Musical Idea 2: Diegeticity
10.08.2013 § 1 Comment
In film theory, screen sound can be either diegetic or non-diegetic. Diegetic sound is understood to exist within the film’s world: words spoken by the characters to each other, footsteps or any other retained profilmic sync-sound, etc. Non-diegetic sound is understood to not exist within the film’s world: soundtrack, voice-over/narration/commentary, etc.
I submit that we can apply this film sound theory to music, too.
Let’s begin by framing a familiar example. Consider when a rock song fades out. We know that the performers didn’t play softer and softer until their guitars and drums were inaudible, nor did they step farther and farther away from the mics until the mics could no longer pick anything up. We know this not just because it’s silly to think that the band would do it this way, in light of our familiarity with recording or post-processing technologies; we know it also because playing either softer or playing from farther away would have created sounds that were not merely quieter: the timbre and reverb, respectively, would have been altered. Traditional song fade-outs are non-diegetic, as they are achieved essentially by turning a knob, controlling an entire world of sound from outside of it. I hear that one of the first examples of a fade-out, though, for the last track of Holst’s Planets album, was achieved by slowly closing a door on the chorus in another room — now that’s what I call diegetic.
It’s not uncommon to see ambiguity, transition, or reveals between diegetic and non-diegetic sound in the cinema, whether deployed to artsy or jokey effect. A biopic slips from its titular pianist’s name-making performance into a montage of his touring life scored by a continuation of the music he was performing in the original scene. Or in Blazing Saddles, a character is introduced with big band music as he rides across the prairie — only for it to be revealed that Count Basie has apparently, absurdly, actually been here on the prairie just off screen playing this music with his band. Perhaps it’s because it’s lacking the visual anchor, but you don’t see this type of play in music near as much; imagery can carry such helpful information for classing sound as diegetic or not. But there is tremendous potential here — even more so because of the challenge and purity in classing the sounds using sound alone.
Suppose, for instance, some music involves a vocal refrain that alternates between lower and higher pitch. Sometimes, recordings of the vocalist having actually sung at a different pitch are used. Other times, after-the-fact pitch-shifted recordings are used. These two methods affect the sound profile in recognizably different ways, and are associated with different possible artifacts. The singer can only sing so high or low until her voice cracks and breaks. The digital effect gets choppy if you stretch a file too much and sounds robotic. Or you could have a performer render a melody in retrograde versus reversing the recording. Or it could be as simple as whether the singer is repeating a phrase over and over himself, with the infinitesimal organic variations that entails, or a sample is on loop. Etc. It doesn’t have to be black or white, either — use achieve desired pitch with varying ratios of diegetic and non-diegetic transposition — or even drift fluidly between those ratios.
And it goes without saying that with music, you’re not limited to one diegesis, right? Or that these multiple diegeses can be made ambiguous between each other in all sorts of ways, too? This is only the beginning.