by Per Vilez

Josep Comjuncosas is a rare combination of International prize winning pianist, electronic instrument designer and computer scientist. Josep has developed some complex instruments which pay exact attention to the musical and sonoric characteristics of it's acoustic version. One of these is the "Gong Exporer" created in the 'Sync Modular' environment. Josep addresses some important aspects of design using digital modular environments.

Your Csound and Sync Modular instrument designs are pretty involved and detailed. How do you approach instrument development.

It is essential to plan ahead. First you must have a good idea of what you are trying to achieve. This is the key to success. Making some kind of diagram which shows the fundamental design of the instrument is important. Normally I do a formal flowchart of the DSP involved. This is followed by a simplified but working version of the final instrument. This should be robust and work without a hitch. This can eliminate many errors and problems.

How do you personally use these within a composition framework?

I can be a purist in this respect and I prefer a minimum of tools. This minimalism allows me to know their environments throughout. I use Sync Modular, Csound and sometimes Reaktor. The composition and production takes place using an audio/midi sequencer. Ideally I would only use Csound, but this is not practical. Sometimes I get locked into the design of theinstrument, which I have a passion for and is a very beautiful art form, yet is very different from the composition process.

Being an accomplished pianist where do you see the place of physical modelling in terms of imitating acoustic instruments. Is it just the challenge of trying to realistically imitate the instruments or do you see them as compositionally useful?

As a pianist I can be very critical when comparing a synthetic instrument with a real instrument, however, it can unveil new horizons, even with the simplest design.

The virtual modelling of instruments is an interesting discipline in which scientists and artist can equally participate.

We are very near to realising virtual equivalents to the acoustic versions, but with a radically different domain of parameter editing, control and automation. A domain only offered by the computer.

I get very involved in playing the part of luthier. This can suggest many compositional possibilities, for example:

The real / virtual dialectic. The orchestration of hybrid instruments, not possible by acoustic means Performance techniques not realisable by humans.

The Grand Piano has been very well represented using multi-sampling techniques. We have seen that it is a matter of memory rather than anything else. Could physical modelling improve on this?

Not at present. Physically modelling a piano is a titanic endeavour - just imagine the number of simultaneous string models needed to create an accurate representation. The mechanical production of the sound and it's radiation are not fully understood, however, 'improve' does not necessarily mean the modelling of the piano note itself but should also encompass the expression of musical phrasing, response to interpretative gestures and the capacity to modify timbre. In this respect a real-time physical model would be superior to a sampled piano.

As I understand it, 'Sync Modular' is a reversed-engineered 'Reaktor'. What are the main differences between the two softwares.

At the user level I would cite three; Sync is more efficient and uses less resources than Reaktor, the GUI is simpler yet retains the same interactivity. You can also work DSP at a lower level, which is far more interesting when analysing a module such as an oscillator or filter. This way you can see how it works. Reaktor does not permit this. Reaktor can work as a DirectX plugin, something that I don't think Sync can at present, however, and it can work as a VST processor or Virtual instrument. There is a lot of expectation to see what Dr Sync (The creator of Sync Modular) will port to Reaktor now that he works for Native Instruments.