Directional Audio Processing and Visualisation on the GPU

De Grandis, Lorenzo (2025) Directional Audio Processing and Visualisation on the GPU. Bachelor's Thesis, Computing Science.

Preview	Text thesis-s5221609-.pdf Download (8MB) | Preview
	Text Toestemming.pdf Restricted to Registered users only Download (180kB)

Abstract

Accurately perceiving the direction of sound is a fundamental part of human spatial awareness, enabled by auditory cues such as phase shifts and spectral dampening described by Head-related Transfer Functions (HRTFs). However, these cues are often lost in digital playback systems, reducing immersion and accessibility in virtual environments. This thesis explores a system for real-time directional audio processing and visualisation on the GPU, aiming both to enhance immersive experiences and to provide alternative visual cues that could potentially assist hard-of-hearing users in understanding spatial soundscapes. The proposed audio engine combines ray tracing, FFT-based convolution, and clustering techniques to reproduce directional sound, while a spherical visualiser shows the perceived locations and properties of audio surrounding the listener. The system, implemented in Vulkan, employs asynchronous compute pipelines to achieve low-latency performance. A small user study demonstrates that the approach improves spatial awareness through both auditory and visual modalities, though limitations in the simplified propagation model remain. Overall, the results highlight the GPU’s potential as a platform for real-time spatial audio and lays a foundation for future enhancements to be built upon.

Item Type:	Thesis (Bachelor's Thesis)
Supervisor name:	Frey, S.D. and Kosinka, J.
Degree programme:	Computing Science
Thesis type:	Bachelor's Thesis
Language:	English
Date Deposited:	19 Sep 2025 06:13
Last Modified:	19 Sep 2025 06:13
URI:	https://fse.studenttheses.ub.rug.nl/id/eprint/37042

Actions (login required)

View Item