3D Camera Tracking in Fusion Part 2: Improving 3D Camera Solves

From Point Cloud to 3D Compositing – Position and Mask Images in 3D Space

In Part 1, we established a solid camera solve with optimal point distribution and low solve error. Now comes the real work: transforming that seemingly chaotic point cloud into a practical compositing tool.

Point clouds from a 3D camera track can look confusing at first glance – hundreds or thousands of white dots scattered across 3D space with no clear organization. Yet these points hold the geometric key to accurately positioning elements in your scene. The challenge isn’t generating the point cloud; it’s organizing it in a way that supports precise compositing work.

In this Insight, we tackle that challenge head-on. You’ll learn to organize tracking points using user-defined colors and naming conventions, add targeted points to specific areas using magic masks, and build simple 3D reference geometry that serves triple duty: verifying your track, positioning elements, and creating believable masks. We demonstrate these techniques by compositing a fire element into a scene – placing it in a location where no direct reference points exist.

“Sometimes it’s even easier if you think ahead during the camera tracking and work already in 2D to define which points belong to where and…organize them a little bit.”

Bernd Klimm, VFX Artist

Key Takeaways

By the end of this Insight, you should understand how to:

Related Mixing Light Insights