'Real' Instruments

Many people believe that synths can never sound as good as the 'real' thing (i.e. the acoustic instruments that they are emulating), and that they should not try. They reason that it's OK for a synth to sound like a synth, and that synth makers should leave well enough alone and not muck things up by doing poor emulations of acoustic instruments.

This perception, I believe, is the result of hearing too many poor emulations and unexpressive samplers. Once you hear a well-voiced physically modeled synth such as a VL1-m played by someone with some chops, you will quickly see that it can sound every bit as good as a 'real' acoustic instrument. Even considering a complex voice such as a saxophone, which is often played with a very heavy dose of expressive gestures, including all kinds of expressive attacks, decays, breathy sounds, growl, and so on -- the VL1-m can really do all of these things very convincingly! Now, of course, nobody's sampler can come within shouting distance of doing even a small portion of the expressive range of an acoustic sax. So, being accustomed to sample-based synths, you might think that this is a property of synths in general. You have to hear a good physically modeled (PM) synth played expressively, then you will know what I'm talking about. A good PM trumpet patch does not sound like 'synthy' trumpet, it sounds like trumpet, in all modes of expression that a trumpet can do. As an example, my favorite trumpet voice on my VL1-m has all of the timbral and attack characteristics of a real orchestral trumpet, including lip gliss, but it has a richer, nicer vibrato than any acoustic trumpet I have ever heard, and I can adjust the voice so I can blow soft, billowy attacks that cannot be done on a 'real' trumpet. To me it sounds better.

The VL1-m, the current reigning champion of PM synths, is not yet up-to-snuff on all of its acoustic emulations, but it is remarkably good at a great many of them. Some progress has been made in tweaking in and improving many of the patches or 'voices' as Yamaha calls them. I suspect that bowed strings will be the last bastion of resistance to emulation (not counting the human voice). The base algorithmic structure in the VL synths does not appear to be ideal for bowed strings, even though it does quite well with various types of plucked strings (which are much easier to model). Once you have heard a good sampling of the VL voices, you will have a good feel for the state of the art in PM synthesis, and you will see where the weak spots are and what is well in hand. The basic truth, however, is that PM, properly done, really does give you the capability to emulate acoustic instruments with full expressive control.

To the question of 'Why would you want to emulate acoustic instruments?', you will find that many people (including this author) really like the sound of various acoustic instruments, especially when played expressively. The reason that most synths sound so 'synthy' is that the simple waves like the sawtooth, sine, and triangle waves that form the basis for the early synths are easy to do, both with analog circuits and in a digital synth. The tinny sound and poor emulations that you hear in old sound cards and the cheapest synths are based on FM (Frequency Modulation) synthesis, also pioneered by Yamaha. The FM chips that Yamaha makes and everyone has used are inexpensive to manufacture, but they don't model the actual acoustics of traditional instruments. With sample tables, which are used in the 'deluxe' sound cards and conventional pro synths, you have photographic accuracy of whatever sound that was sampled, but extremely limited expressive capabilities. Even when you control layer & crossfade, envelope, filter, pitch-shift, and vibrato with conventional samples, you are nowhere near the expressive capabilities that are natural to acoustic instruments. The experience of this author is that most people who dislike synthesized music and prefer acoustic music are expressing a preference for that organic expressive control that is native to acoustic performances, and simply cannot be done with conventional synths. In the same breath let me say that conventional synths can emulate percussive instruments such as the piano and some others like organs just fine, and such synths often give the performing musician far more expressive capabilities than a 'real' piano or organ would have. This is where we are headed with PM instruments. We are getting the emulations tweaked in for all of the instruments whose notes cannot be merely triggered, but which must be 'played' through their duration, but we can also go far beyond that -- allowing greater range both in octaves, in tonal control, in attack and decay possibilities, and so on. Syncoustic hybrids can easily be fashioned (e.g. plucked flute layered with oboe), unusual 'things' can be crafted (e.g. a 24 inch diameter, 10 foot long clarinet), and so on, but in my view this is not the most important potential for physical modeling.

A tremendous potential in physical modeling is to improve the human control interface, and hence the expressiveness of the 'instrument'. In part, this gets outside the realm of PM and gets into controllers, but the PM tone generator allows the musician to control the sound in a very intuitive, simple, fluid, and organic way. My contention here is that the hallmark of a 'real' instrument is expressiveness, and PM instruments have that in spades. In many cases technology has given us cheap, sterile imitations of 'the real thing' (e.g. processed foods, artificial female bodies, imitation gold or chrome, etc.). In the case of PM, however, the technology has gone beyond earlier efforts, and you could call the PM synth 'the real thing' and the brass sax the 'cheap imitation'. After all, a brass instrument is only a tone generator, and one with severe limitations at that. All acoustic instruments that you see on stage are 100 % artificial creations of man. They are not more 'real' than an electronic box. Even a hollowed-out ram horn is an artificial instrument. None of them make music by themselves. They are only tools for the musician to use to add beauty and emotional expression to our lives. The questions for a musician are 'How do I like this tool? What can I do with it? What is it good for? What are its limits? When you dig into a good PM synth and evaluate it along these lines, you will find some nice answers, indeed.

The reason that most musicians don't play more than one or two instruments is that a considerable amount of time is taken to get and maintain your chops with a particular instrument. This time is a cost. With PM you have one control interface, which you may choose to suit your taste. You may add or modify it at any time (e.g. add a foot pedal or use a slider to modify a different parameter). It might be a guitar controller, a keyboard, a wind controller, a Buchla Lightning, Midivox, you name it. (With the first two listed you would want to add a breath controller.) You then develop your chops with that control setup, and that is all. You can do expressive playing of all of the instruments of the orchestra, plus many ethnic and popular instruments. There is a significant expansion of expressive power here for the musician, while economizing on the 'chop cost'. In this light it bears saying that using a breath controller and keyboard to control a VL1-m is a heck of a lot easier than playing a tuba, oboe, trumpet, or sax (take your pick). In other words, even if you only consider a single acoustic instrument, it is more difficult to play the acoustic model than the VL + controller. Bear in mind that these acoustic instruments are the result of many decades of refinement in design, materials, and workmanship to make them easier to play well, and the PM tone generator combined with midi controllers has them beat in this regard already.

So what is a 'real' musical instrument? One that is easy to control and express great music with! PM instruments stand at the top of this list.