The DEPTH Workflow - 3D Postproduction: Stereoscopic Workflows and Techniques - page 136

of the camera. To make parallel imagery viewable as 3D frames, teams must i rst
move L and/or R to the left and/or right. Once imagery has been adjusted without
exceeding either near or far maximum parallax targets, the illusion of Z-Space is
created. The key advantage to shooting parallel is that it can reduce or eliminate
some common geometry issues such as keystoning, that is, a change in shape from
rectangle to trapezoid, and cardboarding, or the degree of l atness in an object.
Parallel imaging systems are also considerably easier to setup on set or location
because there is no need to check convergence angle or adjust it during the shot.
Converged
In converged imaging systems , the axial lines of L and R imagers rotate toward
each other, meeting somewhere in front of the rig (Figure 5.3) . Converged
imaging replicates some of the aspects of human vision, with a greater angle of
convergence when viewing near objects, a smaller angle when viewing far objects,
all the way out to parallel, which we discussed above. This angle creates an image
in which horizontal parallax varies for different areas (or objects) of interest in the
scene. The result is that objects in the scene live in different places in Z-Space.
Shooting “converged” allows the on-set creative team to adjust creatively the
location in Z-Space of the interest area during the shot. In this scenario, the
L CENTER LINE
R CENTER LINE
exaggerated for clarity
Figure 5.3 Shooting Converged.