On the utter death of musical arts

May 4, 2021 by Lucian Mogosanu

... where by "musical arts" we mean the crafts whose output consists of finite "acoustic objects" pleasant to the human ear; where an "acoustic object" is a thing in itself through its mere recognition by the human mind.

In other words, I am saying that the broad field of music-making has by Anno Domini 2021 reached the state of sheer rubble. And I can bet my ass that at least the late classics foresaw this utter perversion of the... natural? well, on some order of things!

In order to explain (first and foremost to ourselves) what happened there, let us first take a step back and a quick look at the realization that music itself is a sort of perverse phenomenon. Take, say, an archetypal Joe, the first human (and perhaps the last) to ever exist in this world: his ear and his geometric machine are connected in some way that we don't quite understand in all its intricate details; regardless, beyond all the neurons and whatnot, it's clearly observable that Joe has a measurable reaction to certain acoustic patterns, patterns which from Joe's point of view sum to more than just information, to more than a practical tool for survival. This is what Joe calls "music", and not only Joe calls it that from the point of view of the listener, but moreover he can produce sequences of discretely-observable sounds that himself and other humans may gather under the common label of "music" -- assuming, of course, there are other humans with whom he can enjoy this phenomenon.

Moving on, the hows of this music-making activity come from a long, thick string of technical considerations, starting, say, with a quick detour into the world of... guts. Say, those of a horse. They're an integral part of the normal functioning of the horse in question, and one'd expect our archetypal dude Joe knew this. Well, despite, or rather because of his knowledge of the inner workings of horses, Joe has learned to tie a piece of gut between two pieces of wood, which he uses to create nothing other than our beloved Music! Not to mention that he only went there sometime after he learned to use his own vocal cords for the job. He then found out that the tools he uses do indeed give some sort of distinctive "shape" to his music, through the very physical properties of the tools in question. This went on for a while, say, when Joe decided to create an abstract, symbolic, algebraic representation of music, or when he used electronics to make loud music, or music that sounds completely otherworldly, or whatever other musical pieces his human mind could conceive.

The means have changed on multiple coordinates, as sound synthesis, storage and processing were eaten by electronics; then some decades later this evolution was to be amplified in orders of magnitude by digital computers. As usual, the average Joes and Schmoes couldn't and can't keep up with the ocean of technology for making music, each with their scope, strengths and flaws, from the 44.1kHz CD standard to the shitton of compression formats, and from "phone apps" to expensive "professional" tooling. The means have changed: the DACs in your mobile phone can synthesize whatever information "you" will give them; the sad part being that it's not "you" who feeds the information to the DAC: you'd have no idea which bits go where, so you instead have to rely on Google/Steinberg/whomever the fuck's idea of what constitutes "the right approach to music making". The even sadder part being that the underpinnings of both Joe's rudimentary horse-gut string and of (post-)modern synthesizers are the very same and "you" are unable to see that, which makes "you" more fit for a piece of horse gut than a "mobile music-making app".

There will, I believe, come a moment when, along with the disappearance of "free knowledge for everyone and their dog", the entire process of "music-making" will be outsourced to computers, much like in Arjen's Pink Beetles in a Purple Zeppelin. Not only encoding and decoding musical pieces is cheap, at whichever level you can think of, but any off-the-shelf computer can easily generate enough unique musical pieces to last one a lifetime. Any gargle about "creativity" and whatnot will easily go down the drain, as the human ear is so easy to trick and the human mind so quick to forget. Actual professionals will have no choice but to (re)train their computer programming skills along with their grasp of music theory, which begins with Fourier transforms and the underlying physics and merely touches the Western system as a particular (yet predominantly relevant) case -- Bach's works remain written precisely as they are on paper, but let's not delude ourselves into believing that a bunch of MIDI files can capture their essence.

So, to end this on a more optimistic note: I've begun writing some code to "make music using computers", which I may share with you sometime. Maybe.

Filed under: asphalt.
RSS 2.0 feed. Comment. Send trackback.

4 Responses to “On the utter death of musical arts”

  1. #1:
    whaack says:

    My first common lisp project I wrote, mostly to get myself familiar with the language, was a synthesizer of sorts. I created a parser that would take a lisp list of notes and a tempo and play a corresponding sound. For example you could write (1/4 G3 A3 B3 C D E F# G) to play the G major scale. You could make chords by nested lists, i.e. (1/4 (A C E) (F A C)) to play an A minor triad followed by an F major triad.

    At first I played pure tones for the notes, but that sounded uninteresting, so I tried to write an equation that generated a sound wave that simulated the one generated by a plucked guitar string. This took me on a calculus and trigonometric adventure that I never finished with satisfaction. But the exercise showed how understanding music is tied to understanding our physical world. In order to create that sweet plucked nylon string sound, you need to have a solid grasp on the laws of physics that govern the movement of a vibrating string.

  2. #2:
    spyked says:

    This is very similar to what I have in mind: there's about half a century's worth of sound engineering that would very well fit in a single computer program. And I do mean everything, from musical composition (which would look like a domain-specific language of its own, not necessarily limited to Western notation, which was meant more for human than machine interpretation) to filters, mixers, modulators and synthesizers -- it's all just a bunch of textbook circuits/algorithms that'd go well along with some nice scaffolding to gather them together. The closest readily available solution I had used in the past was LMMS, but to be honest I'd rather steal the code from all those plugins piece by piece than try to build/use the thing again.

    It's not trivial to realistically emulate e.g. an acoustic guitar. I remember back a decade ago or so going through some heavy books on sound modelling, the differential equations were tough to understand, let alone implement correctly. So then you're stuck with at least two books to read, one on acoustic modelling and one on advanced numerical methods (and why not add some control theory in the mix while we're here), which is a great path to follow, except by the end of it you'll probably be left with very little time to make any actual music.

    To be honest, I sorta like 8-bit music and I'm a sucker for analog synths, so...

  3. #3:
    spyked says:

    While we're on the subject, I guess this doesn't look like such a bad reference for digital sound processing.

  4. [...] as my previous article on the demise of music-makin' ended on an overly optimistic note, not even a week passed and one anonymous reader made a point of [...]

Leave a Reply