Kawai's latest synthesiser combines samples with sounds created using additive synthesis. Simon Trask investigates what could be a new dimension in synthesiser technology.
With their K3 and K5 synths, Kawai have begun to make a name for themselves among "serious" keyboard players. Now their K1 synth looks set to take them into the big league.
WHEREAS THE K5 made the "mechanics" of additive synthesis available to the musician, with the K1 Kawai have adopted a different approach to the creation of sounds using additive synthesis. This time around the company have provided 256 internal ROM-based sounds: 204 have been created using additive synthesis of the first 128 harmonics (these Kawai call VM sounds, standing for Variable Memory) while the remaining 52 have been PCM-sampled. Sadly these are displayed numerically rather than by name in the K1's LCD, so you have to keep referring to a booklet provided by Kawai to see what the sounds are. Basically, it's easier to remember where a particular sound is located if you can associate it with a name.
Kawai have divided the VM sounds into five groups, namely a basic wave group and four frequency groups; low, mid, high-mid and high. The basic wave group consists of 40 variations on familiar synth waveforms, while the other four groups break down into brass, string, piano/electric piano, guitar/bass, wind/lead and bell/percussion/organ categories. Not all of these are intended to be used in their own right; some function best as "colouring" for other sounds.
The 52 PCM samples are divided into one-shot, loop, omnibus loop, reverse and alternate categories. Here you'll find a modest collection of drum sounds lifted from the company's R100 and R50 drum machines, together with, among other things, sampled guitar attacks and harmonics, bowed string attack, pizzicato strings, shakuhachi attack and loop, pan flute attack and loop, voice loop and white noise. The omnibus loops group several related samples into looping sequences. So, unlike the D50, the K1's samples aren't just attack samples; moreover, the attack samples themselves are of varying lengths. As to quality, the K1's sounds are fairly clean if not sparkling (if you get my Dreft); bearing in mind the origin of the drum sounds, I'd say we're talking 12-bit here.
THE K1 HAS two performance modes, with associated patches: Single and Multi. Onboard the synth are 64 Single patches (organised as two banks of 4X8) and 32 Multi patches (4X8 again). Plugging in a RAM card effectively doubles the synth's memory capacity, giving you instant access to a total of 128 Single patches and 64 Multi patches.
You can chain up to eight patches (Single or Multi, internal or external) into what Kawai call a Link, and then step through the chain in either direction using the Link +/- buttons on the synth's front panel (K1 and K1M). A useful bonus feature, if a bit half-heartedly implemented.
The K1 allows you to combine up to four of its 256 available waveforms (known as Sources) into a Single patch. You can choose two- or four-Source operation for each patch, making the K1 16-note or eight-note polyphonic respectively. Programming is done through four Groups: Common, Frequency, Wave and Envelope.
These have dedicated squidgy buttons, and you access individual parameters within each group by repeated pressings of the relevant button; in addition, pressing the Stick button allows you to reverse through the parameters.
Kawai have provided the usual +/- buttons for adjusting parameter values, while the joystick functions as a data entry slider.
"Kawai have concentrated on giving the K1 a thorough complement of performance features which can be used to shape the sounds in a very musical way."
The Common parameters, as you might guess, are global settings for each patch. Here you choose two/four Sources, and define vibrato depth, speed and shape (sawtooth or square); aftertouch control of vibrato; mod wheel control of vibrato; auto-bend depth, time, velocity response and keyboard scaling response (good for pitch glides); aftertouch control of frequency; pitch-bend range; keyboard scaling curve; and poly mode (which essentially allows you to select poly or solo voicing).
The Frequency group allows you to define coarse (±24 semitones in semitone steps) and fine (±50, a semitone either way) frequencies for each Source, along with key tracking on/off (together with a fixed frequency key number, C4-G6, or Off), vibrato and auto-bend on/off, aftertouch control of frequency on/off, and the degree of effect that the keyboard scaling curve chosen under Common will have on frequency (±50). While the last parameter doesn't give you the precise control over frequency that Yamaha's microtonal tuning does, you can quickly set up some weird tuning effects which might just set you off on new creative tangents (alternatively they might just give you a headache).
Wave is where you get to select the Waveforms for the four Sources, together with amplitude modulation if required (1:2 or 2:1, 3:4 or 4:3). Additionally you can copy complete frequency, wave and envelope data from any one Source to any other Source in the same or another Single patch - so now you can quickly insert a tuba sound into a piano patch or a shakuhachi sound into a fretless bass patch. Seriously, though, it can be a great time-saving feature.
Underneath the innocuous-looking Envelope button lurk parameters governing volume level, DADSR, velocity curve (1-8), the effect of velocity on envelope level and attack time, the effect of aftertouch on envelope level, and the effect of key scaling on overall level and on attack and decay times. All of these effects can be inverted, so that for instance a hard keystrike can generate a quiet sound (in this way you can set up velocity crossfades between Sources). Incidentally, the initial D of the DADSR envelope allows you to delay the attack of each Source by up to 12 seconds - a feature which can be put to good use creating pitch sequences and echo effects.
THE D50 SHOWED what an impact a well-programmed collection of sounds can have, and it's a lesson which hasn't been lost on Kawai. The internal sounds which come with the K1 and the first set of extra sounds available on RAM card show the instrument off to its very best advantage.
What is impressive about the K1 is the sheer variety of sounds it can come up with. It can produce sharp, bright percussive sounds such as 'Brite EP' (IA6), 'Vibe' (IB4) and 'HardMallet' (IB5), and smooth warm sounds such as 'String Pad' (IA4), 'StringOrch' (EB1) and 'VeloString' (eB1). Similarly it can produce a good range of, say, bass sounds, from the snappy 'DigiBass' (ID1) to the warm, rounded tones of 'Ac Bass' (ID2).
Kawai's synth has a nice line in "acoustic piano" sounds, such as 'Piano 1' (IC1) which effectively captures the top-end hammer strike, and 'Elec Grand' (EB7) which has a particularly strong bottom end. They won't oust today's digital pianos, but they are convincing. Organ sounds are well catered for, from the clean-edged, slightly Leslied 'Jazz Organ' (IC4) to the ominous church organ sound of 'Phantom!' (EC2) to the growling 'Red Onions' (eC2), a pretty good attempt to recreate the sound of Booker T and the MGs' 'Green Onions'.
Acoustic guitar sounds come off well, sounding warm and rounded or sharp and incisive as required, successfully avoiding "plinkiness"; '6 String' (IA3), 'Backin' Guitar' (IC8) and '12 String' (iA3) are particularly effective.
Atmospheric sounds are very much the K1's forte: 'Peaceful' (EA1), 'MysteryAir' (eA3), 'Breath' (EA6) and 'Water Drama' (ED2) all sound like the programmer wished he'd written the soundtrack to Bladerunner. Another standout atmospheric patch is the gloomy, scary 'Poltergeis' (eD2), which'll have you reaching for a crucifix. At the same time there are some very good glassy, shimmery sounds, such as 'Shimmer' (eA1) and 'Glassy' (eD5). Nothing if not truthful, these names. The K1 also has its share of fashionably breathy sounds - such as 'Voice Ahhh', which Kawai have sensibly placed in the pole position (IAI). If I had to pinpoint weaker areas, I'd say that the woodwind sounds qualify (at least as evinced here), as do (to some extent) the brass sounds. Reservations? Well, the absence of filtering means it's not easy to tweak a sound to make it brighter or darker, though of course this task could feasibly be assigned to external EQing.
Overall the K1 manages to achieve a very successful balance of natural and synthetic sounds, providing you with a broad vocabulary of sounds to draw on. Yes, it is a digital synth, but it's capable of producing a wider variety of sounds than that label tends to suggest. I'd say it has a "translucent" quality which sets it apart from analogue synths you've known and loved, and that it has not an analogue warmth, but a digital warmth - a "translucent warmth", if you can imagine such a thing.
"The K1 manages to achieve a very successful balance of natural and synthetic sounds, providing you with a broad vocabulary of sounds to draw on."
MIDI IMPLEMENTATION HAS been a strong point of all Kawai's hi-tech instruments, and the K1 isn't about to let the side down. You can set separate MIDI transmit and receive channels (1-16) and independently enable/disable the transmission and reception of aftertouch, pitch-bend, modulation, hold pedal, velocity and patch change data. Additionally the reception of volume and SysEx data can be enabled or disabled (the K1 has no means of transmitting volume changes, unfortunately).
SysEx communication allows patches to be saved and loaded individually or as a bulk dump (single or multi, internal or external). Patches can be transferred in either direction without handshaking, but for anyone wanting to write editor/librarian software there are appropriate handshaking commands available for automating the whole process.
The K1 sports an extremely flexible Multi mode implementation, with up to 64 Multi patches available at any given moment (32 internal, 32 on cartridge). A Multi patch has eight Sections, each of which can be assigned a Single patch together with low/high note limits (C2-G8), velocity-switching response (soft, loud or all), polyphony (0-8 or variable), transposition (±24 semitones in semitone steps), fine-tuning (±50, a semitone either way), volume level (0-100) and pan position (L, R or L+R). Additionally, each Section can be set to receive on its own MIDI channel (and different Sections can be assigned to the same MIDI channel, for split and layer textures), but the K1 still transmits on a single channel. While data reception on/off states apply equally to all Sections in a Multi patch, each Section is completely independent in its response to MIDI performance data. This means you can warble your choir, sustain your strings and bend your trumpet all at the same time.
Unfortunately the increasingly sophisticated multitimbral implementations of today's synths isn't being matched by a corresponding increase in polyphony.
Dynamic voice allocation (as used on the K1) can give the impression of increased polyphony by assigning voices to musical parts as and when they're required (on the principle that not every "instrument" will be playing constantly). But this voice-stealing can cause its own problems - where, for instance, you have a prominent bassline that is suddenly silenced by voice-stealing at just the wrong moment. Kawai have thought of this one, and allow you to assign a fixed polyphony to any Section up to a total of eight voices (Roland do the same on the MT32).
Additionally the K1 is clever enough to dynamically vary its polyphony according to the combination of two- and four-Source sounds in use at any given moment. But none of this gets around the fundamental problem: not enough voices (particularly if you're also layering Sections). The long-term solution may well be for manufacturers to increase the polyphony of their instruments, but in the meantime you can always link up two instruments to give double the polyphony. For this to work, the instruments must have an Overflow mode (so called because it only transmits notes via MIDI when an instrument's voice capacity is at full stretch). Given the K1's multitimbral power, Kawai's omission of such a mode on their new synth is a real disappointment.
To match the various performance modes on the K1, Kawai have provided their synth with three alternative ways of responding to patch changes. Norm means that patch changes 0-63 will call up a Single patch while patch changes 64-127 will call up a Multi patch. You should choose Sect if you want to change patches in individual Sections of a Multi patch (patch changes 0-63) as well as the Multi patches themselves (patch changes 64-127).
Finally, Link allows incoming patch changes to call up patches in the K1's Link chain of eight patches.
It's important to realise that the K1's Multi mode isn't exclusively a MIDI feature. Each of the eight Sections in a Multi patch can be assigned to MIDI, Keyboard or Mix (MIDI and Keyboard). By selecting either of the latter two, you can organise up to eight sounds at once on the keyboard. The K1's low and high note-limit settings and velocity switching allow you to define an impressively wide variety of keyboard textures, which can of course be used in conjunction with sounds triggered remotely from a MIDI sequencer. In fact many of the 32 Multi patches which come programmed into the K1 are "extended" Single patches rather than multitimbral configurations for sequencing.
THE K1 IS destined to put Kawai in the synth big-league. Its extremely competent handling of a wide range of sounds coupled with musical responsiveness and ease of programming make Kawai's latest synth a rewarding instrument to play. And its flexible handling of multimbrality means that it should find a place in many a sequencing setup.
Don't hesitate to investigate. The K1 is already a serious contender for best synth of the year.
Prices K1 £595, K1M £395; DC8 RAM card (which will come with sounds programmed into it for no extra charge) £30-40 (tentative); all prices include VAT
Gear in this article:
Review by Simon Trask
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!