jwilzewske: AntiPatterns: Classical Banjo, Code Refactoring, and Short Time-scale Acoustical Physics

O ne has merely to hear the banjo’s strings touched, or one double-stop drawn out on the fiddle, to recognize the sense in which the meaning of bluegrass begins simply with what these instruments essentially are: it is the aim of bluegrass style to set forth what the instrument—or for that matter the person playing the instrument—essentially is.”
— Robert Cantwell, Bluegrass Breakdown, p. 203.

A thousand kilometers away, one of my nephews is tackling learning banjo. I give him Bullard’s book (link below) and begin listening to exotic recordings by Bela Fleck and John Bullard and others, in hopes of having mutually fun dialogues with my nephew, via email and face-to-face when I go and visit him.

T he banjo sounds on these recordings remind me of the sound—the ‘interiority’, the ‘earnestness’—of traditional shamisen or samisen (三味線, literally “three-taste strings”) music. Shamisen is similar to banjo in that the dō, the skin-covered box, is a lightly-damped resonator, much like the banjo’s. Both instruments are highly percussive and relatively ‘bright’. Both are plucked with the fingers or with a plectrum.

S o I go looking around to see what spectral analysis studies have been published about banjo or shamisen. I am curious to better understand the reasons that explain the unusual tone or timbre of each. Välimäki and coworkers spent a lot of effort on the matter in the mid-1990s (link below). Unlike guitar and other stringed instruments, for banjo the 5th harmonic (fourth overtone; 2P8+M3) and higher harmonics are rapidly attenuated, except for the 8th harmonic (3P8; 3 octaves above fundamental).

S olid body electric guitars don’t lose much string energy to the bridge. As a consequence they have a distinctive sound and are very long in sustain. Banjos on the other hand dump energy from the strings very easily to the rest of the instrument, and their characteristic sound is in large part related to that.”
— Liutaio Mottola.

T he directionality of the instrument’s sound radiation has a lot to do with banjo-ness, too, I think. Plucked string instruments have complex sound radiation patterns for various reasons. The resonant-mode frequencies of the box account for most of the radiated sound pressure. And different mode frequencies of the body have their own directional patterns, such as monopoles, dipoles, quadrupoles, and their combinations (see examples at UNSW, Joe Wolfe’s site, link below).

A nother distinctive factor is ‘masking’ caused by occultation of the sound by (and reflections from) the performer’s body. Masking plays an important role in environments where the listener (or the recording microphones) and the sound sources are positioned off-axis from each other or are freely moving with respect to each other.

T he influence of “X” vs. “Y” hand position for passages that involve the right-hand moving up the neck of the instrument is yet another factor, one I must remember to explore with my nephew when I next see him. The “X” position lends a bright timbre, and the “Y” position gives a rounder, darker quality. What the the player does with the right-hand on the head of the banjo also changes the timbre/color a lot.

A nd then there is picking technic. Fleck’s jazz-influenced single-string picking style lends an extra “impulsion” to the banjo timbre—a consequence of sympathetic acoustic coupling to and from the other un-plucked strings? The “impulsion” is one of the qualities that I am most attracted to in these recordings.

O h, the sheer exoticness of them, though! The sound calls into question the locale, the culture and politics, and the epoch of each musical text and of each of the speakers (composer, performer) and hearers. The ‘rurality’ of banjo timbre... the instrument’s architecture preserved even in the most refined, expensive ones of modern manufacture with technical polymer engineered-composite materials and high-tech strings.

T he ‘interiority’ of chamber banjo sound is radically different from the interiority of lute or harpsichord. If our own interiority is an object, then the music embodies that ‘objective’ world. Not ‘mimesis’ or ‘representation’; more like incarnation...

P hysical strings vibrate in both the vertical and the horizontal directions (well, actually, more like a lissajous curve in the plane perpendicular to the string’s length, so the polarization directions are changing and are only intermittently orthogonal to each other). If the effective length of the string is not the same in the two predominant polarization planes, we get mixing of the subsignals of slightly different frequencies, which in turn creates beat-frequency pulsations in the sound. This is a distinctive feature not only in banjo but in the Finnish kantele’s timbre as well, where the mechanism has to do with the peculiar way in which the strings have been terminated on the box.

V älimäki and colleagues emphasize that distinctive timbral properties of banjo are associated with the “attack”, the early transients in the waveform in the few tens of milliseconds immediately after the string is plucked or frailed. But there are wonderful, deeply banjo-ey aspects in the notes’ “decay” as well. We hear those in the Bullard and Fleck recordings. The ‘openness’ of these transcriptions of baroque, classical, and romantic compositions affords closer examination of those timbral spectral features than would ordinarily be possible to do in, say, idiomatic bluegrass and other banjo repertoire.

I t seems to me that these recordings of Bach and other composers’ works on banjo draw our attention to what neurophysiologists call “event fusion” (the formation of unitary musical events from the micro-evolution of waveforms on a time-scale of tens of miliseconds, too small to exhibit rhythm or voice-leading) and its inter-relationships with harmony and melody and rhythm and larger-scale form.

T hese banjo transcriptions and arrangements are unusual, musically speaking, but they register for me as very familiar in other ways. ‘Refactoring’ and ‘porting’ musical “code” from one performance “platform” to another—in the manner that Bullard and Fleck exemplify—feel to me very much like refactoring and porting computer code from one computing platform/language/operating-system/architecture to another. The process calls into question the [psychological; physical] conditions necessary for making large-scale—that is, ‘formal’—interfaces and system services-stack clear in music, just as the refactoring process does in software. The process in both cases invites or even demands reconsideration of the work’s design constructs and architectural components.

I n both, it vividly reveals the different levels of abstraction that are supported or not-so-well-supported by the different platforms. And it can uncover the essence of our interior selves, by making us pay attention to neural logic-races, multi-timescale communications phenomena, and hard-coded “antiPatterns” that the composer or performer (or listener) might not otherwise have noticed. So even if the target transcribe-to platform (banjo) manifests some antiPatterns of its own, it is still always healthy and illuminating to be outside our “comfort zones” and gleaning what we can from the refactoring process.

T he beauty and elegance of these classical banjo recordings make me want to revisit some of my own code that I have recently [re-]written; improve it; maybe understand it from novel and useful perspectives.

A ll sounds, all colours, all forms—either because of their preordained energies or because of long association—evoke indefinable and yet precise emotions, or, as I prefer to think, call down among us certain disembodied powers, whose footsteps over our hearts we call emotions.”
— W.B. Yeats, The Symbolism of Poetry.

H ow can we know the dancer from the dance?”
— W.B. Yeats.

jwilzewske

Saturday, January 8, 2011

AntiPatterns: Classical Banjo, Code Refactoring, and Short Time-scale Acoustical Physics

No comments:

Post a Comment