One-Toe vs. Two-Toe Camera Rigs

< Previous | Contents | Manuals Home | Boris FX | Next >

‌One-Toe vs. Two-Toe Camera Rigs

Ideally, a camera rig has two cameras next to each other, perfectly aligned. If both camera viewing axes are perfectly parallel, they are said to be converged at infinity, and this is a particularly simple case for manipulation. Usually, one or both cameras toe in slightly to converge at some point closer to the camera, just as our eyes converge to follow an approaching object. Mechanically, this may be accomplished directly, or by moving a mirror. We refer to the total inwards angle of the cameras as the vergence angle.

It might seem that there is no difference between one camera toeing in or two, but there is. Consider the line between the cameras. With both cameras properly aligned and converged at infinity, the viewing direction is precisely perpendicular to the line between the cameras. If one camera toes in, the other remains at right angles to the line between them. If both cameras toe in, they both toe in an equal amount, with respect to the line between them.

If you consider an object approaching the rig along the centerline from infinity, the two-toe rig remains stationary with both cameras toeing in. The one-toe rig moves backwards and rotates slightly, in order to keep one camera at right angles to the perpendicular line between the camera centers.

SynthEyes works with either kind of rig. Though the one-toe rigs seem a little unnatural (effectively they make the audience turn their heads), the motions are very small and not really an issue for people, except for those who are trying to do their tracking to sub-pixel accuracy! The one-toe rigs are mechanically simpler and seem more likely to actually produce the motion they are supposed to (are the two-toe rigs really moving at exactly matching angles? Are the axes parallel? Maybe, maybe not!).

< Previous | Contents | Manuals Home | Boris FX | Next >