Singing in virtual environments: Exploring vocal adaptations to congruent audiovisual scenes of performance venues

This study investigated the vocal adaptations of fifteen soprano singers performing in virtual venues. The virtual reality (VR) system comprised a headphone-based auralization using room impulse responses and a head-mounted display presenting 360-degree still images, both sourced from corresponding real-world venues. Singers performed two songs with differing tempos in each venue, and their audio recordings, captured via a near-lips microphone, were analyzed alongside responses to a comprehensive survey. Results revealed subtle but notable effects of venue: larger spaces were associated with reduced pitch accuracy, brighter timbre, and slower, less pronounced vibrato, while the mid-sized purpose-built auditorium elicited louder singing and greater tremolo extent compared to the excessively dry or reverberant venues. These adaptations primarily occurred at the note level and differed between songs. Interestingly, singers' subjective evaluations often did not align with objective performance metrics, suggesting many of these adjustments were subconscious. The study reveals some subtle impacts of audiovisual scenes on vocal performance and highlights the potential of VR systems for exploring these phenomena. The inclusion of visual cues may have contributed to differences from previous research, warranting further studies on the interplay of audio-visual factors in shaping musical performance.