{"id":57,"date":"2026-05-08T23:11:41","date_gmt":"2026-05-08T21:11:41","guid":{"rendered":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/paf-section\/04\/"},"modified":"2026-09-19T19:29:24","modified_gmt":"2026-09-19T17:29:24","slug":"04","status":"publish","type":"paf_section","link":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/paf-section\/04\/","title":{"rendered":"Time: Delay and Temporal Extension"},"content":{"rendered":"

4.1 Temporality and Predictive Structure<\/strong><\/h3>\n

If perception is fundamentally predictive, time is not merely a dimension in which experience unfolds: it is the medium through which prediction operates.<\/p>\n

Auditory perception is intrinsically temporal (Large & Jones, 1999). Unlike visual perception, which can stabilize a spatial configuration for extended inspection, sound unfolds irreversibly: each moment replaces the previous one, the listener cannot pause auditory flow, and anticipation is structurally embedded in the act of listening itself.<\/p>\n

Within predictive frameworks, temporal expectation operates at nested scales: micro-temporal (milliseconds, integration of acoustic features), meso-temporal (seconds, rhythmic and gestural continuity), and macro-temporal (extended duration, formal development and narrative anticipation). Contemporary dynamical models of rhythm perception formalize this nesting explicitly, describing entrained neural oscillations across multiple coupled timescales as the mechanism by which the auditory system generates temporal predictions and coordinates action to them (Large, Roman, Kim, Cannon, Pazdera, Trainor, Rinzel, & Bose, 2023). Music intensifies this temporal layering: it conditions listeners to expect pattern continuation, rhythmic recurrence, harmonic arrival, and structural resolution. Time in music is rarely neutral; it is directional, and it moves toward something.<\/p>\n

The introduction of delay structures alters this directional expectation in ways that are now well characterized in the temporal-prediction literature.<\/p>\n

4.2 The Linear Gesture in Conventional Musical Expectation<\/strong><\/h3>\n

In many tonal traditions, particularly within Western equal-tempered frameworks, musical gestures are organized linearly. A note or chord is articulated, progresses forward, and resolves into the next event: causality is clear, and event A leads to event B.<\/p>\n

This linearity reinforces predictive compression. Each gesture carries implicit forward momentum, and even in slow or ambient music listeners anticipate development or arrival. Delay, as a structural element, complicates this linearity: a delayed signal reintroduces past material into the present. The original gesture does not disappear; it returns, transformed by temporal displacement, and this reintroduction destabilizes the straightforward cause and effect chain.<\/p>\n

4.3 Delay as Temporal Multiplication<\/strong><\/h3>\n

From a structural perspective, delay creates temporal duplication. A single sonic gesture becomes layered across successive moments; the present contains traces of the immediate past; the acoustic event extends beyond its originating instant.<\/p>\n

This extension alters the temporal contour of perception in three concrete ways: it blurs the boundary between past and present, it reduces the finality of discrete events, and it creates overlapping temporal strata. This is not true of every delay configuration. A single delay line with feedback tuned to a metrically related interval reinforces a periodicity rather than multiplying it, functioning as entrainment in the ordinary sense described by dynamic attending theory. The diffusion effect described here applies specifically to multi-tap delay, non-integer or modulated delay times, and delay times short enough to interact with sound localization: below roughly 5 to 40 milliseconds, delayed copies of a signal fuse with the original into a single perceived event through the precedence effect rather than being heard as a separate temporal repetition at all (Litovsky, Colburn, Yost, & Guzman, 1999). In these configurations, atmospheric or non-percussive delay generates temporal diffusion rather than reinforcement: it multiplies competing periodicities rather than sharpening one. In such cases, prediction shifts from anticipating the next discrete event to maintaining coherence within an overlapping field of echoes. The auditory system continues to predict, but the object of prediction becomes less about singular events and more about sustained temporal texture.<\/p>\n

4.4 Temporal Extension and Causality Modulation<\/strong><\/h3>\n

One of the defining characteristics of delay is its modulation of perceived causality. In strictly linear sequences, cause and effect are temporally adjacent; in delay-rich environments, cause and effect become temporally distributed. The listener no longer experiences a gesture as a self-contained event but as the origin of a temporally extended ripple.<\/p>\n

This distribution of causal structure interacts with the entrainment mechanisms described by dynamic attending theory, which holds that the auditory system tracks time-varying events by synchronizing internal attentional oscillations to external periodicities (Large & Jones, 1999). These attentional oscillators are not fixed: recent work modeling their properties directly shows that a listener’s preferred oscillator rate and, critically, their flexibility, the ease with which they re-adapt when the rhythmic context changes, vary systematically between individuals and can be measured behaviorally by manipulating tempo across trials (Kaya, Kotz, & Henry, 2023). That study did not use delay; flexibility there is a property of the listener, estimated from how their internal oscillators re-lock to a new rate, not a parameter that a stimulus manipulates directly. What follows from it is more modest than a direct manipulation claim: overlapping delay periodicities plausibly place greater demands on this flexibility, since they offer no single periodicity to lock onto, and a listener’s individual flexibility is therefore a candidate moderator of how readily they relax a single dominant entrainment target in favor of a diffuse temporal field, an empirical question, not yet an established effect.<\/p>\n

Importantly, this does not imply confusion or cognitive overload. Structured delay, especially when continuous and non-abrupt, sustains temporal coherence while altering its density. In predictive terms, the distribution of prediction error across time becomes less concentrated in singular moments and more evenly dispersed across the temporally extended field the delay structure creates.<\/p>\n

4.5 Modulation-Based Techniques: Chorus, Flanging, and Phasing<\/strong><\/h3>\n

Delay is the clearest case of temporal extension, but it is not the only one used in this practice. A family of related techniques, chorus, flanging, and phasing, all built from short, continuously varying delay times mixed back with the original signal, extend the same logic into the domain of periodic spectrotemporal modulation.<\/p>\n

These techniques do not add discrete echoes; they impose a slowly moving pattern of constructive and destructive interference across the frequency spectrum, so that the same sustained sound is heard as continuously shifting in timbre. The auditory system has been modeled as analyzing incoming sound through a bank of filters tuned jointly to spectral and temporal modulation rates, a representation formalized in the multiresolution spectrotemporal model of Chi, Ru, and Shamma (2005) and supported behaviorally by perceptual-learning studies showing that training on a specific combination of spectral and temporal modulation rate produces narrowly tuned improvements in sensitivity to that combination (Sabin, Eddins, & Wright, 2012). Chorus, flanging, and phasing are, in this sense, manipulations of the representational dimension, joint spectrotemporal modulation, that this computational model proposes the auditory cortex encodes.<\/p>\n

This connects directly to predictive processing. Sohoglu and Davis (2020) showed, using MEG recordings of degraded speech, that spectrotemporal modulations are more strongly represented in neural responses than alternative representations of the same signal, and that the brain’s representation of those modulations is shaped by prior expectation in a manner consistent with prediction-error computation: expected modulation patterns are suppressed relative to unexpected ones. Slow, continuous modulation of the kind produced by chorus, flanging, and phasing therefore engages the same predictive machinery already discussed for delay, but applied to a continuously evolving spectral pattern rather than to discrete temporal repetitions. The effect on predictive density is structurally similar: a single, precisely anchored spectral prediction is replaced by a moving target, redistributing precision across a slowly evolving trajectory rather than concentrating it on a fixed timbral identity.<\/p>\n

4.6 Time-Varying Pitch-Shifting and Glissandi<\/strong><\/h3>\n

Pitch-shifting covers two distinct operations that this section keeps separate. A fixed-ratio pitch shift, transposing a sound by a constant interval, is a static transposition: it changes which pitch is heard but does not itself introduce a temporal trajectory, and the discussion below does not apply to it. Time-varying pitch-shifting, applied continuously so that the transposition ratio itself changes over the course of a sound, is a genuinely different case: rather than fixing an interval at an unfamiliar value, it moves pitch continuously along a trajectory over time, and it is this continuous case, equivalent in its perceptual structure to a glissando, that is addressed here.<\/p>\n

The auditory system’s handling of continuous pitch trajectories of this kind has been studied directly under a predictive framework. Henry and McAuley (2013) tested an auditory pitch-motion hypothesis using brief frequency glides in a laboratory discrimination task, showing that listeners judging a comparison glide use the velocity of a preceding standard glide to generate an active prediction about the velocity of the next one, producing systematic perceptual distortions when the two velocities differ. This is behavioral evidence, within that specific paradigm, that the auditory system does not merely track a moving pitch passively but extrapolates a predicted trajectory from prior motion and evaluates new input against that extrapolation.<\/p>\n

Extending this finding from brief, isolated laboratory glides to continuous, sustained pitch-shifting within an immersive musical passage is an extrapolation this paper makes explicitly, not a direct consequence of Henry and McAuley’s result. If the same velocity-extrapolation mechanism operates in that richer setting, sustained or irregular time-varying pitch-shifting, applied without a single stable target, would be expected to reduce the precision of any one predicted trajectory endpoint, in a manner structurally analogous to how overlapping delay reduces the precision of any one predicted temporal event. That expectation is plausible given the cited mechanism, but it has not itself been tested.<\/p>\n

4.7 Granular Synthesis and Time-Stretching: From Event to Texture<\/strong><\/h3>\n

A third family of temporal techniques, granular synthesis and time-stretching, operates at a different structural level than delay, modulation, or pitch-shifting. Rather than extending or modulating a discrete gesture, it fragments sound into very short grains, typically a few milliseconds to a few hundred milliseconds in duration, and reassembles them at a density and rate that can be decoupled from the original recording (Roads, 2001).<\/p>\n

At sufficient grain density, the result is no longer perceived as a sequence of discrete events at all: it is perceived as a texture, a continuous acoustic substance with statistical rather than event-level structure. McDermott and Simoncelli (2011) showed that sound textures, acoustic signals built from many similar, temporally dense events such as rain or insect swarms, are recognized by the auditory system through time-averaged statistics of peripheral auditory filter responses rather than through the tracking of individual constituent events. Their stimuli were natural or synthesized environmental textures, not musically granulated recordings of harmonic or microtonal material; whether densely granulated musical content is processed by the same statistical mechanism is an extension of their finding, not something their study tested directly. With that caveat stated, granular synthesis and extreme time-stretching construct a stimulus with the same defining property their model identifies: by decoupling grain density and playback rate from the source recording, they convert a formerly event-based gesture into something whose identity depends on statistical properties accumulated over time rather than on any single onset.<\/p>\n

This is the clearest possible instance, within the temporal axis, of the shift from event-based prediction to field-based continuity that Section 4.8 formalizes below: granular processing does not merely soften the edges between events, as delay and modulation do, but actively converts events into a field with a different perceptual structure. It sits, for that reason, at one structural extreme of the temporal axis, with unprocessed, sharply onset sound at the other, and with delay, modulation, and pitch-shifting occupying intermediate positions along the same continuum.<\/p>\n

4.8 From Event-Based Time to Field-Based Continuity<\/strong><\/h3>\n

Event-based prediction and field-based continuity, the two poles introduced across this section, describe two different ways perception can organize a stream of sound: around discrete units, each separately anticipated and resolved, or around sustained texture, tracked as a continuously evolving whole. Delay structures intensify the shift toward the second pole.<\/p>\n

When delay creates overlapping layers of sound, discrete edges soften: the listener is not only anticipating the next event but inhabiting a temporally thickened present. This thickened present reduces segmentation. Temporal segmentation is closely linked to narrative construction: when events are sharply delineated, narrative progression is pronounced; when events overlap and dissolve into one another, narrative linearity weakens. Within immersive sound practice, delay contributes directly to the transformation of time from sequence to continuum.<\/p>\n

4.9 Predictive Redistribution in the Temporal Axis<\/strong><\/h3>\n

Just as microtonality redistributes predictive density across intervallic possibilities, delay redistributes predictive density across temporal scales. Instead of anticipating singular moments of resolution, the listener shifts toward maintaining continuity within extended temporal textures. This redistribution corresponds to reduced temporal urgency, decreased emphasis on arrival points, and increased tolerance for durational ambiguity.<\/p>\n

These are not presented as universal or guaranteed effects: individual differences in oscillator flexibility, of the kind documented by Kaya et al. (2023), predict that listeners will vary in how readily they relax a single dominant entrainment target in favor of a diffuse temporal field. They are presented as plausible perceptual outcomes within a predictive framework, grounded in an established literature on temporal entrainment, spectrotemporal modulation, and pitch-motion extrapolation, when temporal extension is structured rather than chaotic. The crucial distinction lies in coherence: random or unstructured application of these techniques produces fragmentation, while delay, modulation, pitch-shifting, and granular processing integrated within an overall acoustic architecture support continuity instead.<\/p>\n

4.10 The Temporal Axis in the Context of PAF<\/strong><\/h3>\n

Delay, modulation-based techniques, time-varying pitch-shifting, and granular processing together, not any single one of them, constitute the temporal axis of the Perceptual Attenuation Field. It should be stated plainly that this grouping is a design-level, operational category, not a claim that the four techniques share a single underlying auditory mechanism. They do not: delay and its extreme forms engage entrainment and, at short lag times, the precedence effect; modulation engages joint spectrotemporal filter tuning; pitch-shifting engages velocity extrapolation over a frequency trajectory; and granular processing engages summary-statistics texture perception, a representational regime that does not track individual events at all. These are four distinct computational problems for the auditory system, studied in four largely separate literatures. What unifies them here is a functional description rather than a mechanistic one: each, through its own distinct route, redistributes predictive precision away from a single, sharply anchored temporal event and toward a more extended or more statistically defined structure. That functional description is this section’s claim; a shared underlying mechanism is not.<\/p>\n

The axes of this model are similarly not perceptually orthogonal, and this section’s own findings make that explicit rather than incidental. Delay times short enough to trigger the precedence effect are processed as a spatial cue, not a temporal one, which is precisely why Section 4.3 restricts the temporal-diffusion claim to longer or multi-tap delay configurations; the temporal and spatial axes overlap at that boundary rather than sitting in separate domains. Chorus, flanging, and phasing likewise operate through spectral interference and are, in a strict sense, frequency-domain techniques applied for a temporal purpose. Frequency, time, and space are used in this paper as operational categories that organize the practice and the routing architecture described in Section 9.4, not as a claim that the auditory system itself partitions perception along exactly these three lines.<\/p>\n

When combined with microtonal frequency modulation, this temporal axis deepens ambiguity not by increasing dissonance but by extending it across time: frequency destabilizes tonal anchoring, the temporal techniques described in this section destabilize temporal finality, and together they reduce the dominance of sharply bounded predictive endpoints.<\/p>\n

The temporal axis must, however, be considered alongside spatial modulation to fully account for its contribution to field formation: without spatial diffusion, temporal extension can still retain discrete localization. The following section addresses the spatial axis, reverberation and immersive continuity. Only through the integration of frequency, time, and space can the theoretical coherence of the Perceptual Attenuation Field be fully articulated.<\/p>\n","protected":false},"menu_order":4,"template":"","class_list":["post-57","paf_section","type-paf_section","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/wp-json\/wp\/v2\/paf_section\/57","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/wp-json\/wp\/v2\/paf_section"}],"about":[{"href":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/wp-json\/wp\/v2\/types\/paf_section"}],"wp:attachment":[{"href":"https:\/\/www.lovefrequencies.org\/microtonalia.live\/wp-json\/wp\/v2\/media?parent=57"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}