Kawai Duts its acclaim-winning K-1 synthesiser on the rack. Jez Ford turns the screws.

Kawai knew it was on to a winner when it created the K-1 synthesiser last year. And if it wasn't sure then it found out pretty pronto from the reaction at music shows and out in the real world. The keyboard alone was worth the asking price, full-size velocity sensitive keys and zoning to split the keyboard into eight sections. Add to that the high quality sounds it could produce and it began to look as if Kawai might have stuck the wrong price label on the box. But they hadn't. It cost £600 and it was a gem.

Delighted and determined, Kawai set about to consolidate its sudden grasp of the mid-range synth market. This year it has introduced a stable of K-1 products, perhaps the most useful of which is the K-1r.

This is the rack-mounted version of the K-1, without that handy keyboard but with everything else intact and neatly squeezed into a 1U height unit. Mind you, Kawai has chosen to omit the joystick it used so effectively on the original front panel. Perhaps it was considered a little incongruous for a rack unit. Its omission is a surprise but no disaster.

So then, without the keyboard the K-1r will have to stand up on its sounds alone. Can it hold up against the competition? The answer (shout it from the rooftops) is yes. The sounds are wonderful and the organisation of programming is masterful.

There are two different types of patch on the K-1r, single sounds and multi sounds. Not too surprisingly the multi sounds are made up of lots of single sounds, all mixed up and layered around as you see fit. Not that the single sounds aren't perfectly serviceable on their own. Although the style of synthesis is reasonably straightforward for anyone trained up on an algorithm-type synth, the results far outclass any 4-operator DX synth and are at least equal to 6-operator machines.

Much of the credit for the richness and accuracy of the K-1r's timbres can be attributed directly to the excellent preset waveforms that you can choose to make your sound. There are 204 VM waveforms produced by additive synthesis of the first 128 harmonics. These vary from straight sine and square waves right up to complex bell and vibes shapings. Here we can see how this synthesis method gains power over FM programming. In FM you start out with only the relatively simple waveforms. That can make it difficult to go in the direction you desire unless you're very well versed in the laws of acoustic synthesis, and on the way you use up your operators pretty quickly.

Most of us poor uneducated fools would prefer to leapfrog that stage and start off with a prefabricated waveform as our original operator. This does of course leave us relying on the skill and integrity of Kawai's original programming team. Happily they seem to know what they're doing and anyway if you don't like it then the basic waveforms are always there for you to work upon snail-like into the night. But these additive synthesis sounds are not the end of the story. No matter how carefully you build up a sound from harmonics, you will not get a perfect representation of a real instrument or sound unless you build in an infinite number of those harmonics. Attack shapes (the first part of a sound) tend to be particularly complex and consequently rotters to get right.

The K-1r's ace in the hole is a selection of PCM samples that can also be used as your initial waveform. There are 52 in all, a variety of attacks, central loops and overall shapes from pianos, strings, brass, guitars, flutes plus a selection of drum and cymbal samples. The inclusion of these as primary operator waveforms brings spectacular results. Writing Your Own Programming from scratch takes time to learn but is worth the effort, far more so than on FM synths. Once you grasp the terminology and realise what on earth parameters like KS autobend actually mean, you can twist and fine-tune sounds to perform exactly as you wish. Very simply, to produce a single sound you take up to four waveforms and for each one set the base pitch, envelope it, twiddle with the parameters, (understatement of the article here), then scale the keyboard response. All this is done using the patch selection keys and the up/down yes/no pads.

It is not difficult to absent-mindedly push the wrong button and exit editing mode but Kawai has included an edit buffer (praise be to Kawai) so you can get it back again. You get the choice to trade the K-1r's polyphony against the complexity of the sound. If you're happy designing with just two waveforms then you can keep the maximum 16-note polyphony. Using all four waveforms takes you down to eight.

The preprogrammed single patches in the K-1r demonstrate the versatility of all this. There are eight banks of eight, the noises ranging from stunningly ethereal dry-ice-and-aircraft-landing-light sounds, to good powerful organs, to delightful panpipes. Each has a programmable 10-character name which is displayed on the backlit LCD display. With excellent reaction to velocity sensitivity, aftertouch, pitch-bend, modulation and sustain, there is nothing at all that can be criticised.

The Multi Story So far, so great

Now let's take a peek at the 32 available multi sounds. Imagine putting eight of the single sounds all under each other in a single patch. Or split into eight areas across the keyboard. Or chosen by velocity, hard for one sound, soft for another. We are talking power here, enough to keep most megalomaniacs dancing in their underwear for weeks. The only trade-off is against polyphony. If you layer eight single patches on top of each other then you could end up with a monophonic sound, albeit one which could fell a yak at sixty paces.

However, sensible splitting and voice allocation will avoid most such problems and at least it is you that gets the choice. Apart from producing weird and wobbly megapatches, the multi patches can be used for customising sounds for a single purpose. As an example, I was recently called upon to provide the keyboard parts for a cover of "Biko" so I set up the lower end of the keyboard to give earthy and swelling chords.

Knowing that the lead keyboard line would have to cut through both this and one of the world's loudest and most inconsiderate guitarists, I set the top octave some three times louder than the rest of the keyboard. The result was perfect, my lead line was extremely audible and my guitarist fell over. Such designer patches are only possible with this sheer quantity of zoning capability.

As an expander for computer sequencing, the K-1r offers eight channels of multitimbral sound. Which sounds are on which channel can be varied pretty sharpish since this information is part of each multi patch, with each of the eight single sounds allocated to receive on a different MIDI channel if you so wish. Gosh. The only possible criticism is that you can't get separate outputs for each sound, although you can allocate each to the left, right or centre of the stereo output. A comprehensive MIDI spec is provided with the keyboard, complete with hex data for system exclusives. In fact special commendation should be given to Kawai for the manual and data sheets which are among the most comprehensive and clear I have ever seen.

I was particularly interested by the description of the K-1r as a "multidimensional" synthesiser. As even the most stupid PhD physics graduate will tell you, a single dimensional synthesiser would be a dot floating in free space. Certainly the K-1r is multidimensional (three to be precise, I counted them) but this could hardly be described as a unique selling point. Perhaps a little marketing-speak creeping in here?

All in all then, the K-1r is a faultless little number that deserves to sell in large and fruitful quantities to people who don't need the master keyboard of the K-1. Its factory-supplied sounds are excellent and the programmable possibilities endless. The RAM card slot provides ample ability to expand your sound library.

Wholeheartedly recommended to all.