Repurposing Audio Playback Tools to Test Human Localization with 6DoF Sound
summary
A practical method records directional human speech with a first-order spherical microphone array, builds a triangulated octahedron mesh for spherical-weighted interpolation, and repurposes Unity playback tools to mix simultaneous chunks by volume to deliver object-based 6DoF audio in AR. Deployed on a HoloLens 2 in a mobile localization task, the technique yields effectively 100% accuracy across conditions while monaural HRTF playback shows much lower and more variable localization (interquartile near 70%), demonstrating a low-overhead, working pipeline for translational plus rotational audio cues.
AI-written summary · about
chapters (10)
- 0:00Introduction to study
- 1:51Ambisonics limitations
- 3:37Recording method overview
- 5:00First study results
- 5:41Second study setup
- 7:03Playback technique explained
- 8:26Experiment demonstration
- 9:50Study results analysis
- 10:51Limitations and future work
- 11:53Q&A session
related talks
21mComposing Spatial Music with Web Audio and WebVR
20mSpatial Communications at Scale in Virtual Environments
46mHoloLens 301: Advanced Mixed Reality Development and Best Practices
10mInside-Out Instrument Tracking for Surgical Navigation in Augmented Reality
03mSoundTrek: A Virtual Reality Audio Perspective-Switching Game
01h 00mReal World HoloLens Mixed Reality Development with Unity - Be Part of the Future
40mHolographic Telepresence with Hololens
59mHoloLens Development: The Next Steps