Assessing realism in virtual architecture: Neural and subjective responses to changes in level of detail

• EEG and surveys assessed responses across a fixed progressive LoD sequence in VR • The largest realism gain occurred at LoD2→LoD3 (materials/texture/color cue bundle) • Condition-dependent EEG differences were concentrated at the LoD2→LoD3 transition, particularly in frontal alpha and occipital beta • EEG differences relative to the real-image reference were largest at low LoD and were reduced at higher LoD • EEG-derived valence/arousal proxies showed weak coupling with subjective realism In this study, we investigated how neurophysiological responses and perceived realism vary across a fixed, progressive increase in architectural level of detail (LoD) in virtual environments. Using electroencephalography (EEG) and subjective survey data, we conducted a two-part experiment with five LoD levels and benchmarked each virtual scene against a single real-world reference image (RI) of the same scene and viewpoint. Results showed that perceived realism increased with LoD, but not linearly. The LoD2→LoD3 transition showed the most pronounced change in perceived realism, coinciding with the introduction of material, texture, and color cues. Across conditions, frontal alpha asymmetry (FAA)-based valence proxy indices were predominantly negative at LoD1–LoD2 and shifted toward predominantly positive values from LoD3 onward, with LoD5 maintaining positive values as semantic and contextual cues were added. EEG frequency-band analyses indicated significant neural differences, particularly in frontal and occipital regions, concentrated in earlier LoD transitions; these patterns are discussed as potentially consistent with changes in processing demands and attentional engagement. At higher LoD levels, EEG patterns showed fewer condition-dependent differences under the present analysis. Correlations between EEG-based valence/arousal proxies and subjective realism ratings were weak, suggesting a temporal and cognitive gap between immediate neurophysiological responses and subsequent evaluations of realism. Overall, these findings indicate that progressive LoD cue-bundle additions within the fixed presentation sequence were associated with improved visual fidelity and systematic shifts in EEG response patterns and perceived realism. These findings may support efficient level-of-detail management and user-centered visualization design in architecture, extended reality environments, and digital twin urban simulations.

Authors

Institutions

Publication Details

Journal
Brain and Environment
Published
2026-04-14
DOI
https://doi.org/10.1016/j.braen.2026.100019
Primary Topic
Architecture and Computational Design
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Assessing realism in virtual architecture: Neural and subjective responses to changes in level of detail

Jongho Lee, Dongwook Sohn, Soyeong Kim, Kyujin Choi et al.
Brain and Environment
Architecture and Computational Design
article

Assessing realism in virtual architecture: Neural and subjective responses to changes in level of detail

Jongho Lee, Dongwook Sohn, Soyeong Kim, Kyujin Choi, Uemee Jung
article en

Abstract

• EEG and surveys assessed responses across a fixed progressive LoD sequence in VR • The largest realism gain occurred at LoD2→LoD3 (materials/texture/color cue bundle) • Condition-dependent EEG differences were concentrated at the LoD2→LoD3 transition, particularly in frontal alpha and occipital beta • EEG differences relative to the real-image reference were largest at low LoD and were reduced at higher LoD • EEG-derived valence/arousal proxies showed weak coupling with subjective realism In this study, we investigated how neurophysiological responses and perceived realism vary across a fixed, progressive increase in architectural level of detail (LoD) in virtual environments. Using electroencephalography (EEG) and subjective survey data, we conducted a two-part experiment with five LoD levels and benchmarked each virtual scene against a single real-world reference image (RI) of the same scene and viewpoint. Results showed that perceived realism increased with LoD, but not linearly. The LoD2→LoD3 transition showed the most pronounced change in perceived realism, coinciding with the introduction of material, texture, and color cues. Across conditions, frontal alpha asymmetry (FAA)-based valence proxy indices were predominantly negative at LoD1–LoD2 and shifted toward predominantly positive values from LoD3 onward, with LoD5 maintaining positive values as semantic and contextual cues were added. EEG frequency-band analyses indicated significant neural differences, particularly in frontal and occipital regions, concentrated in earlier LoD transitions; these patterns are discussed as potentially consistent with changes in processing demands and attentional engagement. At higher LoD levels, EEG patterns showed fewer condition-dependent differences under the present analysis. Correlations between EEG-based valence/arousal proxies and subjective realism ratings were weak, suggesting a temporal and cognitive gap between immediate neurophysiological responses and subsequent evaluations of realism. Overall, these findings indicate that progressive LoD cue-bundle additions within the fixed presentation sequence were associated with improved visual fidelity and systematic shifts in EEG response patterns and perceived realism. These findings may support efficient level-of-detail management and user-centered visualization design in architecture, extended reality environments, and digital twin urban simulations.

Brain and EnvironmentVol. 7
Yonsei University (KR), Korea Institute of Civil Engineering and Building Technology (KR)
Korea Institute of Construction Technology, Ministry of Science and ICT, South Korea
Openalex Percentile: Top 12%
Architecture and Computational Design
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.