This shader is used to shoot final gathering (FG) rays into the scene in a predefined way. Without this shader, mental ray normally shoots eye rays through the camera into the scene during the final gathering precomputation stage. FG points are calculated in the scene based on the location caused by how these eye rays are shot.
The advantage of this is that the density of FG points is related to image space, and is therefore automatically adapted to exist in areas where the points are necessary. On top of this is an adaptive calculation of the FG point density, creating the optimal solution for the image.
However, when the camera moves during animation, the location of where FG points are calculated actually moves along with the camera. In an animation, this can potentially cause artifacts in certain situations, especially if the camera moves slowly or only moves a little (such as in a small pan, dolly, tilt, truck, or crane move).
The mip_fgshooter shader safeguards against these situations: it allows you to use a fixed (or a set of fixed) transformation matrices as the root location from which to shoot the eye rays that are used (during the final gather precomputation phase only).
