Technics SX-WSA1

Technics' first foray into the pro synth market, the WSA1, catapults them to the front of the pack. Martin Russ discovers whether acoustic modelling synthesis really can create gold sounds from base metal.

When I reviewed the Technics KN2000 some time ago, I can remember thinking that with a slight change of emphasis that machine could become a genuine challenger to some of the major players in the pro synthesizer market. But at the time, digital pianos and sophisticated home keyboards were the order of the day for Technics, and I thought little else of it. So when I was offered the opportunity to test the company's first 'professional' synthesizer, the SX‑WSA1, I was particularly intrigued because the KN2000 has indeed influenced many recent and forthcoming releases from traditional synthesizer manufacturers — particularly the adoption of large LCDs with lots of soft‑keys. New manufacturers often bring a radical set of ideas to an established marketplace, and the WSA1 is no exception, as we'll discover...

Workstation Ethic

The WSA1 is actually the abbreviation for the slightly less memorable Technics SX‑WSA1 (or the rackmount variant, the SX‑WSA1R). It's a 'workstation' — which is often interpreted as merely the combination of a synthesizer and a sequencer. I must admit that my own bias has always been towards the synthesizer part — the workstation sequencer is usually relegated to the 'musical notepad' role in my mind. I've always preferred using a separate computer‑based sequencer, because I like to see a graphical representation of the music, rather than just being able to hear the notes. Well, the WSA1 is probably the first workstation that I have seen which contains a sequencer that I might actually use for more than just a quick 16‑note backing pattern.

But the main focus of any workstation is the sounds that it makes. After all, an onboard sequencer is merely a tool to help you present those sounds. The WSA1 is not just a preset sound engine coupled to a General MIDI playback device — this is a true synthesizer which happens to have a usefully powerful sequencer included. It's not very often that something genuinely different appears in the synthesizer world — I think we've all become jaded by too many variations of Sample & Synthesis (S&S) technology. Most supposedly 'new' ideas turn out to be minor tweaks to old ideas, and very little in the way of innovation. Well, I'm happy to report that Technics seem to have avoided this well‑worn rut and have succeeded in producing an instrument which is powerful and innovative, yet easy to use and versatile — and it's a neat amalgam of various technologies.

Acoustic Modelling

New methods of synthesis turn up rather too frequently. Dressing up old and tired ideas with new names doesn't really fool anyone for long, and each of the familiar methods has its drawbacks:

• Analogue synthesis is great for making synth brass sounds, but not much else.
• FM is great for tinkly pianos and evolving sounds (which are now so cliched that they are about to come back into fashion!).
• S&S suffers from 'snapshot syndrome', where the sounds it can produce are so closely tied to the sample provided that every musician ends up sounding the same.
• Physical modelling is so complex that we aren't even allowed to programme it ourselves. Instead, we have to use fixed models created by experts.

The synthesis methods above (and others) all have one major flaw: players want to make, interact, and perform with sounds. Most synthesizers fail in this basic requirement, which is why breath controllers, foot pedals, ribbon controllers, and numerous other performance devices are used to try and provide some sort of improved performance interface between the user and the instrument. The catch is that most synthesis techniques just don't let you make quick, simple, intuitive changes to sounds — and as instruments become ever more sophisticated, it seems to grow harder, not easier to achieve this.

Thankfully, the WSA1 is startlingly different. It uses a synthesis method called 'acoustic modelling', which attempts to take the best parts of physical modelling and combine them with the familiar parts of S&S. In the process, it also provides gobsmackingly good user control. My personal rule‑of‑thumb for determining the viability of a synthesis method is the name: if it makes sense then it's probably going to be a winner; if it is just technical gobbledegook, then the customers are going to figure it out eventually, and it might as well turn up at Christmas, stuffed and roasted. Rest assured, however — acoustic modelling is no turkey.

In a nutshell, physical modelling attempts to produce a mathematical model for a complete musical instrument. It then uses this model to produce sounds which emulate the behaviour of the real thing — often with astonishing realism. But the depth of detail required is formidable — you need to know a huge amount about the physics of the musical instrument, acoustics, and mathematics, and then convert this into software and electronics.

In contrast, acoustic modelling takes a step back, and looks at physical modelling in a different way. Instead of trying to work out how a complete instrument works, is there any way that the important elements can be encapsulated in a form which provides control, but is easy to manage? It turns out that there is — and it stems from research into speech. When you speak, your vocal chords are vibrated by the air which rushes past them, and this raw sound is then modified by the complex set of tubes and spaces formed by your throat, nose, mouth, teeth, lips and tongue. A physical model of this would need to consider the velocity of the air, the pressure, the tension in the vocal chords, the space between them, their elasticity, and so on — and trying to work out exactly the mechanisms for how they vibrate could take a long time. A more pragmatic approach would be to ask what the raw sound produced by the vocal chords sounds like, and what sort of filter the throat, mouth and nose form.

Acoustic modelling does just this. It assumes that musical instruments can be split into three parts:

• Drivers, which produce the raw sound. Examples are the hammer hitting a piano string, or the plectrum plucking a guitar string, or the reed vibrating in an oboe.
• Resonators, which colour the sound from the driver. Most musical instruments exhibit some sort of resonance: often the whole of the instrument vibrates along with the sound to some extent, and the way that it vibrates affects which frequencies are emphasised and which are suppressed.
• Modifiers, which shape the coloured sound. Filters and Envelopes are familiar examples.

In a real instrument, the drivers and resonators are very closely connected, and interact with each other: the hammer hitting a piano string causes the string to vibrate, but the vibration of the string is affected by the fact that the hammer is touching the string, has probably stretched the string slightly when it moved the string, and has added in a low frequency thump. The act of setting the string vibrating depends on the hammer — you can't have the sound without it, but the hammer affects the sound. The two are inextricably interconnected. In acoustic modelling, the two elements have been separated but the same interactions can be produced by controlling the way that the driver and resonator are connected together — and Technics call this the 'connection'.

With the WSA1, you get over 300 drivers and lots of resonators (you can couple together up to four of the 64 resonators to form even more complex resonators), and you get lots of control over how the driver and resonator interact. The connection/interaction part is quite interesting, because it really is very 'physical modelling' in its approach. Let's use the analogy of a bowed string. You can control whereabouts on the string you put the bow (Position), how hard you press on the string with the bow (Depth), and how the bow and string interact (which Technics call 'Fitting'). There's also control over how the position changes, and how the resonance changes (whether it is fixed or not). Only the drivers are fixed — you can alter all the parameters for the connection and the resonator.

If you listen to the raw drivers without any resonator, they sound very bright and the initial transient is highly emphasised, almost like high‑pass filtering. But since most resonators act as band‑pass or low‑pass filters, connecting this driver to a resonator turns it into a more normal sound, except that by changing the resonator you can change the timbre. The WSA1 offers all manner of resonators — some which behave like strings, some like tubes and cones, others like flared tubes, drums, and even customised resonators.

But worry not — if you don't wish to get involved in the deeper aspects of programming, then the top level of the WSA1's edit pages allows you to choose just a driver and resonator, and these act as templates for all other parameters. Programmers can delve deeper and alter individual settings. Still, it's refreshing to find an instrument that offers both immediacy for the player and real depth for the programmer. When I began editing the WSA1, it took me some time to explore all the recesses, which is usually the mark of an interesting synthesizer. If you were frustrated because the Yamaha VL1 did not allow you to delve into it, then the WSA1 could be for you. Just remember that this is not a physical modelling instrument, although it's similar. What you lose is the transition between notes — whereas a physical model will move from one note to another in much the same way as a real instrument, the WSA1 will merely play two realistic notes, one after the other.

Once you have chosen a driver and resonator (and connected them together), then the rest of the WSA1 sound generation is the familiar S&S filter and amplifier, but with separate multi‑mode filters (Hi, Lo, and Band‑pass, with additional EQ) for each of the four tones which make up the sound. There are separate envelopes, LFOs and scaling for filter, amplifier, and pitch control of the sound, and the only thing that seems to be missing is a Sample & Hold waveform on the LFOs. Downstream of the individual sounds, there are the conventional layering, velocity switching, panning, and octave transposing features that you associate with good S&S instruments. Combis enable up to eight sounds to be combined in much the same way, and there's the standard 'one effect setting applies to the mix' assignment of the built‑in effects processing.

Internal Effects

The WSA1's effects processing comes in two parts: Digital and DSP. The 'Digital' effects are a neat way of pulling together all the detuning and envelope trigger delay edits that are normally part of the pitch controls. Instead, by choosing a chorus effect here, you alter the tuning of the tones, or by selecting an echo you force the envelopes to keep retriggering. It's a very clever way of providing one place where a variety of otherwise separate functions can be controlled.

The DSP section is a conventional Digital Signal Processing type of effects unit. There are two general purpose sections and one optimised for reverb effects. The familiar collection of chorus, phasing, flanging, single and multi echoes, enhancer, distortion and EQ are all to be found here. Sound quality is high, and several parameters are available to be controlled by one or more performance controllers. So you can do things like opening up the filter with a modulation wheel whilst simultaneously increasing the reverb time.

Modern Design

The WSA1 front panel follows the modern trend of being quite sparsely populated with few buttons and controls. It is dominated by a very large (320 by 240 pixels) backlit blue LCD, surrounded on three sides by 26 assignable soft‑keys. Below this display are two sets of soft‑keys, intended for up/down control of parameter values. These sense your finger pressure, and with a light touch move in single values, whilst pressing harder makes values change in 10s. This makes editing much faster, and very intuitive. Full marks for innovation here!

On the left of the display are the performance controls: one Pitch Bend wheel and two Modulation wheels, plus a 'Real‑time Controller'. The latter is a combination of a trackball and joystick: you move it around and it springs back to the centre position automatically. Next to the volume control is the other major performance control, the 'Real‑time Creator'. Although this resembles a smaller version of the joystick/trackball combination controller, it is actually more like a 2‑way wheel: you move it around and it stays in position — with four LED lights to indicate the position, and a 'Reset' button to clear any movements you make.

These two real‑time controls can each be assigned to two of the major controlling parameters of the Acoustic Modelling technique, and moving the controls alters the sound accordingly. The default setting puts the Fitting and Position parameters onto the two axes. Moving the control and playing a sound enables real‑time changes to be made to its timbre. With the Fitting and Position parameters, the changes vary depending on the sound you are modifying. For example, the Piano sounds can be made brighter, more muted, more hollow in character, or thinner in tone — words really aren't adequate to describe what happens. Moving the controller around also produces a slight phasing or chorus effect as it moves, which players can put to good use as a performance effect. You really have to sit down and play around with these controls to appreciate how sophisticated the changes are — it's not at all like the VCF cutoff or vector mixing that you might have experienced on past synthesizers. After a short while, I got so used to selecting a sound and moving the Creator control to vary the timbre that I began to miss this feature on my other keyboards! (Incidentally, you can save these settings with each sound for later recall.)

On the right of the display are the rotary dial, increment and decrement switches, plus the sound group selection and editing buttons. I didn't use these controls anything like as much as on other workstations or synthesizers. The LCD and soft‑keys very quickly become the focus of your attention, almost to the exclusion of everything else. So, although you can use the soft‑keys to select a value, and then tweak it with the rotary dial, you tend to use just the soft‑keys because they are so conveniently placed. Sound selection is made by choosing a group number from the group buttons, and then pressing the soft‑keys around the display to choose from the eight sounds in that group (this also fits into the General MIDI naming conventions).

It's not very often that something genuinely different appears in the synthesizer world... Technics have succeeded in producing an instrument which is powerful and innovative, yet easy to use and versatile.

The keyboard itself is quite springy, has metal weights built into the keys, and felt quite sluggish to my hands — but then I'm very used to my Yamaha SY99 keyboard. It is attack velocity and aftertouch (pressure) sensitive, and the velocity curve felt slightly too hard for my own playing style. However, a quick visit to the 'Touch Sensitivity' page of the 'System' menu enabled me to change the curve (10 options) and the velocity offset (1‑127) of the velocity sensitivity. plus two parameters for the touch/pressure sensitivity. This is rather more user adjustment than you normally find on workstations, and is more like a master keyboard. Luxury!

Rear panels are boring. This one is no exception, but it does offer two sets of MIDI sockets (32 channels), stereo supplementary outputs, and space prepared for additional outputs and a digital audio output. The sockets are labelled along the top edge of the front panel (hurrah!), which is incredibly useful when you are always groping over the back of keyboards trying to connect things up.

Performance

64‑note polyphony makes for lots of sustained notes, which appeals to my love of the sustain pedal, but the real‑time controls are the winner here. Being able to take a piano sound and alter its tone with a single movement is wonderful, and changing the timbre of a solo sound while it plays is superb. I couldn't make these controls glitch, although sometimes the changes were a bit abrupt, so you need to move around slowly at first to find the 'hot spots'.

The wheels have a smooth, but slightly stiffer feel than I prefer. The detent in the second modulation wheel is very light, and the deep ribbing makes it difficult to find the notch, which can be very disorientating. The volume control is just black plastic with no white (or gold) line, and would be virtually unreadable on a darkened stage, which is a shame because the metallic gold lettering should be very visible, even in monochromatic lighting.

The feet are made of hard, smooth plastic, rather than the soft rubber feet I prefer, and don't grip onto smooth surfaces very well. They also seem to be held on with plastic push‑rivets, which look like they could get sheared off rather too easily. Given that this is Technics' first attempt at a professional synthesizer, these are minor criticisms, but I suspect that future versions will be more roadworthy, since the Technics staff I spoke to seemed to be very responsive to user feedback.

16‑Track Sequencer

I don't normally like workstation sequencers, but this one was an exception. Whereas most workstations use the 'tape recorder' metaphor for their sequencer functions, Technics have provided an onboard 16‑track sequencer which is much more like a midrange computer‑based sequencer. You get real‑time and step‑time recording with looping, plus a master track to hold time signature and tempo information. But, most significantly, there is a piano‑roll style display of the track contents, so you can see exactly what you have recorded and edit it easily, without relying on your ears alone.

Technics have spent a long time getting the sequencer right, and it shows. Unlike the text‑only interfaces that infest most workstation sequencers, the WSA1 has neat, structured menus with icons (see screen photos). In fact, the sequencer is probably more logically laid out than the synthesizer edit section — it certainly uses the soft‑keys more consistently.

There are 16 tracks, plus the master 'tempo' track, and up to 10 songs can be chained together. Total capacity is about 47,000 events. Track and Part controls are presented visually using a 'mixing desk' metaphor, which makes the most of the graphical display and the soft‑keys. I never normally read sequencer specifications in any depth, and I suspect that you don't either, but suffice to say that it performs all the usual track copies, dynamic processing, controller insertion and deletion, etc.

The sequencer saves its data as Standard MIDI Files, so transferring to a computer sequencer is a doddle. The WSA1 can play Format 0 files directly from DD (720K) or HD (1.44Mb) 3.5‑inch floppy disks inserted in its built‑in drive, which is very convenient for commercial MIDI File disks. However, you need to load Format 1 files into memory prior to playing them.

MIDI Matters

There are plenty of MIDI filters and switches available to control how MIDI data behaves on the WSA1. The assignment of controllers to MIDI Controller numbers is particularly clear and easy to understand, thanks in part to the superb display. Having two separate sets of MIDI connections enables more flexibility in how you connect up the WSA1. One set of MIDI sockets is intended for System Exclusive transfers and connections to the internal sound generators, whilst the other set is intended for connecting the internal sequencer to external MIDI modules. This gives a 32 channel MIDI implementation, rather reminiscent of the Gem S2.

A written manual covers the MIDI Implementation Charts and supplies comprehensive details of the WSA1's System Exclusive messages (Technics use a handshaking method of checking that SysEx dumps have been transferred correctly). Although the manual's coverage of MIDI is detailed and certainly on a par with the best I have ever seen, there is no separate Implementation Chart for the sequencer; instead, it is lumped in with the main chart for the whole instrument. Despite this, the MIDI manual gets '10/10' for clarity.

Manuals

The other three manuals deserve a mention, too. There's the usual slim 'Getting Started' manual, which shows you how to get noises out of the WSA1 when you first unpack it. The 114‑page 'Practical Applications' manual covers all the menu screens — although it often does not go deep enough into the function of all of the parameters. Extra details are provided in the 'MIDI' manual, along with additional information about the DSP and Digital Effects parameters, but the third manual is just a list of the factory sounds and drum mappings.

I'm afraid that I often find that synthesizer manuals don't go far enough in providing support for the user. Technics have concentrated on the menu screens, but this does not help with the 'How do I do this?' type of question. I encourage readers to complain about manuals, because it's the only way that manufacturers will improve them. Luckily, the WSA1 is easy to use and the operating system is very user‑friendly, so I did not have too much trouble finding my way around, but a little more explanation would make so much difference.

Conclusion

It's always pleasing when a company new to this market produces a synthesizer, because they often bring with them fresh ideas and different approaches. The WSA1 takes a refreshingly alternative look at synthesis and its acoustic modelling delivers just about the right mix of player‑friendly and programmer‑friendly features. Technics seem to have a positive attitude to software updates — which can be very important insurance to a purchaser — especially with the first product in a field. They inform me that ROMs might be upgraded three or four times in a typical product lifetime, which means that any bugs stand a good chance of being ironed out. Review models are often slightly buggy, but with the exception of one minor bug which Technics told me about (and which will be fixed by the time you read this), I didn't encounter any problems.

So how about the competition? There are plenty of 'sample & synthesis' workstations around — even Korg's new Trinity seems to be another variation on S&S with the bonus of a plug‑in Prophecy. If you're looking for something with a little more synthesis power, you are into Kurzweil K2500 territory, and that's about it. Now if the WSA1 was just a little cheaper, it might just be able to break the stranglehold that S&S instruments have on this market. Even so, anyone astute enough to go for a WSA1 can smile to themselves, secure in the knowledge that they have a true synthesizer which delivers rapid and intuitive real‑time changes of timbre, and the nearest thing yet to affordable polyphonic physical modelling.

With a debut instrument like the WSA1, Technics have finally broken away from their 'digital pianos/home keyboards' image. Try a WSA1, you will be surprised — but be warned that the real‑time control is highly addictive!