Spatial Memory Fragment is an evolving system for turning LiDAR scans into interactive web artifacts.

The project started from a practical question: how to show point cloud scans without reducing them to still images, videos with fixed camera renders. I have collected many scans from different places, streets, friends, events, exhibitions, rooms, and temporary situations. In this project, each scan remains navigable. The viewer can choose a distance, angle, and point of view, or leave the scene running in automatic camera mode.

Each artifact is built as a single HTML file. It contains the viewer, interface, configuration, intro animation, camera behavior, animation stages, Strudel panel, and audio-reactive modes. The .ply scan is currently loaded from an external link, while most visual and technical settings are controlled inside the file.

The interface begins with a loading screen containing the title, place, date, credits, and an Open button. After opening, the scan assembles into space. The viewer can rotate, zoom, move through the cloud, switch camera presets, open the Strudel panel, turn on microphone-based audio reactivity, or check controls and technical information in the help panel.

The work is part of my research into point cloud photography. I am interested in scans as spatial images: technically precise, but visually open. They can document a real place and at the same time become an abstract field of points, gaps, color, noise, and movement.

Credits

Spatial Memory Fragment

Created by Grisha Tsvetkov

2026