Two Dutch designers who were influential in the design of the cult '80s Synton Syrinx synth have recently returned to the field, coming up with the exclusive and highly idiosyncratic Fénix synthesizer.
The Chinese do it with animals — every few years the rats, ducks, tigers and pigs come flying round again. We Caucasians, on the other hand, do it with our music industry. The '50s revival, the psychedelic revival, the prog‑rock revival (well, let's not get too carried away here) and, of course, the modular synth revival. I don't know how long the cycle is in the Orient, but over here it's about three decades. From The Small Faces and The Beatles to Pulp and Oasis, and from big Moog modulars to big Analogue Systems, Doepfers, and Technosauruses — it all comes round again if you wait 30 years.
Indeed, 1999 may well become the year of the modular synth. You only have to look at the number of manufacturers offering these instruments to realise that there is a huge resurgence of interest in them. Whereas, five years ago, you would have been limited to hunting down a vintage Moog, ARP, or maybe a Roland System 700, today you can choose from a dozen fully fledged systems from countries as diverse as Canada, Germany, Switzerland, Holland, and the UK. But this Serge, oops... I mean 'surge' of interest has led to the creation of at least one strange and powerful 'modular' synth that isn't really modular at all. Confused? Then read on...
Synton is a name that you may know. Due to the rarity and exclusivity of its products, the company has acquired almost legendary status, yet few players have ever played — let alone owned — any Synton instruments.
Founded in the early 1970s by Felix Visser, Synton was a Dutch company that started building vocoders and competing with the likes of EMS and Sennheiser. Unfortunately, the market for expensive vocoders is extremely small, so the company decided to branch out into importing and distribution, taking on Fairlight, Emu and Linn products and supplying them in the Netherlands, Belgium, and Luxembourg.
At this time, Synton employed two people whose importance to our story will soon become clear. Marc Paping was a product specialist selling and supporting the Fairlights that the company distributed. At the same time, Bert Vermeulen was developing several products that would be released under the Synton name, and which would eventually become highly regarded collectors' items. The largest of these was the System 3000, an analogue synth with 20 modules that fitted in a set of 19-inch cases. Never built in large quantities, this was a powerful instrument distributed and used by no less a luminary than Bob Moog, who supplied numerous systems to Karlheinz Stockhausen for live use.
Far more famous, and undoubtedly the product that made the company's reputation, was the Synton Syrinx. First shown in 1982, it had a UK price of under £400, but now sells to analogue anoraks for as much as £1,500! This is ludicrous, but there's no denying that the Syrinx was a unique synth with remarkably flexible filters and a range of sounds that were hard to emulate elsewhere.
In 1986, just after the Ensoniq Mirage was released, Synton approached Ensoniq in the USA and suggested a joint venture to bring their products to Europe. Ensoniq agreed, and thus Ensoniq Europe was born, and it went on to have offices in six countries across the continent.
So Synton had a highly respected manufacturing division, distribution rights for major brands, and control of all Ensoniq's products throughout Europe. The company couldn't fail to become big, and very successful. Except that it did — fail, that is. Despite everything, the bankers pulled the plug on Synton, and the company collapsed. The staff dispersed, and Paping (who had been with Synton for seven years) became a computer programmer. Vermeulen, following 11 years at the company, went to work for the Dutch railways.
Paping remained a synthesizer player and fanatic, with a significant collection of instruments. For some purposes, however, nothing in his collection was quite what he wanted. So, in 1997, he got together with Vermeulen and the two of them specified and designed a semi-modular instrument that contained all the facilities Paping wanted. Although gainfully employed elsewhere, they also set about building their new design.
Since this project was a labour of love, the partners took great care not to cut corners. For example, standard knobs were not good enough for them so, instead of buying conventional bits and pieces for tuppence ha'penny in Amsterdam's local Components 'R' Ooos, they had bakelite knobs fabricated in Taipei. These, like the original Moog knobs that they resemble, included metal sheaths to ensure that the new synth felt exactly right. And the name of this instrument, which rose from the ashes of Synton? Of course, it was the Fénix.
- VCOs: 3.
- LFOs: 3.
- Noise modules: 1 (2 independent sources with VLFO output).
- Multimode VCFs: 2.
- 24dB/oct low-pass VCFs: 1.
- VCAs: 4.
- General-purpose mixers: 3, plus 1 integrated in VCF3.
- Dedicated CV mixers: 2.
- EGs: 3, plus extra AD in EG1.
- Portamentos: 2.
- Ring modulators: 2.
- Slew limiters: 1.
- Sample & Hold generators: 1.
- Comparators: 1.
- External signal processors: 2.
- Main CV inputs: 2 (3.5mm mini-jacks on rear panel).
- Auxiliary CV inputs: 2 (3.5mm mini-jacks on rear panel).
- Gate inputs: 2 (3.5mm mini-jacks on rear panel).
- External audio inputs: 2 (quarter-inch jacks on front-panel).
- MIDI: No.
- Other connection: Grounding point.
- Power: 17V/17V @ 880mA.
- Power supply: Universal 115V/230V external PSU.
It's worth taking a detailed look at the Fénix's architecture, if only to see how much an instrument can differ from the mass-produced synths churned out by most companies. And what differences there are! Whereas major manufacturers must provide a balanced range of facilities to cater for the widest range of tastes and requirements, Paping and Vermeulen clearly felt no such constraints. As a result, the Fénix is perhaps the oddest synthesizer you'll (n)ever play (see 'The Future' box on page 54 for why you might find it hard to buy one).
The Fénix includes 31 modules, but it's not a truly modular synth. It is a semi-modular in the grand tradition of the Roland System 700. Unlike true modulars — which allow you to select and position the combination of modules you buy — these have predetermined architectures. However, unlike pre-patched instruments, semi-modulars give you the freedom to connect the various bits of the synth in the ways you want. The Korg MS20 does this in a very limited fashion — you can rearrange the modulation routings but not the signal path. The ARP 2600 is more flexible; it is pre-patched internally, but you can augment or override these routings using patch cords. The Fénix and its equivalents are more flexible still because, although they offer no pre-patching (and are, therefore, silent with no cords connected) they have more devices, and give you the greatest range of options once you start sticking cords into holes.
Ah yes... the holes. These are colour-coded to aid patching. Audio outputs are green, CV outputs are blue, and inputs are black. It's a good ploy, and it helps. I wish I could say the same about the silk screening — about which I will say more later.
Anyway, let's get back to those modules. Starting at the beginning (where else?) the Fénix has three VCOs. These are not based on the ubiquitous Curtis chips, but instead consist of low-noise op-amps, and they undoubtedly contribute to the Fénix's individual character. VCO1 is the most comprehensive, with knobs for coarse frequency, fine tuning, pulse width, PWM depth, hard/soft/no sync, and CV2 amount. It also features no fewer than seven CV inputs and five audio outputs. The inputs are CV1 (which, throughout the Fénix, is calibrated to exactly 1V/oct for standard use), CV2, PWM source, sync source, sync amount, a dedicated CV for accurate frequency modulation, and a CV to control the phase of the sub-oscillators. Ah, didn't I mention the 'subs'? There are two of these, so the five audio outputs are: sine wave, square wave, sawtooth, plus sub-oscillator 1 (a square wave one octave down) and sub-oscillator 2 (go on... guess!). The phase shifter allows you to modulate the phase of the sub-oscillators with a maximum deflection of ± one wavelength of the basic pitch. The effect of this is, in isolation, vibrato, but combined with other waveforms it can thicken up the sound considerably. The second and third oscillators are more basic, each with just square and sawtooth waveforms, four CV inputs (CV1, CV2, PWM and FM) and two outputs — one for each waveform.
There are also three conventional low-frequency oscillators. LFO3 is the basic one, which offers sine, triangle, and square waveforms, each with independent outputs, and a frequency control. By contrast, LFO1 offers two waveforms — positive-going triangle and square waves, with independent outputs — and controls for frequency, symmetry, and for the amount by which the CV2 input modifies the LFO frequency. There is one other CV input, a trigger that allows you to reinitialise the wave using an external source. This is, of course, LFO 'sync' and is the same as oscillator sync, albeit at a lower frequency. LFO2 replaces the positive triangle wave with one centred on 0V, and adds a positive triangle at twice the basic LFO frequency. It also replaces the trigger input with a gate input that allows you to switch the LFO on/off/on/off... ooh, I'm getting dizzy.
For many players, it's the filters that determine a synthesizer's worth and here, also, the Fénix eschews dedicated chips for op-amps and custom circuitry. The Fénix has three such VCFs and, again, one of these is distinctly different from the others. The simple ones are VCF1 and VCF2, each of which has a single audio input and dual CV inputs: CV1 is, as always, calibrated to 1V/oct for precise filter tracking, while CV2 is a variable input. These filters offer three characteristics: 12dB/octave low-pass; 12dB/oct high-pass, and band-pass, and there are three controls for each: cutoff frequency, resonance, and the amount by which CV2 influences the cutoff frequency in either positive- or negative-going directions.
VCF3 is, however, a completely different beast. This is a low-pass filter with outputs for both 12dB/oct and 24dB/oct rolloffs. Again, there are two inputs called CV1 and CV2 but, unlike its brethren, VCF3 also has three audio inputs and a 'hidden' audio mixer that allows you to feed three signals directly into its audio path without tying up one of the mixer modules. There is something else special about VCF3. Unlike all other self-oscillating 24dB filters (which, to be honest, rarely live up to their descriptions in my opinion), the Fénix uses five filter 'poles' to achieve a theoretical response of 30dB/octave. This makes VCF3 particularly powerful, and again contributes to the instrument's distinctive capabilities.
Next we come to the envelope generators — of which there are, surprise, surprise... three! And, just to make life interesting, they are all different. Aargh!!! EG1 is a complex beastie, with an architecture described as APDSR. The A, D, S and R will be familiar, but the 'P' may not. It's actually a Peak hold, that sustains the maximum level for a short period before letting the envelope enter its decay stage. The A, D and R stages can all — and independently — be voltage-controlled, and two other sockets provide the necessary input for the Gate pulse and for the EG output. But if this sounds almost too simple for the Fénix... it is, because EG1 is actually two envelope generators in one. The second is a simple AD envelope with a near-instantaneous attack, a single control for the Decay time, and an output. Well, simple is still the wrong word, because you can voltage-control the Decay time and the Attack level for a wide range of effects. As with everything else in the Fénix, nothing is as basic as it seems.
EG2 is no less strange. It lacks a Sustain control but, according to the position of a switch, acts as either an Attack/Peak/Decay (APD) envelope or an Attack/Peak/Release (APR) envelope. There are knobs for the attack time and peak hold time, plus a knob and a CV input for the Decay/Release time. See, that wasn't too hard (was it?) — except that was only section 1 of EG2. Section 2 is another AD envelope with a CV input for the attack time and a knob for the decay time. Oh yes, and there are three outputs: EG2(1) only; EG2(2) only; and EG2(1+2). Help! My brain is trying to curl up into a foetal position!
With the style I'm coming to expect of the Fénix, EG3 is yet another odd bit of circuitry. As an envelope generator it's just a simple AD response with near-instantaneous attack, and a single knob for the decay time. Its more obscure function is its ability to delay the onset of the envelope by a user-defined amount. After this delay EG3 performs as a simple gate, and its output is either a negative- or positive- going pulse (depending upon the output used) that occurs sometime after an external Gate is presented to its input.
If the Fénix was pre-patched with just these facilities plus some integrated VCAs and mixers it would be very powerful. But the VCOs, LFOs, VCFs and EGs represent fewer than half the modules available. So let's continue...
There are two dedicated CV mixers, CVMIX1 & CVMIX2 Found in a dense cluster of modules on the far right of the panel, its inputs and output are hard to locate, and the knobs that control it are far from obvious. Once worked out, though, everything performs as expected. CV MIX 1 has a 1V/oct input to which you can add a variable amount of any CV2, plus a user-defined offset dialled in from the appropriate knob. The other CV mixer is identical, except that instead of a fully variable offset, it offers a four-position knob with -1V, 0V, +1V and +2V positions. This makes it ideal for use as an octave selector switch.
There are also three general purpose mixers, MIX1, MIX2 and MIX3, in addition to the mixer integrated into VCF3. MIX1 is the most complex. It has one 'normal' input, one that can be attenuated from 0 to ± unity gain, and a 1V/oct input for use when mixing keyboard and other CVs rather than audio signals. MIX2 and MIX3 are slightly simpler, lacking the gain/polarity control for the second signal input.
At last we come to something simple: the Fénix has four VCAs, but each of these has just a signal input, a CV input for gain control, and an output. Hurrah! And what about the Sample & Hold, the Slew Generator, and Ring Modulator 2? Well, at least these are straightforward. But then complexity sets in again... There are two portamentos, although these share a common input. This means that you can't 'smear' two separate voltages, but instead you can apply two different amounts of portamento to a single CV. Furthermore, there is an on/off CV input that allows you to defeat the portamento using a gate-type signal.
RING1 is yet another weird piece of engineering. Look at it like this: almost by definition, ring modulators have two inputs (of frequencies X and Y) and one output, producing the sum and difference frequencies (X+Y and X-Y) of the two inputs. But this one doesn't. It has three inputs, two outputs, and the number of frequencies produced depends upon the signals applied. For example, used conventionally, the output is X+Y, X-Y, plus the signals applied to inputs 1 and 2 themselves, which are X and Y. The same result is obtained at a lower amplitude if you use inputs 1 and 3. If you then try inputs 2 and 3, you get next to nothing, and what you do get is probably just the breakthrough in the circuit. But what about using all three inputs together? Surely inputs 2 and 3 are roughly the same thing? Well, no they're not. Input 3 applies a further degree of amplitude modulation to input 2 so, if you apply an LFO square wave to input 3, the 'on' section of the square wave allows the RM to function in its 'normal' fashion, whereas the 'off' section passes only input 1, with just a little RM'd signal breaking through. This is interesting, but now imagine applying a sine wave to sweep the RM's effect in and out. Then imagine this at audio frequencies... and now apply complex waveforms to the modulating input. It's a nightmare of sonic possibilities! Of the outputs, one is designed purely for audio frequency signals, while the other also handles low-frequency CVs. You can, in general, treat them as dual outputs of the same signal, although you can, with a little patching, extract the subtle differences and use them to make even more complex sounds.
The noise generator is also, as always on the Fénix, not quite what it seems. Where else would you find two noise generators, an analogue one and a digital one, occupying the same 'module'? Without a spectrum analyser I can't be sure, but I suspect that the analogue noise source produces pink noise. This has a characteristic 'roaring' sound often used for wind and wave effects, although you can of course use a filter to alter its sound. Only one control is associated with this — its output level. By contrast, the digital noise source produces a huge range of tones, ranging from an almost tonal distorted growl to a high-pitched 'blue' noise in which high frequencies predominate. You can set the timbre with the knob provided, and also use a CV to modulate the noise in any way you wish. Even this isn't the end of the story, because the Noise module also produces a random low-frequency waveform that can be applied to, for example, oscillators and filters to introduce subtle variations into the sound.
The Comparator module is another oddity. It has two inputs (A and B) and, if the voltage at A is higher than the voltage at B, the output is 'high'. If the voltage at A is lower than B, there is no output. This means that you can use the comparator as a switch, allowing signals to pass at times determined by LFOs, envelopes, or even other oscillators.
Finally, the Fénix offers two modules that allow it to handle audio from the outside world. The first is an external sound processor (ESP) that accepts audio and outputs three signals: the audio itself, a Gate derived from the presence (or not) of the external audio, and an envelope derived from the amplitude of the external audio. The second ESP also accepts external audio, and produces a trigger, a 'step' (which is an amplitude sample taken when the external signal changes abruptly) and the audio signal itself. These modules have independent (although mislabelled) gain controls, and the former has a threshold control to help prevent false gates being produced. Like all ESPs, the Fénix's are not entirely reliable, but they add the ability to process sounds through the filters, or to create triggers from external drum tracks, and so on. Thankfully, the audio inputs are quarter-inch jack sockets, so you should be up and ESPing in no time at all.
And that, finally, is it! A mere 31 modules of mind-numbing diversity and complexity — simply patch up your sounds, direct the audio to the stereo outputs, and you're laughing. The only difficulty you'll encounter is finding a 'Y' cable with a stereo quarter-inch jack plug to take the sound to your mixer, amp, or whatever.
Writing this review has been as much about will-power (stop playing and write, boy!) as it has been about investigating the pros and cons of the instrument. The Fénix invites you to experiment with sounds and real-time control, and to try to work out what it's doing.
On the design side, the Fénix has some bloopers in its layout — putting the portamento as far as possible from the input CVs is a classic — but I understand that many of these were determined by design considerations, not oversights. More culpable is the ambiguous silk-screening, with legends applying to one control almost always closer to the one above, and with the names associated with several sockets closer to the wrong ones. This is especially confusing in the densely packed area to the bottom right of the panel. And what about the size of the thing? It's built like a rackmount, it looks like a rackmount... but it's several millimetres too wide to fit into a 19-inch case! Ooh-er!
But the Fénix sounds excellent, it's incredibly flexible, and it boasts impressive engineering; for example, I was impressed by the inclusion of heater circuits for each of the VCOs and VCFs. These keep the devices at 60°C for accurate tuning and stability (a two-colour LED switches from red to green to tell you when the instrument has reached the correct operating temperature). The ±4V audio levels are another nice engineering touch — they're hot enough to pass a full-level signal from your left hand to your right (go on, try it!) and the source of excellent signal-to-noise ratio figures. Similarly, the envelopes and triggers will respond to a wide range of voltages. But, on the other hand, the Fénix's CVs generally lie in the ranges ±4V or 0V to +8V, making it harder to interface it with the more common ±10V ranges of other synths. You win some... and you draw some!
Of course, the use of banana plugs (see the 'Right Connections' box) won't please everybody, but I believe that the benefits outweigh the limitations. Indeed, two of the earliest customers asked whether custom units could be built with 3.5mm mini-jacks, but they came round to the Fénix way of thinking in the end. And that's just as well, because Paping said that it would be against his design principles!
Give me a year with the Fénix, and I'll tell you exactly what I think of it. Give me two weeks, and I'll tell you what I think that I'll think of it in a year's time. It's a question of detail, you see. You can put a Fénix next to a Roland System 100M and, in terms of complexity, they'll look little different. But this observation is superficial — the Fénix is very much more elaborate. For example, each of the 100M's oscillators are identical, whereas the Fénix's show marked differences. Similarly, the Roland's filters, envelope generators and so on are homogeneous, whereas the Fénix's are all distinct from one another.
The consequence of this is that there are many more possibilities hidden within the Fénix than a summary of its facilities would suggest. Furthermore, it makes any statements regarding its sound almost worthless. Does it sound fat like a Moog? Well, yes it can. So it doesn't sound crisp like an ARP or fizzy like a Korg? Well, it can do that too. What about Roland- and Yamaha-type sounds? Given a few patch cables... yeah, no problem.
The bottom line is this: the Fénix feels like a piece of laboratory equipment but behaves like a musical instrument. It's nowhere near as imposing as a Technosaurus, and visually it's not a patch (oops, sorry) on the new Analogue Systems RS8000. It doesn't have the range of modules the Serge does, and it's not as clearly laid out as any of them. But these systems all share the most important quality: they can sound great!
To my mind, the Fénix is undoubtedly the oddest of the bunch, and it's unlikely to appeal to most players. But for those prepared to experiment, it will be a source of never-ending pleasure, producing sonic garbage and audio candy in equal measures. And, at the price, it's a steal. The chances are that you'll never own one, but I think that you should at least try and get to play one, if only to discover how truly odd and intriguing a synthesizer can be.
There's no reason why it must be so, but conventional synthesizer design decrees that the information carried by a patch cable is determined not by a current, but by a voltage. This means that the cable must be capable of sustaining a potential between the primary conductor (usually the core) and a reference 0V or 'earth'. Conventional quarter-inch and 3.5mm cables do this by wrapping the core in a 'screen' that is grounded at both ends of the cable. This has the advantage of protecting the core from stray interference, and it ensures that the modules at either end have the same earth potential. The disadvantages are that the cable and connectors are generally less robust, and more likely to suffer from intermittent crackles.
The same purpose can be achieved using a single conductor that carries a voltage referenced to a common chassis ground, if the modules within the synthesizer are robustly earthed internally. You can then use thicker cables with much sturdier connectors, although the system may be more prone to interference, and maximum cable lengths are much shorter.
Although twin-conductor cables are the norm, single-conductor patch cables are nothing new; indeed, Don Buchla was using them before many of us were born. The type of connector used is called a 4mm banana plug, and offers two important advantages over standard cables. Firstly, they make much firmer connections than jack plugs. (If you stick a single banana plug into its socket, you can almost lift the Fénix using just that one cord!) Secondly, banana plugs are stackable, so you can distribute a single CV or audio signal to a variety of destinations without needing space-hungry 'multiples'. I found that I could stack connectors up to six high before there was any noticeable degradation to the CV. Unfortunately, you can't use the same principle to do away with mixers; the modules' outputs interfere with each other, so cables stacked at an input do not sum signals linearly, and the results are at best unpredictable.
Inevitably, there's a down side to using these cables, and here it's the problem of inter-connectability. The Fénix may be compatible with a handful of rare Buchlas and Serges, but you can't patch it directly to an ARP, Moog, or any of the new batch of modular synths such as the RS Integrator or the Technosaurus.
However, the Fénix has an earth point on its rear panel. If you connect this to a suitable chassis ground on another modular system, you should then (in principle) be able to use a converter cable (with just the tip connected to the conductor) to transfer CVs or audio signals to and from the Fénix. I suspect, however, that this could be a bit of a nightmare. Furthermore, these single-core cables are limited to about 1 metre before interference problems set in, so I wouldn't advise interconnecting a Fénix and other modular synths if your studio is located next door to a taxi company!
Chris Whitten is a musician with impeccable credentials. He spent three years from 1987 working with Paul McCartney, and was the drummer on the latter's hugely successful world tour of 1990. He then spent the next year with Dire Straits, and played on their last tour in 1991. But in 1992, Chris switched careers to write scores for film and TV, and it was then that he became able to indulge his interest in analogue synthesizers. In 1997, Chris bought the fourth Fénix, and brought it back to the UK. This is what he has to say about it:
"At first I found it difficult to use analogue synths in my soundtracks, but as I became better known I was able to integrate them into the stuff that I was doing. I particularly like modular synths because I can do things like play drums through the filters, so I bought a small Serge panel to use as an audio processor. I would have bought more, but at $3,000 per panel it takes a bit of time to work up to one. Shortly after that, a friend in Spain let me know that Marc Paping was developing a new semi-modular synth, and he told me that he expected it to be very good. So, when Marc announced that he was building 25 Fénixes I contacted him immediately to put my name on the waiting list.
"My first impression of the Fénix was that, because of the sheer number of modules, it would take a bit of time to try everything and find out if it was any good. My first experiments were treating audio using the multi-mode filters, which I loved. I also discovered that, because both use banana plugs, the Fénix really opened up the Serge for me (I now have three Serge panels) and I was very excited by that. Unfortunately, it's almost impossible to get the necessary banana cables here, so I recommend that anybody considering buying a Fénix or a Serge should also place a big order for cables.
"Now I'm very happy with the Fénix, and I use it both as a monosynth and as a sound processor. Mind you, it's not without its faults, and I remain disappointed by the look of it, which is a bit cheap compared to the Serge and the Analogue Systems that I have. But I won't criticise Marc and Bert for this because they have put the money into the modules rather than the cosmetics, and that is the most important thing. Mind you, it would be nice if it looked sexy too. And this thing about it not fitting a 19-inch rack is crazy. I ordered a dedicated rack for my Fénix, and when it arrived the synth wouldn't fit in it. I was very annoyed about that.
"There's also been a bit of a debate about the quality of the low-pass filter, VCF3. Some people find it a bit non-descript. Part of the problem is that everybody is always looking for that 'Moog' sound, and I think that VCF3 also suffers in comparison with the great character of the multi-mode filters. But I understand that Marc and Bert have listened to the criticism, and that later Fénixes have been tweaked a bit to make them more Moog‑ish.
"Despite these few niggles, I never hesitate to recommend the Fénix, because of the combination of sound quality and value for money. Mine is in Holland at the moment having the filter mods done to it, but I've already used it on many soundtracks and documentaries for companies like the American Discovery channel, so if you've got a satellite receiver, it's quite possible that you will have heard it at some time or another."
The background to the Fénix is intriguing, and Paping's and Vermeulen's plans for its future are shrouded in rumour and confusion. So I got on the phone to Holland, and spoke to Paping to find out more. This is what he told me...
"Bert and I designed the Fénix because none of my existing synthesizers gave me exactly the combination of facilities that I wanted. We did it for that reason alone — it was the synthesizer that I wanted. Both of us work elsewhere so, even though the Fénix is now a reality, producing it will never be the main job for either of us, and we have no wish to become a mainstream synthesizer company. That's why we built just 25 Fénixes. Having said that, we've shipped these already, so we are going to build a further 50, but that will then be the end of it. We already have a queue for these 50, and at the current rate of production, we'll still be building them in the year 2000."
I suggested that a total run of just 75 Fénixes would instantly catapult it into the 'collectable' class. "I hope not", replied Marc, "because we want these instruments to be played. I suppose that they will have a certain rarity value, and collectors might want them because of their Synton background. But neither Bert nor I like the high prices of so-called collectables, and we don't want to see the Fénix go that route. That's why we have not charged too much for them — I'm sure that we could have sold the Fénix for more than 3,200 Dutch Guilders, but this was primarily an enthusiasts' thing. We hope that we're adding to synthesizer culture, not squeezing for blood. We have produced the Fénix to share and enjoy, and our customers are our friends. Some of the existing Fénixes are in Amsterdam, there are three elsewhere in Europe, but most are in the USA. Whatever happens to the rest of them, I will be keeping the last two. After all, that's why we designed it in the first place."