In the 1980s and early '90s, with Yamaha's help, Korg expanded dramatically, producing some of the first affordable digital recorders and physical-modelling instruments. But it was their world-class synths, such as the M1 and Wavestation, that made them the company they are today...
At the end of last month's article, we left Keio Electronic in a strange position. By the mid-'80s, they had carved a significant niche for themselves, and won the respect of keyboard players everywhere, without ever becoming a market leader. A handful of products — most notably the PolySix and Poly 800 — had sold well, but it wasn't likely that you'd see these on Top Of The Pops or on stage at Reading Festival. What's more, the company had not attained permanent financial stability, and its proprietor, Tsutomu Katoh, had injected his own money as development capital on numerous occasions.
So in 1985, Katoh embarked upon a series of strategies and developments that would see his company rise to become a pivotal player in the synthesizer and keyboard markets. In this, the second of our three-part history of the company now known as Korg, we chart the evolution of some of the most successful electronic products ever to grace the stage or studio.
Major Product Launches
- DDD1 digital drums.
- DSS1 61-note sampling synth.
- DVP1 digital voice processor.
- KMS30 synchroniser.
- MPK130 MIDI pedalboard.
- SG1 76-note digital piano.
- SQ8 sequencer.
- DRV1000 digital reverb.
- SDD1000 digital delay.
- SDD1200 dual digital delay.
- SDD2000 sampling digital delay.
- SDD3300 triple digital delay.
However, before looking at the products themselves, we should first turn our attention to the corporate developments that were taking place within Keio Electronic at this time. To begin with, there was the creation of a permanent offshoot of the Japanese company in America. The roots of this lie in the early 1960s, with the assimilation of ACA (the Amplifier Corporation of America) by a company called Unicord. This led to the manufacture of a series of valve amplifiers released under the Univox name. In 1967, Unicord merged with a guitar importer, Merson, and was subsequently purchased by an oil company, Gulf & Western, adopting the memorable name 'Merson Musical Products, A Division of Unicord Incorporated, A Gulf + Western Systems Company'. Around this time, the company began its association with Keio Electronic, eventually importing synthesizers, rhythm machines and effects units, and rebranding them as Univox products for the US market.
After separating from Gulf + Western in 1975, Unicord decided that the Korg name (which, at that time, was merely the brand name under which Keio Electronic sold its synths and organs) had more value than the Univox name, and stopped rebadging them, becoming instead the importer/distributor for the Korg brand in the USA. This continued until 1985, when Unicord was purchased by Keio Electronic and subsequently renamed Korg USA.
Secondly, Keio Electronic realised that their Korg brand was far better known than the company's actual name, and changed their corporate name to Korg Inc, thus giving the company and their products a single, focused corporate identity.
As far as the products went, the big news of 1996 was the introduction of the DSS1 sampling synthesizer. This marked Korg's entry into the field, but nobody could have expected that it would be another 13 years before the company would release another keyboard sampler. At the time, all the omens looked good... the DSS1 offered sampling, additive synthesis, and waveform drawing, just as the phenomenally expensive Fairlights had done a few years earlier. Each of its eight voices consisted of two oscillators that could draw upon a memory of 16 such samples or waves, and a full range of looping, truncation, mixing, reversing and linking tools were provided, together with keyboard splits and multisampling capabilities.
Although it's now fashionable to deride the DSS1, its oscillator sync of sampled data, variable sample rate output, powerful analogue filters and twin delay lines made it one of the most interesting of all mid-'80s keyboards. Unfortunately, it was not a commercial success. The same was true of 1986's other innovative product, the DVP1 Voice Processor. A combination of a digital vocoder, a pitch-shifter, a primitive harmoniser and a DWGS-based vocal synth that produced a pleasing range of 'ooh' and 'ahh' patches, this was a very flexible 2U rackmount module. Sure, it was restricted by limited polyphony and occasionally poor fidelity, but it was nonetheless an unusual product, and one that has never since been emulated.
Korg also made big strides in 1986 with their range of digital effects units. With a complete rack series that included the DRV1000 digital reverb, GR1 gated reverb, SDD1000 digital delay, SDD1200 dual digital delay, SDD2000 sampling digital delay, and the SDD3300 triple digital delay (some of which are now considered minor classics) it appeared that Korg were well on the way to becoming a force in professional effects technology. Other innovative products included the DDD1 dynamic digital drums and the tiny SQ8 MIDI sequencer. But on the other hand, who now remembers the AT12 auto-chromatic tuner, the CPG01, CPK01 and CPS01 chord processors, the DTM12 tuner/metronome, the MP100 music programmer, and the RP100 rhythm programmer?
Major Product Launches
- 707 performing synth.
- DDD5 dynamic digital drum module.
- DRM1 digital rhythm module.
- DS8 61-note FM synth.
- DSM1 digital sampler module.
- PSS60 programmable super section.
- SQD8 MIDI sequencer.
- Concert 2500 76-note piano.
- Concert 3500 88-note piano.
- Concert 5000 88-note piano.
- Concert 7000/7100 88-note piano.
- DP2000C FM 76-note piano.
- DP3000C FM 88-note piano.
- DP80 FM 76-note piano.
- SG1D 88-note piano.
- DRV2000 digital reverb.
- DRV3000 dual multi-effects/reverb unit.
- SDD1200 dual digital delay.
Busy though 1986 had been, the following year saw even more product launches that extended the size and breadth of Korg's range way beyond anything that had come before. But again, it was behind-the-scenes activity that, in retrospect, has proved to be most significant...
An increasing collaboration with Yamaha was evidenced by the appearance of the DS8 polyphonic synth and its little brother, the 707. Although Korg's marketing described the DS8's digital oscillators as "a musical potion that only Korg knows how to brew", these instruments were, in fact, thinly disguised FM synths, designed using Yamaha components and built under licence. This is not as strange as it sounds: Katoh's relationship with Yamaha had flourished since Keio Electronic had built rhythm units for Yamaha organs in the 1960s, and it appears that Yamaha had owned a small stake in Keio/Korg for many years.
However, in 1987, the relationship took another huge step forward when Yamaha bought a controlling interest in Korg Inc, effectively making it a subsidiary. Apparently, the arrangement was entirely amicable, with the exact details drawn up by Katoh himself. One of the aspects of the agreement was that both companies remained separate entities, free to develop independently, and to compete head-on in the marketplace.
Of course, the capital injection from the share purchase came too late to influence the products launched that year, so Korg's new range retained much of the flavour of 1986. There was the DDD5, a revised version of the DDD1, and the DSM1, an expanded rackmount version of the DSS1 keyboard. This doubled the number of voices, increased the RAM, increased the sample resolution, and added multitimbrality, but like its predecessor, it was not a success, being described in print as "a pig to operate and a lemon to sell". Then there were the DRM1 digital rhythm module, a 1U rackmount that offered card slots as well as a 5000-note drum sequencer, and the PSS60 Programmable Super Section, a long-forgotten combination of rhythm machine, composition tool, and traditional auto-accompaniment machine. None of these was ever further developed, and all died early deaths.
So perhaps the year's most significant releases were digital pianos... of which there were no fewer than three series. Most famous was the SG1D sampling grand, an update of 1986's SG1, a 76-note piano that had singularly failed to take the world by storm. Despite offering just 12-note polyphony, just 12-bit sample resolution, and just four sounds plus a card slot, the 88-note SG1D briefly became one of the keyboards of choice for serious players. The same cannot be said of the domestically styled DP3000C, DP2000C, or the DP80. With FM-generated voices that included organ and strings in addition to pianos, these lacked the realism and expression that had become the standard from other manufacturers. The third range combined sampled sounds with domestic packages that included stereo amplifiers and speakers. Not a huge commercial success, the Concert pianos were interesting because their circuit boards were wholly made for Korg by... you guessed it... Yamaha.
In 1988, the idea of combining sound generation, percussion, and sequencing in a single MIDI instrument was not new, but the M1 brought the concept up to date in a package that cost a tiny fraction of the asking price for the earlier Fairlights and Synclaviers. Nevertheless, something else made the M1 special: Korg had crammed its 4MB ROM full of looped 16-bit PCMs from which programmers could fashion hitherto unimagined sounds. In addition, the digital effects section dramatically extended the range of effects available on any previous keyboard, adding phasing, flanging, delay, distortion, overdrive, excitation and numerous reverbs to the delays and choruses featured on earlier instruments.
The M1 excelled in other areas, too. For example, Korg had ensured that, by using standard synth concepts such as envelopes, LFOs, and filters, the M1's AI (Advanced Integrated) operating system was simple to understand and edit. It was also expandable. There was a dedicated slot for PCM cards, while a second accepted ROM and RAM cards. But perhaps the most significant leap forward was the variety of ways in which players could allocate its 16 voices across its five-octave velocity- and aftertouch- sensitive keyboard. Three different modes of operation provided up to eight simultaneous patches that you could layer, split, crossfade, velocity-fade, or drive using up to eight independent MIDI channels. Common enough today, this was sensational stuff 14 years ago. If there was a shortcoming, it was the lack of filter resonance, which was to limit all Korg's AI-based keyboards, modules and workstation for the ensuing seven years.
Major Product Launches
- M1 61-note music workstation.
- O3 'Symphony' orchestral sound module.
- P3 piano module.
- Z3 & ZD3 FM guitar synth and driver.
Throughout the mid-'80s there had been little sign that, before the end of the decade, Korg would come to dominate the keyboard market. Indeed, the company's own corporate history on www.korg.com has a huge hole, describing no products launched between 1983 and 1988. But in the summer of 1988, Korg announced the keyboard that was to change music synthesis and production forever. It was the Korg M1.
To understand the significance of the M1, you have to appreciate the history of music technology. The Minimoog brought monophonic analogue synthesis to the masses in 1970. Seven years later, the Prophet 5 did the same for polyphony and then, in 1983, the Yamaha DX7 established the domination of digital synthesis. The DX7 reigned for four years before Roland rewrote the rules with the D50. This was the first S+S (sample & synthesis) instrument, was bi-timbral, and was the first keyboard to incorporate a digital reverb section, immediately raising everybody's expectations regarding the 'finished' quality of sounds produced by a professional keyboard.
Each major innovation was occurring more quickly than the previous one, but it surprised almost everyone when, just one year after the appearance of the D50, Korg launched the M1. Following the Roland far too soon to be a copy, this was sample-based, boasted two digital effects units, was 16-part multitimbral, and incorporated a useable multitrack sequencer. Furthermore, thanks to their adoption of VLSI (very large-scale implementation) and surfacemount circuit-design technologies, Korg were able to sell the M1 for less than £2000 and, in doing so, defined everyone's concept of the keyboard workstation. Inevitably, other manufacturers copied the formula, and the M1 soon faced stiff competition. Nevertheless, it was Korg that set the standard. Innovative, simple to use, and with a remarkable range of sounds, the M1 gave birth to a dynasty that still rules today.
Two of Korg's other new products — the P3 piano module and O3 Symphony module — showed a distinct resemblance to the M1. Indeed, the Q1 sequencer and S1 drum machine, sampler, and sequencer even used the same aluminium extrusion as the M1. However, you're unlikely ever to see either of these; the Q1 flopped horribly, and the S1 did not progress beyond its prototypes, being axed before it ever reached the market.
Elsewhere, Korg expanded their domestic piano range with 76-note instruments such as the Concert 600 and Concert 800, plus a full-size model, the Concert 6000. Nicely styled and built, these were never going to be more than a footnote in the company's history.
The final products of note that year were the Z3 guitar synth and the ZD3 guitar synth driver that accompanied it. Children of the marriage between Korg and Yamaha's FM technology, these proved to be Korg's only dalliance with pitch-to-MIDI conversion. The Korg system was quite good for the time, and it's likely that — had Korg redesigned the product to incorporate the M1's sound generator — it would have been a huge success. But the Z3 arrived just a few months too soon, and the FM engine, while adequate, was never going to persuade potential purchasers to rush out to buy it.
Major Product Launches
- A3 signal processor.
- M1R rackmount workstation.
- M1R EX rackmount workstation.
- M3R rackmount synth.
- T1 88-note workstation.
- T2 76-note workstation.
- T3 61-note workstation.
- 'New' SGX1D 88-note digital piano.
As already noted, the M1 proved not to be an end in itself, but the progenitor of a wide range of Korg keyboards and modules. There was the T3, a 61-note synth with an expanded sample ROM; the T2, a version of the T3 with a 76-note semi-weighted keyboard; and the T1, an 88-note piano-action keyboard that looked and felt like a modern, sampled piano. All of these offered bigger displays, disk drives as standard, more sequencer memory and improved MIDI capabilities.
At the other end of the scale, the M1R was an M1 in a 2U rackmount case, while the M1R EX was a T3 in a rack. Least expensive was the M3R, a 1U rackmount that was, in essence, half the M1's sound generator. But even this wasn't the limit to Korg's use of AI technology, because the A3 rackmount multi-effects unit used the AI effects processors without, of course, the AI sound generators.
There were to be many additions and upgrades for the M and T series. These included program and PCM cards, the RE1 remote control for the M3R, ROM upgrades for the M1, M1R and T3, and additional boards that enabled the larger T-series keyboards to sample sounds and use these alongside (or in place of) the internal PCMs. But, again, it was for action behind the scenes that we should remember 1989...
Throughout the 1980s, Dave Smith and John Bowen were renowned as the men behind Sequential Circuits, the men who had designed and voiced the Prophet 5, who had contributed to the design and adoption of MIDI, and were in no small way responsible for major innovations such as multitimbrality and vector synthesis. But despite their justified fame, their company was never financially stable and faced continual financial difficulties.
In 1988, Sequential Circuits folded and, in a cruel twist of fate (FM synths were in no small way responsible for the demise of many synth companies) Yamaha bought the company, re-establishing Smith and his team as an R&D division called DSD Inc. Unfortunately, despite successful progeny in the shape of the SY22 and SY35 vector synthesis keyboards and modules, the marriage was not a happy one so, less than a year later, Yamaha closed the division.
Apparently, Yamaha told Katoh in advance that they were to close DSD and, never one to miss an opportunity, Katoh approached Smith and his team and persuaded them to work for Korg. In May 1989, he established a new company in Silicon Valley. Smith was appointed Vice President and, with a dozen or so staff, Korg's R&D started work on yet another variation of vector synthesis...
When professional keyboard players are asked which single instrument they would have on their desert island, the answer is never "a Wavestation". When the same players are asked which they would choose for the second, the answer is often "a Wavestation". There's a good reason for this. The centrepiece of most keyboard rigs is a workstation capable of all manner of composition and performance duties. Lacking a sequencer, this can never be the Wavestation's forté. But when it comes to spicing up a track, the Wavestation is the double chocolate chip, the pecan and maple, the rum and raisin of digital synths.
Nonetheless, Wavestations were not without their faults, with a number of bugs and omissions in the operating system that remained uncorrected throughout the product's life. There was even a Wavestation 'virus' which was transmitted on one of Korg's own patch cards. This caused all manner of faults, and the only way to eliminate it was to reinitialise the affected synth.
But perhaps the most severe limitation was one that many owners did not even notice. When you switch on a Wavestation, it offers a Performance comprising up to eight Patches played through the internal effects units. Most players never bothered to delve any deeper than this, but if they had done, they would have discovered that the sounds of individual Patches played multitimbrally were not on a par with those of the Performances. This is because the Wavestation's effects are fundamental to the quality and nature of its sounds. With just two effects units, there are simply not enough to go round if you try to use the instrument multitimbrally.
Despite these problems, Wavestations remain powerful and popular sound sources. Like other classic synths, what they do, they do best.
Major Product Launches
- Wavestation 61-note polysynth.
- S3 rhythm workstation module.
The fruits of Korg R&D's work appeared the following year. Yet again employing vector synthesis as a primary synthesis method, the Wavestation used PCMs as oscillators, allowing users to layer samples or to morph smoothly from one to the next. But the Wavestation's real coup was its ability to string PCMs together to create 'wave sequences'. The result was a synthesizer of extraordinary depth and warmth, which produced evolving pads that remain unsurpassed well over a decade later.
Stepping beyond John Bowen's original ideas regarding wave sequences, Korg's voicing programmers saw their potential for generating all manner of rhythmic patches. Thus were the Wavestation's archetypal (and now clichéd) sequences born. Despite the instrument's lack of piano and drum samples, many of these sequences proved to be almost complete backing tracks, playable at the pitch of your choice by pressing a single key. This made the Wavestation very popular in the then-emerging electronic dance genres because it required little manual dexterity in order to produce rhythmically complex music.
Soon, the combination of vector synthesis and wave sequencing attracted a wide range of players, from the hip-hop and house artists who concentrated on its 'groove' capabilities, to almost every electronic and New Age composer, many of whom used it to create the endless washes of sound that characterise so much of their output. Unfortunately, few people took the time or trouble to program the synth. However, you can't blame them for this... the Wavestation's editing system was nothing if not confusing, and given its wealth of superb presets, many players elected to use the factory sounds without modification.
Elsewhere, the S3 was the descendent of the aborted S1 rhythm workstation, and looked to be another milestone for Korg. It embodied a number of new ideas, most radical of which was the separation of the attack and decay components of the drums. Korg called these the 'head' and 'shell' PCMs, and you could recombine these as you wished to create new, impossible instruments. Another innovation was the combination of pattern-based tracks and linear tracks within a single sequencer. A third was the addition of dual effects processors in a drum machine, and a fourth was the inclusion of timecode capabilities. Strangely, the whole turned out to be rather less than the sum of the parts, and the S3 was not a success.
The company also continued to launch effects units, tuners, and digital pianos. The last of these included such long-forgotten entities as the domestic Concert 30, 40 and 50 series, whose only claim to fame is yet another sound-generation acronym, MARS, the Multiplexed Analysis and Re-Synthesis engine (which, in truth, simply meant velocity crossfading of AI piano samples!)
Major Product Launches
- 01/W 61-note workstation.
- 01/W FD 61-note workstation.
- 01/RW rackmount workstation.
- Wavestation EX 61-note polysynth.
- Wavestation A/D rackmount polysynth/signal processor.
- A1 multi-effects processor.
- A2 multi-effects processor.
- A5 multi-effects unit.
- A5 Guitar multi-effects unit.
- A5 Bass multi-effects unit.
With sales that would eventually reach 250,000 units, the M1 was still selling well three years after its introduction. Nonetheless, Korg's designers had not been sitting on their laurels. In 1991, without discontinuing the M1, the company launched its successor, the 01/W, which soon became a hit product in its own right.
Apocrypha has it that the name was an accident: the original intention was to call the synth the M10, but a Japanese graphic artist who did not read English characters attached the image rotated through 180 degrees. Whatever the truth (or not) of this, the 01/W was a little more than just a redesigned M1, incorporating an enhanced sound generator called AI2 (AI squared) that replaced the M1's three processors with a single chip, thus reducing manufacturing costs and improving long-term reliability.
In many ways AI2 was identical with AI, merely doubling the polyphony, doubling the number of tracks in the sequencer, and adding more PCMs and effects. It still lacked filter resonance, but offered a new feature called 'waveshaping', which allowed programmers to distort the waveforms of the basic samples, thus adding harmonics and creating new timbres. In truth, most experiments with the 60 waveshaping options yielded horribly distorted and unuseable results, but a handful of sounds benefited from the high frequencies introduced by the process.
On the Wavestation front, Korg bowed to the inevitable by adding piano and drum samples, plus a handful of new effects algorithms, and called the result the Wavestation EX. Like its predecessor, this was moderately successful, but was rarely a first-choice instrument, perhaps because there was no 76-note version. Then came the Wavestation A/D, a 2U rackmount version of the EX with the added benefit of analogue inputs which allowed you access to the synth engine, making it a powerful vocoder and signal processor.
Another noteworthy family released in 1991 is of no more than passing interest today, but the A5 and its siblings demonstrated that Korg had no intention of walking away from the effects market. Sure, these were sophisticated stompboxes rather than studio-quality products, but they were to be the progenitors of numerous A and AX units that would appear over the next few years. At the same time, Korg continued to develop studio effects, and released the powerful A1 multi-effects processor. At £1200 at UK prices, this cost three times the asking price of a typical budget unit, but its huge range of algorithms, seven simultaneous effects, and professional I/O ensured its place in discerning 19-inch racks. The lower-cost A2, like the A3 before it, used the M1's effects chips, and is no more than a footnote.
Soundlink comprised three major units. The 5U processor contained the brains of the system and all the I/O. There was no shortage of the latter; it included eight channels of balanced, analogue I/O, stereo analogue outputs, digital AES-EBU inputs, assignable AES-EBU outputs, auxiliaries, and even niceties such as a dedicated metronome output. In keeping with its high-end status, there were also synchronisation and control interfaces including LTC/SMPTE, composite video, VITC, RS422 and a Sony nine-pin VCR remote control. Oh yes... and dual MIDI In/Out/Thru. This was not as 'low-end' as it might seem. Soundlink was perhaps the first professional recorder to offer MTC synchronisation.
The second unit was the Digital Audio Recorder itself, another 5U rack containing the hard disk drive, an Exabyte backup unit, and two slots for additional SCSI devices. Recording quality was standard for the era; 16-bit, with sampling frequencies of 44.1kHz and 48kHz.
The third unit was the beautifully designed, dedicated control surface (below). For almost the first time, this integrated control over digital recording with an automated digital mixer, with large level indicators complemented by a backlit LCD, the information on which could also be displayed on a standard CRT monitor. The unit came with its own hydraulic stand, which even offered a sliding tray for a PC keyboard.
The system's capabilities were first-class in 1992, setting new standards for 'low-cost' recording and post-production systems. There was three-band EQ, gating, compression/limiting, and a powerful reverb processor, all with snapshot automation that offered up to 400 cues per session. The MIDI sequencer would have looked familiar to anyone with a Korg keyboard workstation, and offered 50,000 events at up to 96ppqn. But Soundlink's greatest strength was that it was designed specifically for the job and, unlike a 1992 PC and a bunch of miscellaneous outboard boxes, it was fast and efficient in use. Its weakness was that some functions fell halfway between those of a top-notch system and those of a cheaper system; it was neither the best, nor the most affordable.
In the end, Soundlink did little more than demonstrate that Korg were capable of stepping way beyond the boundaries of synthesis and effects units. Surely, the meagre sales never recouped the design costs, but perhaps that's not the point.
Major Product Launches
- 01/W Pro and 01/W Pro X workstations.
- 03R/W synth module. Soundlink digital audio workstation.
- Wavestation SR synth module.
For four years, Korg had dominated the keyboard market, and they used the next 12 months to consolidate their product line. The 76-note 01/W Pro and the 88-note 01/W Pro X joined the 01/W, replacing the ageing T2 and T1 respectively. And, as the M1 had spawned the M3R, the 01/W had the 03R/W, a 1U rackmount incorporating half the AI2 engine of its big brothers.
But even after four years of remarkable success in the musical instrument industry, Korg remained largely unknown in pro-audio circles. So it came as a surprise when, in 1992, the company launched Soundlink, an eight-track digital audio recorder and automated digital mixer at a UK price of £20,000 plus VAT. Though it may seem absurdly expensive now, at the time, this was relatively competitive and indeed, the Soundlink could compete against products costing far more than this. However, this was an arena in which Korg had no experience. With no track record (no pun intended), no distribution channel, and no brand recognition, going in at the top was doomed to failure. Soundlink flopped horribly, selling no more than a handful of units worldwide. When asked how such a mistake could have been made, Rob Castle, the Managing Director of Korg UK, said, "You have to see Soundlink as an experiment. Korg are not known for thinking in straight lines, but for having the courage to try something new. In retrospect, Korg Japan should have employed someone to develop this market properly. We did so in the UK, taking on a chap named Mark Lawrenson, but Korg's worldwide distributors didn't take Soundlink seriously enough. They didn't understand that you couldn't sell a serious DAW through synth dealers."
The year's other launch was to be the final version of the Wavestation, the Wavestation SR. Housed in a 1U rack, this was the Wavestation for players who chose not to program their synths. There were two reasons for this. Firstly, the SR offered far more memories than any other Wavestation, and for many players there was a suitable patch for every occasion. Secondly, by cramming everything into a single rack space, Korg made the SR the least programmable of the family. You really needed a computer-based editor to make any headway — so most players didn't bother.
In retrospect, it was a sad day when Korg discontinued development of this technology. The company had never fully explored the potential for its vector synthesis and wave-sequencing systems, and had there been 76- and 88- note Wavestations with improved polyphony and multitimbral effects assignment, they might have sold in even greater numbers. But there weren't. Shame.
Major Product Launches
- A4 guitar effects unit.
- A4 bass guitar effects unit.
- i3 61-note interactive music workstation.
- AG10 Audio Gallery GM module.
- 05R/W synth module.
- X3 61-note workstation.
- X3R workstation module.
Korg's next products continued to build on the successes of AI and AI2. Clearly, the company had established a winning formula, and it was loath to mess with it too much.
The X3 and X3R both used an expanded AI2 engine, and other than a GM soundset, they offered little that was new. In fact, the X3 lost the 01/W's waveshaping, replacing it with a parameter called 'colour'. This was a rather ineffectual enhancer designed to emulate filter resonance, which failed. The Xs also lost the Yamaha-built keyboards of the M and 0 series, substituting a Fatar unit with a much lighter and less satisfying action. Sure, the number of multisamples was up from 255 to 340, but the X3 was for many players just a cheap replacement for the M1 and 01/W. There was also a baby module, the 05R/W, which took the X3's innards, threw away the sequencer and half the patches, and added a direct serial connection for Mac or PC.
The company used the same sound engine for the AG10 Audio Gallery, a General MIDI module aimed specifically at the games and 'desktop' music markets. At first sight, it seemed good value, with a bundled sequencer, MIDI player and (on the Mac) an editor. But with no physical controls, no memories, and without the usual Korg effects section, it had little appeal for serious MIDI musicians. Perversely, it was perhaps too complex for the people at whom it was aimed. It was not a huge success, and Korg didn't pursue the idea further.
Far more interesting (and far more successful) was the i3, which proved to be the first in a huge series of Korg 'interactive' products. Until the i3, it's probably fair to say that auto-accompaniment keyboards were designed primarily for home use — the modern equivalent of granddad playing the home organ at the annual family get-together. At a stroke, the i3 changed that. Its core was an AI2 engine identical to the X3's, which made it a perfectly useable 'pro' keyboard. It also retained the multitrack MIDI sequencer of Korg's top instruments, but added Styles and Arrangements that allowed players to use it as a band-in-a-box or compositional tool. The only thing it lacked was speakers... Korg's management had decided that internal speakers were just too naff for a serious instrument.
At a time when whole companies were tainted by an association with auto-accompaniment keyboards, the i3 was a very courageous move, but it paid off. Coming in at the top of the market, it was still capable of the usual, horrid backbeats, but its driving rock and dance rhythms were a revelation. Once again, a Korg keyboard succeeded because of the quality of its factory voicing.
But maybe the year's biggest news was something that was never announced, except perhaps to the company's shareholders. Thanks to the products developed using the funds from Yamaha's cash injection in 1987, the previous five years had been very successful, and Tsutomu Katoh now had some cash at his disposal. In fact, he had enough to buy out the majority of Yamaha's share in Korg. So he did.
We can trace the birth of the Wavedrum back to the Sequential design team working at Korg R&D in California. Around 1990, a software engineer named Steve O'Connell had been working on physical-modelling algorithms, and had created a program called Drag, a digital modular synth that allowed you to connect building blocks in much the same way as today's software modular synths.
Given the technology of the time, Drag ran very slowly, and calculated its sounds in much slower than real time. So O'Connell started work on a new version which he called SynthKit, and which he programmed to run on Macintosh computers. This, too, would have run much slower than real time, had it not been for the introduction of DSP boards that fitted the Mac's expansion slots.
Now able to hear the results of their tweaks as they programmed and played, the Korg team began development of a keyboard called OASYS (the Open Architecture Synthesis System) that would employ SynthKit algorithms to generate sounds in real time. But the first instrument to take advantage of Korg's physical modelling was not a keyboard at all... it was the Wavedrum.
Many musicians seemed to think that the Wavedrum was merely a more expensive way to obtain sounds similar to those produced by PCM-based drum pads, but it was far more than this. It could produce all manner of sounds, ranging from traditional percussion through to overdriven lead guitars. The Koto patch, for example, was a true 'string' model, with control over pluck position, string damping, plucking noise, and more. Unfortunately, Korg never managed to convince enough players of the potential of the Wavedrum, and few, if any, plumbed its depths. What a waste!
Major Product Launches
- AX305 guitar effects processor.
- AX30G guitar effects processor.
- G-series processors.
- i2 76-note interactive music workstation.
- X2 76-note workstation.
- X5 61-note polysynth.
1994 was the last year of the second era of Korg. This is because one of the products launched that year was to presage a whole new epoch of synthesis. But before considering this, let's dispense with a bunch of product launches that were merely refinements of existing technology.
The X2 hardly deserves mention, being a 76-note version of the X3. More interesting, if only for its price tag, was the X5, another synth based on the X3 engine, but without the latter's floppy drive and sequencer. Then there was the i2, a stretched i3 with a 76-note keyboard. As for effects, the AX processors were enhanced versions of the previous year's A series.
Far more interesting were the Toneworks or G-series effects processors. Built in Taiwan under licence, these were not marketed worldwide as true Korg products, because Korg USA wanted to differentiate between the company's mainstream releases and its guitar-oriented sidelines.
If the Americans had thought that the G series was going to be an embarrassment, they were mistaken. The G1 distortion unit, G2 acoustic guitar processor, G3 multi-effects unit and G5 synth-bass processor were all well received by their intended markets. Yet the G4 Rotary Speaker Simulator was of most interest to keyboard players. This offered all the facilities of a valve-amplified Leslie cabinet, controlled by an algorithm called IPE — the Integrated Parameter Editing system — which linked the speeds and acceleration of the rotors. Despite the G4's low cost and diminutive appearance, it produced superb results, making any organ-like patch sound as though it had emerged from the guts of a gritty, dusty old C3.
However, as already mentioned, it was neither the synths nor the low-cost effects that were the year's big news. This came at the other end of the market...
For the previous decade or so, academics had been investigating a new method of synthesis called Physical Modelling, which used digital signal processing to recreate the responses and nuances of real (or imagined) musical instruments. As the price of DSP technology fell, it was only a matter of time before someone launched a commercial synth which used modelling as its primary sound engine, so it was no great surprise when Korg announced that they were about to do so. What was a surprise, however, was the nature of the instrument. Described at the time as looking like a Knightsbridge toilet seat, it was the Wavedrum.
In keeping with so many of Korg's off-the-wall developments, the Wavedrum was (to quote one engineer within the company) "the quintessential happy accident". Apparently, one of the senior engineers in Japan had an interest in ethnic percussion, and was keen to apply new ideas to percussion synthesis. Only a company as entrepreneurial as Korg would have given him the time and facilities to develop his ideas, but — at first — it seemed that the gamble would pay off. The Wavedrum was very well received wherever it was demonstrated, and the magazine reviews were uniformly favourable. Everyone was excited by the way that percussionists could play a huge range of sounds, and change them depending upon where — and how hard — they hit the drum skin mounted on top of the unit. You needed the RE1 Remote Controller to get the best from the sound generator, but nobody seemed to object to that.
However, the market for the Wavedrum was tiny. Furthermore, the cost of a single Wavedrum was equivalent to that of a complete PCM-based MIDI drum kit, and approximately three times that of a budget drum kit (including the stands and cymbals). Consequently, Korg sold very few. Had a complete kit been assembled, complete with pads for kick, snare and tom-toms, the story might have been different, but it appears that the technology was not viable or cost-effective enough in 1994 to do this. Once again, Korg proved that something was possible, broke the ground, and then walked away, leaving other manufacturers to exploit the market.
Rose Morris were a British company with a long history. Established in the 1920s, they imported harmonicas in the '30s, built gramophones, banjos and drums, imported accordions and, after the Second World War, imported EKO guitars and GEM organs. In the '60s, the company began exporting Marshall amps, and opened their first London store in 1967. In the '70s, when Marshall took over their own distribution, Rose Morris replaced Marshall by buying Vox.
By the 1980s, the company was very successful, with stores in Central London, Harrow, Hemel Hempstead, Hull and York. They also boasted UK distribution rights for Korg, Takamine, Jupiter, Ludwig and Ovation, and owned two product lines of their own, Vox and Berg Larsen, a mouthpiece manufacturer.
However, events took a severe downturn in the late '80s and, by the end of 1991, Rose Morris were losing hundreds of thousands of pounds per year. Things couldn't continue in this fashion and, in 1992, Korg agreed to acquire a majority stake in the company.
The takeover was followed by an immediate rationalisation. The Vox factory was closed, and Marshall took over manufacture of the brand. Berg Larsen were sold, most of the distribution lines were dropped, the London store was sold to IMP (the printed music company), and the other stores were either sold or closed.
By this time, Rose Morris had lost much of their market credibility, so Korg changed the company's name to Korg UK, and they narrowed their focus to the UK distribution of Korg, Takamine and Jupiter.
In 1995, Korg UK moved from London to Milton Keynes, a move that coincided with their return to trading profitability. However, outstanding debts and leases meant that they had to wait until 2001 before they could retain any profit and contribute to the parent company in Japan. Korg Europe now co-exists in the same building as Korg UK, providing the link between Japan and the European distributors, while Vox R&D was recently established in Milton Keynes to develop future Vox products. In Europe, Korg Italy is a separate development and manufacturing company responsible for many of Korg's digital pianos, as well as the i series.
So we reach the end of 1994 with Korg pre-eminent in the market they had almost single-handedly defined, that of the synth workstation. Elsewhere, the company had a foothold in the field of professional effects units, and were beginning to make inroads into the market for low-cost effects and stompboxes. If there was one area in which they had failed, it was in that of digital recording. Soundlink had disappeared off the map, and nothing had replaced it.
But big things were happening. Other manufacturers — most notably Roland — were developing products to attack Korg's leadership in the workstation market, and Yamaha's 1994 launch of the VL1 and VL7 — the first commercial keyboards to take advantage of physical modelling — was to prove a turning point in the history of synthesis.
It was time for Korg to move on, end its reliance on AI and AI2, and announce something new. Unknown to the public, the groundwork for this had been laid by OASYS and SynthKit (see the box on the Wavedrum, left) and, although they would never be released commercially in their original forms, these — and other developments — were to provide the basis for the third era of Korg. But once again, that's a story for Part 3...
Sound On Sound would like to thank Junko Fukai of Korg Japan and Paul Bundock of Korg UK for their kind help in sourcing pictures for this article.