Within-subjects study in 31 vocally healthy adults: auditory, visual, and audiovisual room cues in immersive VR all measurably change self-perceived vocal loudness, effort, comfort, and acoustic output

Background

Voice therapy traditionally happens in a clinic room - quiet, acoustically dry, with no visible audience. The voice the client produces in that room is often very different from the voice they produce in the real-world settings where their voice problem actually matters (large rooms, noisy backgrounds, social or performance audiences). Acoustic simulation alone (reverberation, background noise) partly addresses this, but immersive VR offers something acoustic simulation cannot: a synchronised VISUAL environment that the client can see, including room size, perceived listener distance, and ambient context.

Whether the visual environment actually drives measurable vocal adaptations beyond the acoustic environment had not been systematically tested in immersive VR.

What the researchers did

31 vocally healthy adults (men and women) were tested under 18 sensory-input conditions in immersive VR. The 18 conditions were created by crossing:

• 2 auditory rooms with different reverberation times
• 2 visual rooms with different volumes
• audiovisual combinations of the two
• each with and without background noise

Each participant completed all 18 conditions, performing four speech tasks per condition: counting, sustained vowel phonation, an all-voiced CAPE-V sentence, and the first sentence of the Rainbow Passage.

Outcomes were self-perceived vocal loudness, effort, and comfort (each rated 0-100); plus objective acoustic measures - sound pressure level (SPL in dB) and spectral moments (spectral mean and SD in Hz, skewness, kurtosis).

What they found

• Self-perceived vocal loudness and effort INCREASED as room volume, speaker-to-listener distance, audiovisual richness, and background noise increased.
• Self-perceived vocal comfort DECREASED under the same conditions - the inverse pattern, consistent with the effort-comfort tradeoff.
• Objective SPL and spectral moments changed concomitantly - speakers automatically adjusted their voice to match the perceived room.
• Visual and audiovisual input - not just auditory cues - measurably shaped voice production. This is the central new finding: the visual environment in immersive VR is a meaningful therapeutic variable.

Why this matters

For voice clinicians considering immersive VR, this study establishes that the immersive visual context drives measurable changes in vocal output and self-perceived voice - BEYOND what acoustic-only simulation can achieve. Clinically, this means the choice of scenario in a VR voice-therapy session (small cafe vs. large auditorium vs. noisy classroom) is a therapeutic decision affecting expected vocal adaptations. The study is foundational evidence for voice-in-VR work that has since proliferated (e.g., Leyns 2025 RCT for gender-affirming voice training, Hoff 2026 voice meditation, Daşdöğen 2026 follow-on).

Limitations

• Vocally healthy adults only - clinical efficacy in voice-disordered populations is not directly tested.
• Single VR system - generalization to consumer hardware (Meta Quest) requires replication.
• 18-condition factorial optimises mechanism-testing over clinical-protocol validation; a clinically-grounded VR voice-therapy protocol is a separate translational step.
• No clinical outcomes (vocal endurance, voice handicap, patient-reported outcomes) - this is a controlled experimental study of voice perception and production mechanics.
• Sample size n=31 is adequate for within-subjects effects but modest for subgroup analyzes (e.g., by sex, by speaking style, by baseline vocal habits).