User Manual

The Camera Tracking Workflow
Camera tracking has two main phases:
1 Tracking, which is the analysis of a scene.
2 Solving, whi7ch calculates the virtual 3D scene.
Once you complete these steps, an animated camera and point cloud are exported from the
Inspector into a 3D composite. The Camera Tracker encompasses this complete workflow
within one tool. Five tabs at the top of the Inspector are roughly laid out in the order in which
you’ll use them. These five tabs are:
 Track: Used to track a clip.
 Camera: Configures the basic Camera parameters.
 Solve: Calculates the 3D placement of the 2D tracking points and
reconstructs the camera.
 Export: Generates a Camera 3D node, a Point Cloud node, and a
3D scene in the node tree.
 Options: Used to customize the look of the onscreen overlays.
The Camera Tracker tab.
Clips That Don’t Work Well for Camera Tracking
Even though the Camera Tracker is somewhat automatic, it sometimes needs your help. If you
can identify potential issues before you even track or solve the shot, you can save yourself
much time. Certain types of clips will cause more significant problems for camera tracking than
others. Some are fixable, while for others you just have to admit defeat and figure out another
solution. Here is a list of the types of shots to look out for, as they can be big headaches for
camera tracking:
 Lack of depth: Camera tracking requires parallax in a clip in order to work. You must be
able to identify objects further away and objects that are nearer as the camera moves.
If everything is at the same distance from the camera, there is no way to calculate
depth. In this case, it’s better to skip the Camera Tracker node and find another
solution.
 Locked-off shots: If the camera does not move, there is no way to calculate which
objects are closer and which are nearer. Again, don’t spend too much time in this
situation; it is better to skip the Camera Tracker node and find another solution.
1590Chapter – 77 3D Camera Tracking