Uncovering the Sound of Immersion: EEG and Self-Reported Correlates of Diegetic Audio in Gaming

Immersion and flow are central to the gaming experience, yet their real-time physiological correlates remain elusive. This study examines the effect of sound, specifically diegetic music, on player immersion and flow by combining electroencephalographic (EEG) data from a commercial device (Muse 2) with self-reported questionnaires. Thirty participants played a custom platform-style video game under three sound conditions: consistent diegetic, inconsistent diegetic, and no audio. Post game self-reports measured flow, competence, challenge, and immersion in response to sound. Simultaneously, EEG features were analysed both across the four electrodes and considering their average. Machine learning analysis revealed key channel-feature combinations predictive of high immersion and flow states, whereas statistical tests aimed to identify features and channel-feature combinations that were statistically different across the sound conditions. Our findings suggest that, even with a low resolution EEG setup, it is possible to identify potential markers of immersion and flow that are worth further investigation.