---
title: Breaks in Presence in Immersive Media
url: https://www.emergentmind.com/topics/breaks-in-presence-bips
type: topic
---

# Breaks in Presence in Immersive Media

Breaks in Presence (BIPs) are disruptions in the experiential state of presence in virtual, mixed, and other immersive environments: episodes in which the user ceases to remain fully absorbed in the simulated or mediated world and becomes aware of physical reality, the medium itself, or a breakdown in the coherence of the experience. Recent work treats BIPs not as a single uniform event class but as a family of temporally structured interruptions spanning perceptual illusion, embodied action, network performance, social interaction, and phenomenology. In this literature, BIPs are relevant both because presence enables immersive systems to produce consequential experiences and because presence is intrinsically unstable: it “may (and always) break” [2405.05926].

## 1. Conceptual foundations

A central starting point is the definition of presence as “a perceptual illusion of non-simulation” and/or “non-mediation,” under which “technology users to some extent do not recognize the role of the technology in their experiences” [2405.05926]. On this view, presence is not inherently positive or negative. It is “about the effect”: mediated events become experientially immediate enough to matter, which is why the same mechanism can support both effective therapy or training and intensified harm.

Within that framework, a BIP is the disruption of this illusion of non-simulation or non-mediation. The conceptual literature on immersive harms makes the point directly: presence is not continuous, and the fragility of presence creates a boundary problem between an intensely presenceful negative experience and a genuine break in presence [2405.05926]. This boundary question is not merely taxonomic. It bears on whether immersive systems are amplifying an event within presence or interrupting presence itself.

A more explicitly phenomenological formulation defines BIPs in awareness terms: they are moments when a VR user becomes consciously aware of the reality outside VR or of the medium itself, and an episodic BIP extends from the perception of a distractor through possible re-immersion [2604.09146]. This awareness-based view also sharpens an important distinction: not every distraction is a BIP. Some potentially disruptive events are perceived but incorporated into the virtual world and therefore do not break presence.

## 2. Presence theory, PI/Psi, and the structure of disruption

Recent work commonly analyzes BIPs through Slater’s PI/Psi model, in which Place Illusion (PI) is the illusion of being physically located in the virtual environment and Plausibility Illusion (Psi) is “the illusion that what is apparently happening is really happening” [2604.09146]. On this account, BIPs can be breaks in PI, breaks in Psi, or mixed episodes involving both.

Micro-phenomenological analysis of 57 BIP episodes found 30 episodes classified as break in PI, 9 as break in Psi, and 17 with indicators of both; in the mixed cases, indicators of a break in Psi preceded indicators of a break in PI [2604.09146]. This suggests that incoherence or implausibility may first destabilize the scenario, after which awareness of physical reality or mediation intrudes more fully. The same study also treated one “mind wandering” episode separately as possibly a break of attention rather than a BIP in the strict PI/Psi sense, reinforcing that attentional drift and BIPs are not identical.

Neural work extends this theoretical distinction into an online perturbation paradigm. In a preregistered ERP study, BIP trials were induced by freezing the expected release of a slingshot ball for 1000 ms at the moment of action, thereby violating the action–consequence mapping while the rest of the simulation continued [2509.01420]. The authors interpret this within a predictive-processing account: low-level sensorimotor contingencies support PI, and violations of those contingencies generate prediction error that can manifest phenomenologically as a BIP. Their results dissociate components: the N2 was more negative for BIP than normal trials only in the embodied condition, whereas the N400 was more negative for BIP than normal trials in both embodied and non-embodied conditions [2509.01420]. This supports a layered account in which not all presence disruptions are of the same kind.

## 3. Triggers and mechanisms

A persistent misconception is that BIPs are only technical faults. The recent literature is broader. In wireless VR, BIPs are modeled as consequences of excessive transmission delay, degraded VR video quality, and inaccurate tracking information at the base station, with additional sensitivity to body orientation, blockage, and mmWave directional links [1812.01202]. In this engineering setting, the paper formalizes BIP at two levels: a wireless-induced binary event indicator \(\omega_{it}\), triggered when combined uplink/downlink delay exceeds \(\gamma_D\) or delivered quality falls below \(\gamma_Q\), and a broader application- and awareness-aware metric \(P_i(\cdot)\), interpreted as the average number of BIP that user \(i\) can identify during a period [1812.01202].

The conceptual safety literature widens the trigger space beyond networking and system glitches. Because presence, embodiment, and immersion are moderating variables rather than valence-bearing properties in themselves, harmful content can become genuinely harmful while presence is maintained; yet “some disturbing content can break presence” [2405.05926]. This makes content and social meaning potential causes of BIPs alongside technological causes. The same paper uses Social VR harassment as the main example: greater embodiment, presence, and immersion can lead to more intense experiences of harassment, but the paper does not reduce that harm to a BIP. Instead, it asks where the line lies between harmful experiences and breaks in presence [2405.05926].

Mixed-reality distraction studies add another mechanism: cross-world attentional switching. In a within-subject HoloLens 2 study, congruent distractions were handled inside the virtual task context, whereas incongruent distractions required acting on a physical real-world button outside the virtual environment [2411.05275]. Both distraction types reduced presence, but incongruent distraction had the larger effect, supporting the idea that interruptions requiring explicit re-engagement with physical reality are especially BIP-prone. The paper is careful, however, that it inferred BIP from presence reductions rather than directly measuring discrete BIP events [2411.05275].

## 4. BIPs as lived episodes

Phenomenological work argues that BIPs are not best understood as instantaneous “snaps out” of VR but as durative episodes with internal temporal organization [2604.09146]. In a study of 14 users of a height-exposure application, 57 BIP episodes were modeled, averaging 4.15 BIPs per participant, with a mean of 7.23 diachronic moments per episode [2604.09146]. Of those episodes, 37 were triggered by planned distractors and 20 occurred spontaneously.

The generic diachronic structure extracted from these episodes comprises eight top-level moments: perceiving something, noticing something abnormal, reacting, conducting an inquiry, knowing, implementing a VR comeback strategy, telling oneself it is not real, and getting back to reality [2604.09146]. Four generic patterns were identified: reflected-upon, discarded, self-preservation, and contradictory mediation BIPs. The most common was the reflected-upon pattern, with 36 of 57 episodes [2604.09146].

Awareness shifts recur strongly across this phenomenology. Reality awareness occurred 46 times across all 14 participants, experimentation awareness 45 times across 13 of 14 participants, and media awareness 29 times across 12 of 14 participants [2604.09146]. This supports the awareness-based definition while also showing that BIPs need not be simple transitions from one fully coherent state to another. Some are investigated, some are discarded, some are used for coping, and some leave the user partially “between worlds.”

## 5. Measurement and operationalization

BIPs have been operationalized through markedly different methods: retrospective drawing of presence trajectories [1905.05673], formal wireless-performance models [1812.01202], trial-locked ERP perturbations [2509.01420], and behavioral-proxy approaches in mixed reality [2411.05275].

| Approach | Operationalization of BIP | Selected findings |
|---|---|---|
| Post-experience drawing | Downward deviations in a drawn presence-over-time curve | Drawings were consistent; the method detected all introduced BIPs at the study level |
| Wireless VR modeling | \(\omega_{it}\) and \(P_i(\cdot)\) from delay, quality, application, and awareness | BIP minimization coupled networking, prediction, and user association |
| ERP induction | 1000 ms freeze at action release on 20% of trials | N2 was embodiment-dependent; N400 tracked BIP trials in both conditions |
| MR distraction proxy | Presence drops, cognitive load, and reaction time under distraction | Reaction time covaried strongly with reduced presence, but BIPs were inferred indirectly |

The retrospective drawing method was proposed specifically because questionnaires do not capture temporal variation and continuous in-experience measures may themselves disturb presence [1905.05673]. Users draw presence over time on a paper template; BIPs appear as drops toward the real-world side. The descriptive model formalizes points such as \(P_{transition}\), \(P_{experience}\), \(P_{mentalexit}\), \(P_{physicalexit}\), \(P_{return}\), \(P_{dropping}\), and \(P_{break}\), together with parameters such as \(t_{transition}\), \(t_{dropping}\), \(t_{raising}\), and \(sh_{break}\) [1905.05673]. In the exploratory study with \(N=30\), drawings were structurally consistent, the method detected all introduced BIPs at the study level, and among 118 detected breaks, 97 were positioned correctly relative to the corresponding task tick [1905.05673].

The ERP approach aims at online measurement without relying on retrospective report. In the slingshot task, embodiment questionnaire scores were higher in the embodied than non-embodied condition (median .73 versus .50), and exploratory Slater-Usoh-Steed scores were also higher (median 5.33 versus 4.58), validating the manipulation [2509.01420]. The authors therefore treat N2 as the strongest candidate online marker of PI-related sensorimotor mismatch, while cautioning that P3b is better interpreted as information extraction or context updating and N400 as broader incongruity rather than a clean, unitary presence index [2509.01420].

Mixed-reality distraction work instead evaluates whether reaction time can serve as a practical proxy for disrupted presence. In that study, combined presence means were \(4.65\) in the no-distraction condition, \(4.34\) with congruent distraction, and \(4.19\) with incongruent distraction; NASA-TLX means were \(1.62\), \(2.92\), and \(4.03\), and reaction times were \(3.90\), \(5.50\), and \(7.88\), respectively [2411.05275]. The reported regressions were \(\text{Presence} = 5.64 - 0.43 \cdot \text{Cognitive Load}\) and \(\text{RT} = 13.87 - 1.78 \cdot \text{Presence}\) [2411.05275]. The paper’s main caveat is decisive: “significant drops in presence scores were used to infer Break in Presence (BIP), given the absence of a reliable and validated BIP measurement tool” [2411.05275]. Reaction time therefore functions as a correlate of BIP-proneness or disrupted presence, not as a validated detector of individual BIP episodes.

## 6. Harm, design, and unresolved questions

BIPs matter not only for usability but also for safety and ethics. The recent conceptual harm literature argues that presence enables simulated harms to register as experientially real, while simultaneously asking where the boundary lies between a harmful immersive experience that remains presenceful and one that becomes a BIP [2405.05926]. The distinction is consequential: severe harm can occur without causing a break in presence, and disturbing content may either intensify harm within presence or rupture presence instead. Social VR harassment, traumatic content, and stressful content are the primary examples used to frame this threshold problem [2405.05926].

For system design, this implies that maximizing presence cannot be treated as an unqualified good. Safety work must address content and social behavior, not only technical fidelity, because some harms are intensified precisely when presence works well [2405.05926]. In mixed reality, the distraction literature points in the same direction at the interaction level: congruent interruptions are less presence-disruptive than incongruent ones, suggesting that system notifications and secondary tasks should remain as integrated with the virtual task context as possible [2411.05275]. In wireless VR, the same concern appears as an optimization objective: federated deep ESN prediction of user location and orientation, combined with association control, reduced users’ BIP by up to 16% and 26% compared to centralized ESN and deep learning algorithms [1812.01202].

Several open questions remain unresolved. The phenomenological literature proposes an awareness-based definition and three BIP-related design opportunities, but very few studies characterize how BIPs are actually lived and managed [2604.09146]. Neural work suggests that different ERP components track different layers of disruption, yet it does not independently manipulate PI and Psi [2509.01420]. The harms literature poses, but does not answer empirically, whether the decisive threshold is set by level of immersion, level of harm, embodiment, context, user differences, or their interaction [2405.05926]. A plausible implication is that immersive systems may sometimes benefit from intentional, protective interruption of immersive intensity under harmful conditions, but that remains an inference rather than an established design rule [2405.05926].

Taken together, current research positions BIPs as a cross-cutting problem of immersive media. They are at once perceptual ruptures, lived episodes, network-level failures, behavioral slowdowns, neural mismatch events, and safety-relevant boundaries in the moderation of immersive harm. The field has moved beyond treating BIPs as mere glitches, but it has not yet converged on a unified operational measure or on a settled account of when a disruption counts as a break in presence rather than a change within presence.

Source: https://www.emergentmind.com/topics/breaks-in-presence-bips