«Combining existing technology into a truly expressive monster-synthesizer Fig.1: Here it is. Another, no holes barred Marc Brassé beast Introduction ...»
Combining existing technology into
a truly expressive monster-synthesizer
Fig.1: Here it is. Another, no holes barred Marc Brassé beast
This article is about my Starship One project. It does not only describe this “instrument” but also explains the
idea's behind it. It is a truly multifaceted tale, even though I'll have to take some shortcuts to avoid turning
this into a complete book about my vision on synthesis.
First, as a sort of introduction, this list of the “specs” of the total setup:
• 4 manuals, including:
1 solo ribbon controller (Doepfer R2M) 2 polyphonic keyboards (Technics WSA and GEM S3) of which 1 has poly aftertouch S3) 1 set of polyphonic basspedals (Fatar MP1) • 3 independent synths (Technics WSA1, Technics WSA1R, Gem S3 turbo)
• User sample loading (WAV's), so a sampler on board in all but the name (GEM S3 turbo)
• More then enough "oscilators" / voices to play with (64 + 64 + 32) • 32 track fully editable sequencer, with extensive MIDI controller compatibility, even including poly aftertouch (Kawai Q80).
• 8 track high spec digital audio recorder (Roland VSR-880).
• a high grade, near modular, real time, stereo effects processor to add a bit of the analog twiddly stuff if needed Eventide Modfactor).
• Hardware Audio patch bay
• Extremely versatile MIDI filtering processor (MIDIMIXFIX)
• Chromatic tuner (Samson CE40).
The more subjective bits:
• Lot's of knobs and switches.
• 8 independent displays!( er, that is actually decadent).
• A snazzy customized casing (well, at least I like it) including a hint of Star Trek TOS bridge styling.
Now does that sound attractive or not?
Don't get me wrong though. I am not claiming that I have built a full instrument myself. This is more of an exercise in combining existing elements into an ergonomic and thus inspiring package. It has been a worthwhile effort though. Somehow the total seems to be more then the mere sum of its parts. This has come very close to my ideal.
What can the reader gain from my experiences? Just like with the Son OF GX project I hope to inspire others to do something similar. So the next time you think that your most expensive synth still doesn't totally fit the bill try to formulate what it exactly misses and get your thinking cap on. You to might turn a mere compromise into a winner.
Philosophy What is the sense behind this particular project one might ask. There are so many instruments on the market so why choose this particular combination of elements?
Well, that has everything to do with he sort of music I am most interested in and thus want to create.
Basically I am a fan of electronic music in its widest sense, from the earliest pioneers to the latest incarnations and then I mean music made with electronic means in general. In this respect I do not only mean electronic dance music, which is often put under the same general label but is a much more restricted in it's form.
Furthermore I am a fan of a lot of other musical styles, including classical music. One can say a lot about western classical music. That it is to elitist, too complex, to brainy and that other styles are much more rhythmically sophisticated but one cannot deny that it is very refined when it comes to harmonic content and expression.
It is my ideal to compose music that is melodically understandable and interesting but still original enough to avoid normal clichés. In practice this boils down to w wish to combine the best of classical expression with the sonic width of electronic instruments. The sounds that I actually find the most interesting are those that blur the line between electronic and acoustic. If a listener says ”I dunno what this is but I love it” I have hit the spot I wanted to hit.
The Starship One project is my attempt to “build” a hardware instrument that is the ultimate tool for creating such music.
That leads to some key principles. My ultimate instruments should:
• be extremely versatile
• be extremely expressive
• enable the performer to separately record all the elements of a life improvisation in one go
• enable further refinement of these improvisations after the initial recording
Initialisation This particular project more or less started when I bought my first Technics WSA1 synthesizer in 2012. I knew about this synthesizer since it was released in 1995 but never heard it, let alone play it myself. 17 years later I did. And boy was I impressed.
That deserves some explanation: The WSA1 was one of the first so called virtual modeling synthesizers. In these instruments the sounds are no produced by dedicated oscillators but by a software program describing the properties of the simulated sound. This was more or less a natural reaction to sampling technology.
When samples are used the individual sounds as such are very detailed and realistic but what you put in is what you get out; a recording of something that already happened and not something that can be reshaped in real time. Therefore samplers tend to be rather bad in producing sounds that react in real time to the players input.
Nowadays that disadvantage seems to have disappeared but that is only a superficial impression. By taking a lot of samples, often even of the same instrument at the same setting and levels but played ever so slightly different and reproduced after another the impression can be gained that a real instrument can be heard. In reality this however is a bit of a scam. In essence every sample layer is still a static snapshot.
Enter physical modeling. In the mid 1990ties computer speed and power had become affordable enough to design instruments that could calculate in real time what was expected from them. First on the market was Yamaha with the VL1. Later Korg followed with the Wavedrum, the Prophecy and later still the polyphonic Z1.
And then there was Technics, a brand normally known for its homekeyboards.
Modeling the Technics way Technics decided to do things a little bit different. Multitimbrality and sample realism had become sort of standard in synthesizer land. Also the concept of the workstation had established itself. The typical user expected an instrument that was a total music solution in itself. Virtual modeling synths where now possible but it was not yet practical to try and build a total workstation around the concept. Korg for instance failed with it's first try, the original, pre-Z1, Oasys (not to be confused wit the much later production version). In the early 90ties the price of such an instrument would still have been prohibitive.
So Technics went for a compromise. Now compromises normally spell trouble or at least average results but this time it led to one of the most flexible (and underrated) synthesis engines ever designed.
A typical model of an instrument normally includes a few stages:
1. The exciter If one plays an instrument there always is something that makes the sound happen. It can be a hammer striking a string, like in a piano, a finger or pick exiting a guitar string, a bow on a violin, a stick striking a drum skin, etc., etc., etc..
2. The resonator.
The exciter might make start the sound bu it is not the only factor determining the total sound. The rest of the instruments can be described as a so called resonator. This is the part of the mechanism that harmonically resonates with the exciter. It can be a pipe or an acoustic body made from a series of materials (wood, metal, etc.).The original intention behind such bodies was amplification (when electric amplification did not yet exist) but sooner or later is was found out that the total sound of the instrument also depended on the layout of the resonator body.
Those are the basics but what did the people at Technics do differently? In short: They still included samples in the concept. In stead of fully modeling a sound from scratch the Technics engineers still took samples as the starting point, They said: The prevailing trend is to use samples and treat them with a subtractive synthesis engine (filters, amplifiers, envelope generators, etc.) Why not still use samples, look at those as the exciters and stick a resonator program between these and a subtractive synthesis engine? Then the sound of the samples can be modified by a modeling algorithm and subsequently still treated by a subtractive end stage. One could call it a sampler Plus approach.
Sounds like a cheap copout at first but it actually made a lot of sense. It thus became more of a natural step then a revolution but one that combined all the best of (affordable) existing technology and simply added the advantages of modeling. They called the result acoustic modeling in stead of physical modeling.
BEEP! Something smelly is going on here. If one uses a sample as an exciter a lot of resonance information is already contained within that original sample (of a piano or whatever). So how could Technics claim they completely separated both aspects? Was that acoustic modeling name only commercial blurb then? It depends how you look at it. The name seems to indicate the new Technics synthesizer should be extra good at simulating existing acoustical instruments. Well, it rather was and it even offered a lot more.
So let's forget about the smaller details and concentrate on the fact that this indeed is a very impressive combination.
The most important difference is that although other physical modeling synthesizers have even more control over certain parameters then the WSA but since most stay so close to their original models (wood, string, brass, etc.) that they are rather bad in producing new, sonically interesting sounds.
Commercial disaster (?)
So in 1995 Technics jumped on the commercial trend that physical modeling was.
Shortly after that the whole trend went belly up and so did the WSA..
All this stuff was becoming way too complex for the typical commercial users. They had accepted new technologies (FM, sampling, etc.) as long as the presets became ever better but learning to program these digital beasts oneself was becoming a bit of a life's work. These synth engines where so complex and sophisticated that even the works programmers had their trouble getting the best out of these instruments.
Take the WSA factory sounds. Now they where very good indeed, they still sound good after all these years, but they hardly made use of all that was possible with this extremely flexible synth engine.
Fig.3: Modeling the Technics way The WSA1 and the WSA1R rack All brands found it difficutl to actaully sell these new virtual modeling wonder synths. On the surface most of them even seemed to offer a lot less then the previous generation (fewer voices, no multitimbrality, no inbuilt sequencers). And in the practice they didn't even sound much better.
But was that true?
On the surface it was. In the shop all these instruments seem to do about the same as a state of the art sample driven synth. To reap the advantages of physical modeling one must develop ones playing skills. In other words: The user has to really work hard to make such an instrument sing. If you want something to sound like a guitar, or a clarinet, or a trumpet you also have to think like a player of such instruments and modify your keyboard playing technique accordingly.
My God, so to ge the best out of such a contraption one not only needs a degree in computer programming and acoustical physics but also a classical music training including multi instrumentalist and arrangement skills. Surely only geniuses could do it all. Well that is a bit of an stretch but one needed to be rather dedicated to get the best out o such an instrument.
It's probably the most important reason everybody suddenly went back to analog synthesis. And thus a very promising trend ended. Let's get on whit real life. Money must be made.
We are still more or less living in the period that followed. The more analog (real or modeled)a synthesizer is the better its chances are of becoming a success. Hell, Korg even re-released its classic MS20.
Technics had an extra handicap. It was not recognized as a typical synthesizer manufacturer. It's reputation was built on hifi equipment, electronic piano's and homekeyboards. No “serious” musician would ever play a Technics.
So the WSA1 bombed. Technics cut it's losses and never built a programmable synth again. True? Only partly. In the real world the technology went into the KN-series keyboards which where among the best autoaccompanyment keyboards of the next years. So do not feel too sorry for Technics. Maybe they had to sell the remaining WSA's at costprice but they still made a very decent buck out of the technology.
Back to the future I also went back to analog synthesis with glee. In the mean time I did however never forget what modeling technology promised and sooner or later prices of second hand equipment became low enough to enable me to it out for myself at last.
Oh boy did I love what I found!. I was already very impressed by the Korg Z1 but found it to be rather unflexible. Oh, yes it can be very expressive but what I hoped for was an instrument that would enable me to blur the lines between synthetic and acoustic sound. One that would not only enable me to make analog type sounds like the real thing but also sounds that would help me to build my own original acoustic instruments.
Say a 100 foot guitar played with a hammer!. Wasn't that what they promised all those years ago?
Well, it's about time to move on to the actual Starship One project so I will not explain all the ins and outs here. Let my conclusion suffice.
The Technics WSA can do almost everything I can imagine a synthesizer should do. And it's synthesis engine is actually rather easy to program for a menu driven synth. Of course one knob per function would be better still but a WSA with such facilities would have a front panel as big as a room. Thanks to the big display and a logical menu structure one can however live rather well with the real world compromise provided.
And last but not totally unimportant: By waiting so long they have become dead cheap.
As far as I am concerned these beasties are the most underrated synths ever!