Adaptive Radio System Using MetaSounds for Serious VR Game

Background: Adaptive audio can deepen immersion in VR, yet many pipelines still rely on fixed playlists or ad-hoc triggers that break narrative coherence during rapid context shifts in serious games. Methods: We implement an adaptive in-game radio using Unreal Engine MetaSounds and a finite-state machine (FSM). Three affective stations (Happy, Excited, Intense) drive procedural ambient music with equal-power crossfades, loudness normalization, and optional beat-synchronous transitions. We integrate Blueprint–MetaSounds graphs and evaluate usability and perceived immersion with 30 participants in VR. Results: The interface was rated intuitive by 90% of users; 85% reported higher immersion than a non-adaptive baseline; 80% felt the adaptive music heightened emotions at key events. Overall satisfaction averaged 4.5/5, and 70% indicated replay intent to explore audio variants. We provide the audio graph, state logic, and transition envelopes, and quantify transition smoothness via envelope and loudness metrics. Conclusions: An engine-native MetaSounds–FSM pattern is practical to deploy and can improve engagement in serious VR contexts. We discuss design trade-offs, limitations (sample size and demographics), and avenues for broader deployment in training and therapeutic simulations.