Sunday, February 05, 2006

Music, Noise, Acoustics, Electronics, and the Future, Epilog

With the situation I described at the end of that previous post, the technology begins to reach the point where the loop will close and the journey will have come full circle. If the performers of the acoustic music and noise-art elements are all seamlessly connected to the computer and are in fact controlling all of the electronic music and noise-art elements, the computer and attendant electronic elements vanish into invisibility and are only another audible aspect of the music/noise-art amalgam.

This is God-only-knows how far into the future, because the technology to make it happen perfectly every time with zero intrusiveness upon performers of traditional acoustic instruments would have to be so advanced as to be "virtually indistinguishable from magic" to us, as Sagan used to say.

To give an idea of how difficult a job this would be, let's consider the interface(s) used to connect a guitar to a computer. The technology for interfacing a guitar to a computer or synthesizer is, with a few exceptions, the same today as it was when Roland and ARP came up with the idea: A pickup with six individual magnetic sensors - one for each string - is used to sense the pitches. This is an incredibly poor idea because a guitar string - especially a steel one - has a large, unstable, and highly variable amount of harmonic content to it. As a result, filtering programs have to be employed to sort through this over-abundance of information and determine what the fundamental is. Add to this that guitarists use all manner of expressive nuances such as vibrato and pitch-bend, and the difficulty of accurate interpretation simply compounds. This gordian knot was so difficult to untie - and impossible to cut through - that one manufacturer of synthesizers, ARP Instruments, was actually put out of business by all of the problems encountered with it's Avatar guitar synthesizer.

Since I was a pioneer in this area, I was around when all of this happened, and I can testify to the fact that early guitar synths were impossibly problematic: The interfaces were so easily "confused" that the vast majority of guitarists could not use them at all. One had to modify one's technique to minimize mis-cues by playing cleanly and deliberately; you couldn't just pick up the axe and flail away and expect it to work. When a note on the guitar decays, sometimes upper harmonics actually become more prominent than the fundamental, so the units would often begin to track those (I called it the "Arabian Melody Syndrome" because the resultant "melodies" reminded me of something a snake charmer might play).

This was circa 1980, and here we are over a quarter of a century later (!) and the Roland hexaphonic pickup is almost exactly the same unit it was when I was a Synclavier Guitarist back in the mid to late 80's. Guitar synths have become more playable mostly through a combination of designing "dead" guitars (Guitars which generate less than stellar harmonic content) and improved pitch discrimination algorhythms.

Of course, a hexaphonic magnetic pickup is a non-starter for a flamenco or classical guitarist who plays on nylon strings, so Richard McLish of RMC Pickup (No Wiki entries) designed the Polydrive, which my Godin Multiac Grand Concert Synth-Access electric nylon string guitar uses. BUT, this is just another hexaphonic pickup that uses contact transducers to sense the string vibrations: Same concept as the Roland design, but executed with transducers. I actually got it to create hexaphonic distortion - which sounds clear even with complex harmonies - and not to control a synthesizer (Though nylon strings and contact transducers make tracking slightly less problematic when compared to steel strings and electro-magnetic pickup systems). Basically, I'm not going to return to guitar/synthesizer combinations until the technology is advanced enough that the interface is "invisible" and every nuance of my playing is accurately translated. That may be never for me, as a forty-eight year old, with the current pace of "progress."

Over the decades I've seen lasers used in an attempt to obtain an instantaneous and accurate signal for interfacing instruments to computers - and that may be one of the proper paths when the tech becomes ubiquitous enough - but I have changed radically from the optimist I used to be about these possible advances into the... ah... "less than sanguine" agitator I am today.

To be truly transparent and totally invisible, the technology is going to have to "devolve" to using a simple, high quality microphone (Or any pickup), of any standard design type, and the computer is going to have to be powerful and accurate enough to translate that signal in to whatever polyphony it receives and translate every nuance virtually instantly.

Perfectly. Every time. No exceptions.

The idea just seems elementally simple to me, but of course it isn't.

Total. Transparency.


Post a Comment

<< Home