Photo by Roxanne Joncas on Unsplash
Project Overview: Tailoring Audio to the User’s Space
At Meta Reality Labs, I contributed to the Room Acoustics project, a vital initiative focused on enhancing the realism of immersive audio experiences. The goal was to evaluate user preferences for different algorithms designed to map and reproduce room acoustics. By accurately simulating how sound interacts with a user’s physical environment, we aimed to make virtual audio virtually indistinguishable from reality, significantly boosting immersion within Meta’s products.
Methodology: User Preference Evaluation
This study involved users evaluating various acoustic algorithms in a controlled setting:
- Test Environment: Participants wore a Quest 3 headset and used a custom application.
- Audio Generation: The app played sounds through both the headset and a physical loudspeaker to create varied acoustic scenarios.
- User Ratings: Users were prompted to rate these sounds based on:
- Realism: How authentic the sound felt.
- Plausibility: If the sound seemed recorded in their room.
- Overall Audio Quality: The general fidelity and pleasantness of the sound.
My Role & Contributions: Driving Evaluation and Refinement
Working collaboratively with UX Researchers and Software Engineers, my responsibilities were critical to the study’s execution and iterative improvement:
- Sound Sample Testing: I meticulously tested sound sample clips across different room conditions to help engineers achieve audio that was as virtually indistinguishable from reality as possible.
- Coordination & Moderation: I effectively coordinated lab and Research Assistant availability to ensure smooth study operations. I also moderated user study sessions, guiding participants through the evaluation process.
- Continuous Feedback & Improvement: I provided continuous feedback during data collection to the team, directly contributing to the ongoing improvement and refinement of the study design.
Impact: Validating User-Preferred Acoustic Models
While iterations of this study were ongoing at the end of my contract, the completed evaluations yielded clear, impactful results:
- User Preference for Enhanced Algorithm: Users consistently preferred the latest enhanced algorithm developed by the audio team.
- Desired Dimensional Mapping: The study validated that users strongly desired how this preferred algorithm mapped the dimensions of the room and adjusted sound accordingly, confirming the value of accurate acoustic reproduction.
- Informed AI Development: These findings directly informed the development and refinement of Meta’s room acoustics models, ensuring that future audio experiences are designed around validated user preferences for realism and quality. My work helped ensure that the engineering efforts aligned with what users truly perceive as immersive and authentic
