This Viral Pool Clip Was Made Using Blender, Omniscient App & After Effects

The Viral Pool Clip, which circulated as an apparent real billiards mishap, is a VFX piece by 3D artist BrutalDuck. It combines live-action footage with individually animated pool balls, lighting work

The Viral Pool Clip, which spread online like a bizarre billiards mishap video, is a composited video created by 3D artist BrutalDuck rather than footage of real balls moving. After first filming a person making the motions of hitting invisible balls, the artist added billiard balls in Blender and constructed trajectories that look like a chance accident in real life. The reason the short video was mistaken for something between live-action and AI-generated content lies not only in the physical effects of the balls, but also in the convincing quality of the image, which fits naturally with the filming environment.

Moving the Filming Space into 3D as a Reference

The production process used Omniscient, a camera-tracking and LiDAR-scanning tool for iPhone. The creator scanned the pool table with the app and used that information as a reference for the subsequent animation. Omniscient is designed to import camera movements and scene-scan data from footage into Blender and After Effects, making it well suited to matching the perspective of a real space with virtual objects.

Creating the Movement and Light One Ball at a Time

The billiard ball models came from free 3D models on Sketchfab, but the movement was not left to an automatic simulation; each ball was animated with keyframes. The lighting and nodes were then adjusted, and the light was further refined in After Effects. In other words, the core of this video lies not in a single generative tool, but in a production workflow that connects live-action filming, spatial scanning, manual animation, and lighting adjustments.

A Workflow That Creates Plausibility

This example shows that securing the conditions of the real camera and space comes before making only the objects on screen look realistic. Recording perspective and scene information during filming makes the later process of matching the size, position, and lighting of the 3D balls much more concrete. Conversely, no matter how precise the balls' trajectories are, the composited scene is easily exposed if the consistency of the camera and lighting breaks down.

Source: 80 Level production analysis · Official Omniscient website