output pitch and many ancient instruments, in both classes show drillings reminiscent of the holes in a contemporary recorder, clearly intended as a means of selecting pitch.

Plucked string: A later development was the use of a tensioned string or similar medium induced to vibrate by mechanical displacement using the fingers or some form of plectrum. Given that the i vibrating medium would normally have been organic in origin (eg some sort of gut or sinew), it is not surprising that few examples have survived intact. Variants thereof may be found in the hammered strings of the dulcimer and piano families.

Bowed string: Perhaps the most recent to emerge, we may assume this category to have been the result of a chance discovery whereby certain tensionable but texturally rough materials (eg stretched gut) could create vibrations when moved across an appropriate sounding medium such as a string or equivalent. Apart from the obvious examples of the orchestral string family, variants exist such as the hurdy-gurdy where the strings are mechanically bowed by a wheel of bone or similar material.

Prior to the introduction of electricity to the art of musical sound generation, the above categories or variations and combinations thereof remained the entire repertoire of musical sound generation. Their pitching and tonal qualities were determined by relatively simple physical parameters and their output amplitude was, in general, a simple function of the input energy of the player.

Perhaps one of the earliest examples of the application of technology to music making was the organ. In its simplest form, it is known to have been in use by the Romans who knew it as the hydraulis, an instrument known to be in use as early as 757 AD when one was offered as a gift by the Byzantine emperor Constantinus Copronymus although some sources claim its existence as early as 75 BC. The organ has continued its development ever since with accounts of its presence in Christian churches throughout the Middle Ages although the earliest documented organ music dates from the 15th century.

A major factor in the appeal of the organ was its ability to produce, by harmonic combination, a multiplicity of timbres, later referred to as “stops” These were created by careful combination of pipes, each contributing its characteristic tonal qualities - in effect an early form of additive synthesis. These became named for their resemblances: flute, string and even the grandly titled Vox Humana. Organs achieved considerable sonic and mechanical complexity, pushing hard at the mechanical limits of their times and even laying the ground for later technologies by utilising difference tones (to be found many years later in the Theremin, and, later still, in the frequency modulation synthesis techniques propounded by John Chowning and most famously incarnated in Yamaha’s DX-7 synthesiser).

The organ was, however, by no means the only instrument to be based on the technologies of its time. The harpsichord and its successor, the piano were essentially the high-tech instruments of their time. The former began in the 14th century as little more than a mechanically plucked harp but, by the 17th century, had evolved into a sophisticated instrument in its own right , taking a significant place as an orchestral instrument until the second half of the 18th century. The piano has its origins in the work of Bartholomeo Cristofori which first saw the light of day in around 1710 when he produced an instrument known as a “gravicembali col pian e forte” (a harpsichord capable of touch sensitivity). This was achieved by striking the strings with a mechanically actuated hammer rather than plucking them with a similarly activated quill as in the harpsichord.

The modern piano owes much to the original work of Cristofori although it was substantially improved by the later contributions of Silbermann, Stein, Kirkman, Zumpe and Broadwood who, among other modifications, created the escapement mechanism that permits repetition of a note while a key remains depressed, the pedal system and the now traditional grand and upright shapes. The piano made its British debut as a solo instrument, played by JC Bach in 1768.

The design of the upright piano was finally patented by John Hawkins in 1800, a later development being the addition of a recording and playback mechanism utilising a punched paper roll similar in principle to an early computer tape. Known as a pianola or player piano, this mechanism typically uses pneumatic actuators triggered by the data contained on the player roll to actuate the impact of the hammers onto the strings. It is through the agency of the pianola (in a sense a precursor of the digital sequencer) that it is still possible to witness the performances of pianists whose demise preceded the emergence of sound recording technology. A further parallel may be seen in the work of American composer Conlon Nancarrow who uses the pianola medium to create otherwise unplayable (usually by virtue of impossible tempi) piano music. Such options have only become available by other means subsequent upon the evolution of the sequencer.

There can be little surprise that electricity had a massive impact upon music making. Its impact on the creation and use of instruments, perhaps surprisingly, predates the invention of audio recording systems.A number of early instruments vie for the ancestral title but the best known is the Harmonic Telegraph developed in 1874 by Elisha Gray (otherwise best known for competing with Thomas Edison for the title of inventor of the telephone).

Gray discovered (apparently by accident) that an electromagnetic switch would produce an audible note if subjected to sufficiently frequent interruptions of current. Furthermore, Gray was able to employ a feedback circuit to create these interruptions in a self-sustaining manner, their frequency being determined by the inertia of the switch itself as well as that of the system in general. Not only did Gray create a playable instrument but he also went on to develop an early version of what is now known as the loudspeaker. The former, known as the “Harmonic Telegraph”, used a full complement of his electromechanical oscillators to create an instrument with a two-octave range. Of greater long-term significance, however was Gray’s later invention of a tone wheel pitch generator, the precursor of the Hammond organ.

Gray’s Harmonic Telegraph was, in practice, little more than a music-hall curiosity. It would be hard, however to imagine a greater such curiosity than William Duddell and his Singing Arc. In 1900, the incandescent bulb (another Edison invention) was insufficiently powerful for use as a streetlight. The chosen source was far brighter: the electric arc struck by a high voltage between two carbon electrodes. An undesirable characteristic of these arc lights was a whining sound which Duddell sought to eradicate. His researches failed to eliminate this tone but did uncover a way to control it by varying the supplied voltage: another precursor but this time that of the voltage controlled synthesis techniques applied by Robert Moog in the 1960s. So successful was Duddell in evolving this technique that, with the addition of a suitable keyboard, he created a musical instrument upon which he was able to perform to music-hall audiences.

During the succeeding years, a number of electromechanical musical instruments appeared, based on principles ranging from tonewheels to optical disks and neon lamps. None was grander in scale than the Telharmonium. Invented by Thaddeus Cahill and patented in 1896 it boasted control over virtually all the variable parameters of a modern synthesiser: pitch, dynamics, additive synthesis, controlling keyboard and even loudspeaker. The Telharmonium was a machine on a grand scale. Described as “a church organ mated with a weaving loom”, it stood two stories high, sixty feet wide and weighed some two hundred tons. A sub-scale prototype was produced in 1900 and the full - size version first appeared in 1906. Each note was created by an individual dynamo mounted upon one of twelve axles, each rotating at such a speed as to create the desired note.

The performance and propagation medium of the Telharmonium was the recently developed American telephone network which suffered damage as a result of its over-generous output and the machine lapsed into obscurity. Once more, though, the fundamental principles of tone generation established by Elisha Gray had proved their worth.

Not only were the mechanical principles of the Telharmonium regarded with respect but its industrial aspect and sheer scale won the approval of Italian futurist Ferruccio Busoni who, together with painter Luigi Russolo and others launched the “Art of Noises” manifesto. Russolo went on to create a series of “intonarumori” (noise intoners) and later used a keyboard and loudspeakers to create the “Russolophone”, an industrial noise-generator/instrument. The Intonarumori were not designed as “musical instruments” in the conventional sense: rather they sought to recreate the sounds of industry and urban life of the period as part of the Futurist attempt to extend the available palette of sound textures that could be used by the composer following the now-famous “Art of Noises” manifesto of 1913. Notably supported by the notable composer Edgar Varese, a range of instruments were built such as the “scoppiatore” (exploder), “ronznatore”, “crepitatore”, “stropicciatore”, “ululatore” and “sibilatore”, mostly named for the quality of sound that they produced. Following a small number of performances up to 1929, the intonarumori remained unused until destroyed during World War 2.

As is often the case, a physical principle established in an earlier era was utilised in the next generation of technology-based instruments. The first truly electronic instrument, the Theremin appeared in Russia in 1920 based upon the vacuum tube electronics developed onwards from 1910 by Lee de Forest in America. More significantly, however, it employed an electronic manifestation of the difference tone principle first used by organ designers. By then know as heterodyning, this used the principle of addition of two closely pitched but dissimilar tones to create a subsequent tone whose frequency equalled the difference between the two original tones.

This approach was achievable in the Theremin by virtue of the fact that early vacuum tube oscillators naturally operated at frequencies above the audio range with the result that the difference tones created by their interaction were potentially within audible frequencies (albeit only with the intercession of a medium wave radio receiver to convert its output to audio). The particularly significant approach adopted by Leo Theremin in his 1928 design was to fix the frequency of one oscillator and vary the other, thus creating a variable difference pitch. Pitch control was governed by proximity to a suitable aerial and, in many versions, amplitude was controlled in similar fashion via a second aerial. This unconventional interface allowed the Theremin to be used in early interactive multimedia productions: an early 1930s system known as a “Terpsitone” allowed the activities of dancers directly to control the instrument.

Despite its greatly restricted timbral qualities and clumsy control interface (a footswitch was used to short circuit the loudspeakers in order to mute the note), the ethereal musical saw-like sound of the Theremin ensured its continued if somewhat intermittent use as a sound effect until the late 1960s and may have contributed to its recent revival by Robert Moog. The somewhat indeterminate pitching of the Theremin limited its usefulness with other instruments and it was not until 1926 that an electronic device that could be regarded as a musical instrument in the conventional sense made its debut.

The work of French musician Maurice Martenot led to the development of an instrument which used the same valve-based generating principle as the Theremin but replaced its control interface with a more conventional keyboard of up to six octaves in addition to a forward-looking additional device: a ribbon controller. This latter device allowed glissandi and vibrato modulations whereas the main keyboard not only determined the pitch of the note to be generated but, in some measure, controlled its envelope. A further set of controllers permitted timbral changes. This degree of controllability and flexibility facilitated the adoption of the Ondes Martenot as a “serious” orchestral resource: a first for electronic instruments. Notwithstanding the relative flexibility of the Ondes Martenot when compared to its predecessors, it remained a restricted tone-generating device by comparison to the later synthesisers of Moog et al.

The first device to embody their open-ended approach to synthesis (ie to provide fully variable parameters rather than imitative preset ones ) was the Trautonium developed by Freidrich Trautwein in 1930. At this time, neon tubes had reached a high stage of development in Germany and, for sound generation, were often preferred to valves. As used in the Trautonium, they generated a rich harmonic content which was then modifiable by means of filter circuits. The main controller was not a conventional keyboard but closer in concept to the ribbon controller used in the Ondes Martenot. Attack and volume were separately controlled by a second adjacent linear controller. A later development, the Mixturtrautonium has seen substantial (if little recognised) service in the composition of film soundtracks such as Hitchcock’s “The Birds ” and solid-state versions remain in use to the present time, distinguished by their odd generating principle (addition of subharmonic tones), unconventional user interface (2 wire-covered catgut strings and a contact rail) and unusual tonal quality.

Although both the Ondes Martenot and the Trautonium were considered to be serious attempts to broaden the available range of instrumental sounds in an orchestral setting (an early Concert Trautonium boasted a 100 watt amplifier and speaker system), their demanding playing styles limited their usefulness in this role and their physical robustness and tuning stability further restricted their adoption to the extent that the “Turangalia Symphony“ by Messaien is virtually the only widely known work that uses either instrument (in this case the Ondes Martenot). A further limitation of these instruments was their monophonic capability whereas their non-electronic counterparts had, in many cases, virtually unlimited polyphony.

This shortcoming led to the development in 1933 of the 200 valve Orgue des Ondes which used one oscillator per note and aimed to replicate not only the sound but the user interface of a “conventional” organ. This was the first truly polyphonic electronic instrument but was by no means a synthesiser: it was purely an electronic organ. Perhaps the best known such instrument, although it does not use electronic generation, is the Hammond organ, first produced in 1935.

The Hammond organ uses electromechanical generators known as “tonewheels”, in principle not greatly unlike Cahill’s Telharmonium. Timbral modification and other processing is, however, achieved by purely electronic means with the exception of the almost obligatory “Leslie” speaker in which a rotating vane deflects middle and upper frequencies to create a characteristic pitch and amplitude modulation. As with many electronic instruments of the time, the Hammond was initially designed to simulate a pipe organ, being equipped with organ-type stops and a pedalboard for bass parts. It was not until considerably later that it came to be valued for its own unique qualities, especially when electronically overdriven to create warm distortions of what was otherwise a somewhat “pure” pipe simulation. Due to its electromechanical generating system, the Hammond proved to be extremely heavy and hence difficult to move. As a result, later versions were designed to split into two more manageable parts a factor that contributed in no small measure to its later success with rock bands.

Throughout the inter-war period, a  range of developments continued, some, like the Hammond, becoming mainstream instruments but many others being short-lived or purely experimental. One particular oddity but nonetheless a highly influential device was the linear descendant of a specialised organ designed in 1779 in St. Petersburg by Christian Krazenstein. His instrument used a vibrating reed and a series of resonators to create vowel sounds. This was further developed in Vienna by Wolfgang von Kempelen who created an instrument capable of generating intelligible speech.

Interest in this specialised area lapsed until the invention in 1939 of the Voder, an electronic instrument designed to synthesise speech for experimental purposes. Initially conceived as a research instrument to enhance studies of voice transmission by telephone and radio the Voder shot to public fame at the Berlin Radio Fair of 1935 where the idea of a machine that could speak was greeted with widespread amazement and acclaim. This interest notwithstanding, the Voder remained purely a research tool, due in part to it’s specialised interface the use of which demanded lengthy training and considerable skill. The research (by Werner Meyer-Eppler and others) upon which this machine was based led also to many aspects of the development of what was to become electroacoustic music by providing the ability to analyse and define sounds in terms of electronically controllable functions such as filtration, pitch and amplitude envelopes etc. Ultimately, however, the fate of the Voder rested upon its user interface - a highly unorthodox form of keyboard which required operation by specially trained typists - and, in common with most instruments that used other than a standard musical keyboard or at least conventionally recognisable interface, it failed to gain wide acceptance and remained simply a curiosity.

Many other exotic instruments were created during this period. Some, such as composer Percy Grainger’s Free Music Machine, existed more as theoretical concepts than as working systems but others like Henry Cowell’s Rhythmicon reached operational status and provided a unique resource for composers and performers. Grainger created a grandiose design intended to expand the composition and performance processes. To do this, he interconnected a range of devices including player pianos and electronic instruments but with limited success. In some respects the Free Music Machine (1948) anticipated the synthesiser inasmuch as it used eight oscillators as its sound sources, these being controllable for amplitude, pitch and timbre by means of a graphical interface.

The Rhythmicon was a unique instrument in that it specialised not so much in creating sounds but rhythms, especially those that were conventionally considered to be unplayable. In this respect, it may be regarded as the lineal ancestor of the modern drum machine. Developed in collaboration with Leon Theremin, it used a similar sound generation mechanism (heterodyning) to the earlier instrument but there the resemblance ended. It was equipped with a short keyboard and a complex range of electromechanical rhythm generators which, in their turn, governed the pitch of the keyboard.  First built in 1932, the Rhythmicon was used by Cowell to create only two works after which it was abandoned until its rediscovery by record producer Joe Meek as a result of which it appeared on a number of film scores and popular records.

Electronic sound generation had been in existence for many years before the concept of the synthesiser as it is now known was to evolve. Until this time, electronic instruments had generally sought to replicate the sounds of existing acoustic ones, particularly the organ. Electronic sources and treatments had been used by Stockhausen et al but these had been little more than pieces of modified laboratory equipment and no real attempts (with the possible exception of the Trautonium and the later Bode Melochord) had been made to create an electronic instrument from, as it were, first principles.

The first device to lay claim to the name “synthesiser” was the RCA Mk1 developed during the 1940s and 50s by Harry Olsen and Herbert Belar. This machine, and its better known Mk2 descendant were the incidental products of research aimed at developing a machine that could automatically create pop songs. Much time was spent in the analysis of hit records and popular songs (especially, for some bizarre reason, the work of Stephen Foster). In this respect, the RCA was a complete failure but the sound generating and modifying techniques that it employed laid the basis for much later work. The system was wholly modular and could be reconfigured to produce any quality of output. Most significant of all, it could be programmed externally by means of punched tape - at that time the common medium of human/computer dialogue.

Although initially used for its intended musically banal ends, the RCA Mk2 went on to become the focus of the legendary Columbia-Princeton Electronic Music Center and, in the hands of composers such as Babbit, Ussachevsky and Leuning produced many classic works of early electroacoustic music. Its usefulness notwithstanding, the RCA was a room-sized monster resembling an early computer with teletype and punch tape interfaces. By no stretch of the imagination could it be used as a conventional musical instrument. It was not until 1964 - when the first work of Robert Moog appeared - that the synthesiser became a viable entity.

An electronics engineer by trade, Moog collaborated with musician Herbert Deutsch to create a programmable synthesiser, basing it around the concept of voltage control. The basic idea was that parameters such as oscillator or filter frequency, modulation and amplitude would all be controllable from a common source. While each module would retain its specific function and be capable of independent operation, a DC voltage could be applied which would vary a given parameter in much the same way as turning a knob. By this means, the pitch of an oscillator (or almost any other parameter) could be controlled by varying the control voltage. This voltage could come from any number of sources, most obviously a keyboard or ribbon controller but also from another oscillator (usually of lower frequency) or other source such as an envelope generator. Moog created a range of modules both controllers and audio generators/modifiers. These could be interconnected in the desired configuration by means of patch cords which handled the routing of both audio and control signals.

Apart from oscillators which created the basic sound in a variety of waveforms, a white noise generator was supplied and provision was made for external sounds to be used as sources for the processing modules which included ring modulators, a range of filters, voltage controlled amplifiers (usually controlled by a dedicated envelope generator (sometimes referred to as an ADSR - attack, decay, sustain and release being its main variable parameters).

Additional modules followed including a sequencer which would generate sequential control voltages allowing, for example, arpeggios to be created. Any control voltage source could be used interchangeably with any other and, in like fashion, audio could be routed from any source to any destination. This approach, together with high quality circuit design, created a machine of considerable musical power and enormous sonic flexibility. Unfortunately, these qualities came at a price: DC voltage control is, by definition, analogue and is hence less accurate than its digital equivalents, being subject to all manner of undesirable influences such as temperature, cable length etc with the result that tuning (and other controllable parameters) often became problematic. Furthermore, the instrument was large (although minuscule compared to the RCA machines) and awkward to program as a result of the forest of patch cords that would adorn its front panel. Furthermore it had no memory capability so, once a sound was created, it had to be used there and then since there was no guarantee that it could be recreated at a later time. Worst of all, it was monophonic.

Nonetheless, the Moog was a considerable success although it did not reach public awareness until composer Walter Carlos used it, together with a fairly basic 8 track recording facility, to produce his seminal 1968 album “Switched on Bach “, a meticulously crafted recording of well know Bach works using the Moog as the sole sound source. Carlos’ album created huge public interest being seen as the first example of the use of an electronic instrument in a conventional musical context.

Prior to this, the public perception of electronic instruments had been confined to weird sci-fi noises on film soundtracks or clangorously obscure “serious” electronic music. Rapidly seized on by the rock and pop world as a desperately needed source of new and different sounds, the Moog went from strength to strength. Despite its success, the Moog was an ungainly beast and quite unsuitable for live performance so, with a few notable exceptions, it remained a studio curiosity until the release of the MiniMoog in 1971. This simplified and pre-patched machine was a true performance instrument, quick to set up and easy to use. More than this, it generated genuinely musical sounds of great warmth and character. Despite its simplicity and lack of memory facilities, it remains a favourite instrument to this day: at the time of writing, manufacture of a modernised version has just begun - over 25 years after its first appearance.

Notwithstanding their virtually generic title, the Moog machines were by no means the only voltage controlled synthesisers to be produced from the late 1960s onwards: numerous developers, most notably EMS in England and ARP, Oberheim and Buchla in the USA produced machines using the basic principles adopted by Robert Moog. Some machines were, as the original Moog had been, large modular designs intended to be installed and operated in studios whereas an increasingly large number of others were designed to rival the Mini Moog as live performance instruments. With a few exceptions, however, these latter machines were considered inferior to their Moog equivalents which benefited enormously from the exceptional audio qualities of their filter design, perhaps the most identifiable Moog quality. Virtually all such machines used a similar design concept: a small number of voltage controlled oscillators produced the raw source audio which was then operated upon by modifiers such as filters, envelope shapers, ring modulators etc. These in their turn could be voltage controlled by means of lower frequency oscillators, keyboards, joysticks or other devices.

Central to most such instruments was the ability to interconnect the various modules in a choice of configurations. The most original approach to this was taken by EMS who used matrix panels into which pins were inserted to connect between vertical (source) rows and and horizontal (destination) rows. This approach centralised routing into a single panel thereby avoiding obscuring controls with patch cords. Increasingly, though, the majority of users adopted a fairly standard configuration and hence, with the exception of machines used by “academic” and similar studios, the concept of the infinitely reconfigurable synthesiser was abandoned as flexibility gave way to convenience. Instruments were produced with predetermined signal paths which, since the concept of voice memory was still some way off, made the recall of previously used sounds considerably easier.

As the synthesiser slowly gained acceptance as a “legitimate” instrument, its lack of polyphony became increasingly irksome. Whereas the organ designer’s approach of taking a single frequency and dividing it as required to produce more than one note had been acceptable in previous instruments, the precise nature of the subtractive synthesis then in vogue required (at least in theory) a minimum of one oscillator per note if the range of available timbres was not to be excessively compromised.Certain instruments allowed external (and therefore possibly polyphonic) sound sources to be input to the modifier circuitry and treated thereafter as if generated by the synthesiser itself but, with the notable exception of the EMS VCS3, this approach failed to find favour. Where specific sounds could be identified, purpose designed instruments could be created (eg the ARP String Ensemble) that offered polyphony but at the cost of a limited range of broadly similar preset sounds. The demand for a truly polyphonic synthesiser remained.

The solution to the polyphony problem came as a result of the widespread adoption of integrated circuit technology by the electronics industry. The extreme miniaturisation that this permitted was quickly seized upon as the idea of the “synthesiser-on-a-chip” caught hold. As before, Moog was an early leader in this field, launching the PolyMoog (which featured, in effect, one complete synthesiser for each individual note) in 19xx. The enabling technology for this approach was still young and the control systems remained essentially as before with the twin results of limited timbral possibilities and tuning inaccuracy and thermal drift.

Within a relatively short time, the idea of using digital control whilst retaining analogue generation and processing began to gain considerable support since this approach made it possible to store complex parameter settings using computer memory chips, usually augmented for archival purposes by magnetic media such as tapes, or later, disks. By this means, it was possible for the manufacturer to pre-load an instrument with a wide range of useful sounds, especially those that users would find difficult or unduly time consuming to program for themselves.

Despite the fact that, by this time, a consensus of approach was beginning to form among the major manufacturers, a number of innovative systems continued to appear, most notably involving the use of computer-like concepts of synthesis wherein the source audio was digitally generated (and sometimes processed too). For the first time, the computer display terminal began to be seen in studios where its ability to display the actual waveform of a sound rapidly became invaluable to the programmer. Within this set of systems, two subsets emerged: digital generating followed by conventional analogue processing by means of filters etc.and wholly digital generating and processing

The former approach was used by relatively few manufacturers but nonetheless commanded considerable interest by virtue of the “purity” of the original digital waveforms and the fact that they could be mathematically described.Principal among such instruments was the German PPG Wave and its successors which included to VDU-equipped WaveTerm. Towards the end of this phase of development, it became possible to import waveforms into these systems and subject them to the same processing as those that were internally generated and limited inroads were made into the area of sampling later seized upon by the designers of the Fairlight and other machines.

The latter approach, which in many ways anticipated the contemporary computer-based digital audio workstation was epitomised by the Fairlight CMI (Computer Musical Instrument). First introduced in 1979, this was truly a computer in all senses save that it had a musical keyboard as an (optional) controller. Using two of the newly available microprocessors, it was capable of generating audio based upon waveforms drawn on its display screen using a light pen.

As might be anticipated, sounds thus generated tended to be somewhat dull and mechanical and it was that in a desperate attempt to find a stopgap solution to this problem until a way could be found of generating more interesting sounds that its developers, Peter Vogel and Kim Ryrie, turned to the idea of using digital recordings of “real” instruments as the basic source material. The Fairlight could record and replay actual sound, allowing it to be used as if it were an instrument rather than simply being played back as a recording. In other words, if the key an octave above the original key of the sample was played, the sample itself would play back with a transposition of an octave. Any noise could now become an instrument. Peter Gabriel, whose company, Syco, imported the Fairlight into England was therefore able to credit such “objets trouvées” as exhaust pipes and paving slabs as part of his eclectic instrumental armoury.

Original referred to as the CMI (Computer Musical Instrument), the single-box structure of the Fairlight made MIDI unnecessary and, before long, entire works were being created within the one machine. Unfortunately the audio quality (8 bit) was relatively poor and the high cost of computer memory meant that samples had to be short. You still couldn’t record an entire vocal part on the one machine. The first machine to permit the complete studio-in-a-box approach was the Fairlight’s great rival, the American Synclavier. A relatively huge rack of computing hardware, the Synclavier permitted the complete construction of songs, soundtracks and almost any audio work. Amongst many other things, it became the first real direct-to-disk recorder and in upgraded form, is still in considerable use for film work as well as by composers, most notably the late Frank Zappa who was able to persuade it to carry out musical operations impossible by “normal” means. This, to all intents and purposes, was the beginning of sampling. Not only that, but the Fairlight featured a simple but usable “arrangement page” (known as Page R) that was the mother of all the on-screen sequencers that were to emerge in later years.

As digital systems became ever more widely adopted as a result of the ever-reducing cost of computer power and custom chip production, a number of manufacturers, most notably those from Japan, began to look for new synthesis methods, especially those which could imitate “real” instruments and/or provide new and radically different timbres. A whole range of approaches were developed, most of which relied at least in part on digitally sampled waveforms. The outstanding exception to this approach was that adopted by John Chowning of Stanford University. Chowning discovered what later became known as “FM” (frequency modulation) synthesis by chance when experimenting with computer sound generation. This principle was adopted by Yamaha and, after some years of development, saw the light of day in the form of the legendary DX-7 synthesiser which not only generated sound in realtime by purely digital means but was also controlled by the equally new and radical MIDI protocol.

Despite its continued reliance on oscillators, FM synthesis allowed them to interact in a far more sophisticated manner than hitherto, creating waveforms of great complexity and subtlety. For the first time, synthesisers were able convincingly to reproduce the sounds of bells, tuned percussion and even difficult instruments such as the harmonica. The price to be paid for this sonic excellence was high: the DX-7 boasted probably the most inaccessible user interface of all time with the result that few users bothered to modify the factory preset sounds. Even the development of computer-based screen editors to replace the tiny LCD display of the DX-7 failed to convince many users. The sound of FM, by far and away the most promising of the new generation of synthesis systems, rapidly became a tiresome musical cliche and its clinical qualities convinced many users that synthesis by itself would not provide the answers they sought in terms of instrumental emulation.