Last update: 7 February 2011
The Roland Alpha Juno 1 and 2
The Birth Of The Junos
The Alpha Juno's earliest ancestor was Roland's first true polyphonic synthesiser, the Jupiter 4. Launched in 1978, this was a contemporary of the SCI Prophet 5, the Oberheim OBX, the Yamaha CS80, and the Korg PS-series. Unfortunately, it was no competitor for any of them. For one thing, it offered merely one oscillator per voice, and just 4-note polyphony played from a 4-octave keyboard that offered neither velocity- nor aftertouch- sensitivity. Its chorus/ensemble, sample & hold, unison, and arpeggiator were first class, but these could not disguise its limitations, which ensured that most of its sounds were thin and uninspiring. On the other hand, the little Roland was affordable. OK, £1,800 was far from chicken-feed in the late 1970s, but the Jupiter 4 undercut each of its competitors by around 50%. Consequently, the list of its owners became a "who's who?" of the era, including Kitaro, Gary Numan, Tangerine Dream, Stevie Wonder, Tomita and, perhaps most famously, Nick Rhodes of Duran Duran.
Ultimately, the Jupiter 4's home-organ styling and sonic under-achievement guaranteed that the it would never become a classic. That accolade was reserved for the Jupiter 8 - a top-of-the-range instrument with a price tag to match. Unfortunately, the launch of the Jupiter 8 marked Roland's disappearance from the affordable end of the market. So it was arch-rival Korg that smashed the £1,000 price barrier with the landmark Polysix. This was the first "everyman's" polysynth, it used the same SSM oscillator chips as early Sequential Circuits instruments, and it copied their sleek good looks. It was no wonder that it quickly became known as the "poor man's Prophet". Korg probably believed that the synthesiser world was at its feet. But it wasn't to be. With a stroke of marketing genius, Roland announced that its next instrument was a digital synth that would cost less than £1,000.
To understand why this was such a coup, you must remember that, in 1981, the keyboard world was besotted with unaffordable digital technologies. Instruments such as the Fairlight CMI and Synclavier II promised almost unlimited sonic creativity, but cost more than the houses in which most musicians lived. So everybody held their breath and waited to see what Roland was planning. The answer was... the Juno 6.
When it arrived in 1982, the Juno looked a sure-fire winner. It was sleek and professional, and its beautifully arranged control panel echoed Roland's flagship Jupiter 8 instead of knobbier synths such as the Polysix. It was 6-note polyphonic, sported a high quality keyboard, had well-placed performance controls and a powerful arpeggiator. It also incorporated Roland's now-classic (if somewhat noisy) chorus ensemble. At just £699 it was also excellent value. Roland had hit the bull's-eye.
As we now know, the Juno wasn't a digital instrument at all. It was the first of a breed of 'hybrid' synths that replaced the older voltage controlled oscillators (VCOs) with new, stable, Digitally Controlled Oscillators (DCOs). Despite having the word "digital" in their name, these retained a traditional analogue synth's ability to create 'thick' sounds, principally by allowing you to use a combination of sawtooth, pulse-width modulated, and sub-oscillator waveforms simultaneously. There were limitations... for example, the Juno's filters lacked the powerful 24dB/oct response of the Polysix's filters, and the VCF and VCA shared a single ADSR envelope generator. Nevertheless, it sounded superb. The only thing that it lacked was memories...
Released less than a year later, the Juno 60 was in most ways identical to its predecessor. There were a couple of minor changes, such as the shift from a continuously variable High Pass Filter to one with just four possible values. More important was the addition of DCB, a pre-MIDI interface that allowed Juno 60s, later Jupiter 8s, and Roland's sequencers to communicate with each other. But the biggest change, and the one that justified the extra £300 Roland charged for their new baby, was the addition of 56 patch memories. The Juno had come of age.
Unfortunately, 1983 was also the year that Roland, Sequential and Yamaha launched the world's first MIDI synths. Roland's own JX3P and, in particular, the all-conquering Yamaha DX7, appeared to consign the Juno to the trash-can of yesterdays' heroes. So Roland responded yet again, releasing the Juno 106, a sleek redesign that looks as attractive in 1998 as it did when it first appeared in 1984.
The 106 was significantly different to its older brethren. Firstly, it offered a remarkably advanced MIDI implementation that allowed you to record notes, performance information, and control panel changes. This made it possible to sequence the programming controls themselves - a facility that would be very welcome on many of today's mega-synths. Secondly, the number of patches leapt from 56 to 128, and there was also the introduction of Roland's now-standard left/right/push performance lever for pitch-bend and modulation. Not everything was roses - the 106 lost the 60's arpeggiator and one or two minor programming capabilities - but Roland had again hit the jack-pot. At just £799 the Juno 106 quickly became Roland's most successful product, and remains one of the best-selling synths of all time.
In 1985 Roland still had two more rabbits to pull out of the Juno's hat. The first was the HS-60 'Synth Plus' - a Juno 106 with built-in stereo speakers. Never a success, this suffered from its similarity to a 'home' keyboard. Yet it was in every other way an unmodified Juno 106. Finally, there was Roland's first multi-timbral expander, the MKS7 Super Quartet, which combined a velocity-sensitive 7-voice bass/poly/lead synth (based on the Juno 106 engine) with the drum sounds of a Roland TR707. At under £500, the MKS7 might have been the bargain of the decade except for one failing: you couldn't edit and store sounds, even with a computer.
The Alpha Junos
By this time, Yamaha had consolidated its world domination with a huge range of FM-based synthesisers and modules. Partly because of this, Moog had gone out of business, while Sequential and Oberheim were well on the path to doing so. In contrast, Roland had not only survived, but had managed to prosper by concentrating on producing cheaper, yet desirable, analogue synths. But it was time for a change...
In the winter of 1985 Roland announced two new Junos. These were the 4-octave JU1 (Alpha Juno 1) and the 5-octave JU2 (Alpha Juno 2). These offered many improvements over the original instruments, with several new programming capabilities, backlit screens and, on the JU2, a velocity- and aftertouch- sensitive keyboard. The JU2 even offered a cartridge slot for storing patches. At just £575 and £799 respectively, it appeared that Roland had again hit the bull's-eye.
Strangely, the world appeared unmoved, and the Alphas made very little impact. Why? It couldn't have been a reflection upon their quality because their excellent construction would have graced synths costing much more. Nor could it have been a reaction to a lack of facilities because, although Italian manufacturers Crumar and Siel had already produced cheap touch-sensitive synths, the Rolands had far better MIDI specifications, more voicing capabilities, a slick keyboard action, and better programming systems. Nor should it have been a consequence of the Junos' cosmetics because, with the exception of Roland's so-called "alpha dial" (which replaced the up/down buttons found on most other synths), their top panels imitated the all-conquering Yamahas. The Rolands even allowed you connect the optional PG300 programmer, making every parameter instantly accessible and editable with a dedicated slider or switch. OK, so this added £200 to the price, but a quick glance through the adverts of the time shows that a fully featured knobby polyphonic costing less than £1,000 was still something of a bargain in 1986. So what was the problem? Maybe it was something to do with the voice structure itself...?
The Alpha Junos' voices were very different to those of the Juno 6, 60 and 106. For example, their DCOs were much more sophisticated than those found on any other Roland, including the top-of-the-range JX8P and Super JX-10. They offered fourteen basic waveforms including pulse-width modulation of both the pulse waves and sawtooth, and six sub-oscillators (four pitched 1 octave down, two that were 2 octaves down). Many of these additional waveforms were harmonically complex, and they made possible timbres that sounded more 'digital' than ever before. Nevertheless, the Alphas retained the original Junos' ability to mix a pulse-wave, a sawtooth and a sub-oscillator for lush 'analogue' sounds.
The Alphas also retained traditional analogue VCFs, but these were unable to self-oscillate, so several traditional Juno sounds became impossible, particularly those that used the filter as a second oscillator. OK, the Alphas' filters tracked the keyboard and nearly oscillated, so you could create aggressive sounds by adding noise to the sound and tuning the filter to the keyboard's pitch, but it wasn't the same. On the other hand, no previous Juno had a velocity- or aftertouch- sensitive filter. You lose some...
... but you definitely win some. The Alphas' new envelope generator sported seven parameters rather than four, and the additional controls added a huge range of new sonic possibilities. Other additions included a fully variable chorus (which was more flexible but less rich that the original Junos'), a chord memory function, and four useful 'tone modify' buttons that allowed you to adjust the modulation rate, modulation depth, brightness, and envelope times of a sound without entering the full editing system. There were also three sockets for performance pedals. The first of these was a straightforward sustain pedal. The second offered three functions: polyphonic portamento on/off, chord memory on/off, and the ability to step through eight of the patches in a selected bank. Finally, an expression pedal allowed you to control the overall volume of the synth (but not, as other commentators have suggested, parameters such as modulation depth or pressure sensitivity). It all added up to an excellent package of performance features, especially when you consider that even the JU1 (which lacked touch-sensitivity when played from its own keyboard) could be played across eight octaves and responded to velocity and aftertouch over MIDI.
So maybe it was their sounds that let the Alphas down? Well... no, I don't think so. Both instruments incorporated two patch banks, called Preset and Memory. The 64 Presets (stored in eight sub-banks of eight related sounds) were, as their name implies, held in ROM and therefore inviolate. In contrast, the 64 patches in the "Memory" were held in RAM and you could edit and replace these at will. Both banks included excellent orchestral and string patches (always a Roland speciality), punchy brass ensembles, and several luscious synth-pads. There were also some remarkably usable lead-synth patches, surprisingly good organ impersonations, and some very fat basses. (This should not be too surprising: the SH-101 was, in essence, a monophonic Juno 6.) Furthermore, the Alphas produced a range of bright, percussive and FM-esque sounds quite unobtainable from their predecessors. Of course, they were still restricted by their single oscillator per voice, but this was not as significant a limitation as you might expect. Played through a reasonable effects unit such as a QuadraVerb, the results could be very impressive, indeed.
Roland even packaged their new Juno architecture five ways, addressing at least three different markets with them. Like the HS60 and Juno 106 before them, the functionally identical HS10 echoed the JU1, as did the HS80 the JU2, but these offered built-in amplifiers and speakers in addition to the standard specification. There was also the MKS50 rack-mount module, which boasted 128 RAM patches, 128 performances (which incorporated a patch and its associated performance data), and a further improved MIDI specification. Yet, at just £550, even this didn't sell in large quantities. Some writers have speculated that the Yamaha TX81Z over-shadowed the MKS50 which, if true, proves what a fickle bunch we keyboard players are.
So, what is the answer? Why were the Alphas poorly received, and why haven't they been rediscovered in the thirteen years since they first appeared? The most obvious answer to this rests in their lack of extreme filter resonance and self-oscillation. This has always been a significant factor in a synthesiser's popularity, and there is no doubt that Roland lost the plot when they excluded this from the Alpha Juno's capabilities. Perhaps it was the limited 6-note polyphony. Another possibility lies in the Alphas' lack of endorsees. Fashionable acts such as Bomb The Bass, The Prodigy, Coldcut and Howard Jones all used one version or another, but none of them sung its praises in such a way that players rushed out to buy one.
Ultimately, I suspect that there was no single reason for their lack of success. In 1986, the public had temporarily tired of analogue technology, and it's probably a credit to Roland that the Alphas sold as well as they did. At least the MKS50 is finally becoming desirable, and I suspect that it's time that the JU1 and JU2 started to appreciate. That this hasn't yet happened is to your advantage. You can still pick up a JU1 for around £100, and a JU2 for as little as £175, and I recommend that you do so. They're warm, responsive and, short of encountering a large lump of Kryptonite, they'll carry on for ever... two of the analogue world's undiscovered little gems.
You can find my full review of the Roland Alpha Junos in Sound On Sound magazine, October 1998.
Copyright ©1996, Gordon Reid.