Cost-Effective Virtual Reality for Smarter Architecture: Predicting How We Hear Spaces
Angela Guastamacchia – angela.guastamacchia@polito.it
Department of Energy, Politecnico di Torino
Torino, Torino 10129
Italy
Popular version of 3aAAb4 – Subjective and objective validation of a virtual reality system as a tool for studying speech intelligibility in architectural spaces
Presented at the 188th ASA Meeting
Read the abstract at https://doi.org/10.1121/10.0037846
–The research described in this Acoustics Lay Language Paper may not have yet been peer reviewed–
When we communicate, clear speech is crucial—it helps us exchange ideas, learn, and build human connections. But often, poor acoustic conditions in rooms like crowded restaurants, wide lecture halls, or meeting spaces can make it difficult to understand speech clearly. Indoor architectural design significantly impacts speech clarity, so studying how different spaces affect communication, especially when hearing-impaired people are involved, is essential for fostering optimal designs that facilitate effective communication.
Virtual Reality (VR) might provide a practical and time-saving solution for this research, allowing us to reproduce various architectural environments and study how people perceive speech within those spaces without needing access to the real environments. Some laboratories have already implemented systems to accurately reproduce acoustics targeting diverse research goals. However, these systems typically rely on complex and costly arrays of dozens of loudspeakers, making studies difficult to set up, expensive, and inaccessible for architectural designers who are not VR experts.
Thus, a question arises: can even a less complex VR system still replicate a realistic experience of listening to speech in an actual room?
At the Audio Space Lab of the Politecnico di Torino, we set up a simpler and more affordable VR system. This system combines a VR headset with a spherical array of 16 loudspeakers to create immersive and realistic audiovisual communication scenarios surrounding the listener in a 360° experience, using an audio technique called 3rd-Order Ambisonics. We then tested whether our VR setup could consistently replicate the experience of listening in a medium-sized, echoey lecture room.
To test this, we compared the speech understanding of thirteen volunteers in the real lecture hall and in its virtual replica. During the tests, volunteers listened to single sentences and repeated what they understood across five different audiovisual scenes, varying the speech source location and the presence or absence of distracting noise. All scenarios included typical background noise, such as the hum of air conditioning, to closely mimic real-life conditions.
In Figure 1, you can see a volunteer in the real lecture room listening to sentences emitted by the loudspeaker positioned to their right, while a distracting noise is presented from the frontal loudspeaker. In Video 1, a volunteer performs the same speech test within the VR system, replicating the exact audiovisual scene shown in Figure 1. Figure 2 shows what the volunteer saw during the test.
Figure 1. Volunteer performing the speech comprehension test in the real lecture room.
Video 1. Volunteer performing the speech comprehension test in the virtual lecture room using the VR system.
Figure 2. Volunteers’ view during both real and virtual speech comprehension tests.
Our findings are promising: we found no significant differences in speech comprehension between the real and virtual settings across all tested scenes.
Additionally, we asked the volunteers how closely their VR experience matched reality. On average, they rated it as “almost very consistent,” reinforcing that the VR system provided a believable acoustic experience.
These results are exciting because they suggest that even with a less complex VR system, real-life-like speech perception in ordinary environments can be effectively predicted. Our affordable and user-friendly VR system could thus become a powerful tool for architects, acousticians, and researchers, offering an accessible way to easily study speech comprehension in architectural spaces and pursue improved acoustic designs.
Image courtesy of Rob Hamilton.