top of page
skysky.png

A THOUSAND CLOCKS

WebXR spatial-audio work / immersive computational composition

Director & Composer & WebXR Developer

Meta Quest 3 · WebXR · head-tracked spatial audio · networked loudspeaker integration

skysky.png

A Thousand Clocks is an immersive WebXR sound work about impossible relationships to time.
The participant enters a clock tower filled with notes left by others. Memories, regrets, and wishes for moments they would stop, revisit, or change. After adding a wish of their own, the synchronized world begins to break apart. The tower collapses into a field in which incompatible moments appear to coexist across parallel timelines, before the participant eventually returns to the tower and leaves a trace for whoever enters next.
Throughout the work, spatial audio is not only attached to objects in the environment. Listener position, head orientation, and movement continually reshape the participant’s relationship to clocks, voices, music, and the temporal behavior of the world.

Research Question

Overview

What happens when the listener’s spatial perspective
becomes a compositional variable?

A Thousand clock part 1Scarlett Shi
00:00 / 03:00
A Thousand clock part2Scarlett Shi
00:00 / 03:14

A Thousand Clocks exists in two spatial listening configurations.
 
Standalone WebXR renders the work directly for the participant through the Quest, preserving a listener-relative binaural perspective.
Networked Performance Mode sends tracked listener and controller data outward through OSC, allowing the same virtual sound objects to be rendered through an external spatial loudspeaker system.

WebXR & Spatial Audio System

The experience is built for Meta Quest 3 in Three.js and WebXR, with browser-based real-time audio rendering and head-tracked interaction. The same work can also operate as a networked spatial-audio system, transmitting listener and controller data through WebSocket and OSC for external loudspeaker rendering.

WebXR

  • Meta Quest 3

  • Three.js / WebXR

  • Web Audio API

  • head and controller tracking

  • real-time listener-relative spatialization

Networked Spatial Audio

  • WebSocket communication

  • Node.js bridge

  • OSC

  • listener position + yaw tracking

  • Inviso loudspeaker rendering

From Headphones to Loudspeakers

Presentation

2026 — 3rd Spaceout International Ambisonics Festival
University of Michigan, 36.1 speakers
Selected, Live Performance Track


2026 — UC San Diego Spring Festival
UC San Diego Experimental Theater, 62 speakers
Premiere / WebXR spatial-audio experience

© 2026 by Scarlett Shi.

bottom of page