Matchmove

Transferring the camera movement from live-action footage into digital space—that's the core of matchmove. You film a scene with a real camera, and the matchmove artist reconstructs this movement as 3D camera data to place CG elements into the same space with correct perspective. Without matchmove, any insert of CGI objects looks glued on. With matchmove, composition becomes true spatial integration.

In practice, it works like this: On set, you need tracking markers—small, high-contrast dots distributed across the entire image area that the camera sees during its movement. These are later analyzed by tracking software (typically: Boujou, PFTrack, or Nuke's Tracker). The software recognizes how these markers move in the 2D image and derives the 3D camera position and rotation from it. The result: a camera path as a sequence of numbers that your 3D department can directly import into Maya, Houdini, or Cinema 4D.

The quality of the matchmove heavily depends on your shot. The more markers visible, the higher the contrast, the less motion blur—the cleaner the tracking. If your camera is fast and makes large movements, you lose markers from the frame; then it becomes difficult. Lighting changes and reflections can also confuse the tracker. Good matchmove therefore requires planning: clarify with your VFX supervisor where the markers should be placed, how bright they need to be, and which camera movements are trackable.

A practical tip from the set: Attach markers with reflective gaffer tape to boards or objects at eye level and in the middle of the frame. Corner markers help the tracker anchor the image plane. And don't forget—the markers must be keyed out later, so keep a few frames clear without talent, just with markers, as a clean reference for the VFX crew. This saves hours in post-production. Matchmove is not rocket science, but it demands care from the very beginning—from your side as DoP to the final camera calibration.