In 1949, Miller discussed the rotating mirror framing camera, having devised an individual frame system that could record a sequence of two-dimensional images. In principle, the rotating mirror framing camera requires that an image of the object under study be formed at or on the reflective face of the rotating mirror. The cones of light forming this image may then in turn be successively reflected into box cameras adjacently located on a circular locus having the axis of the rotating mirror as an approximate center. Over the years numerous improvements have been made to rotating mirror cameras, and some of these are listed in the noted references.
When using the same equations to describe the image locus as used for a streak camera design, we note that the basic mathematical difference in a framing camera is simply that the optical lever arm Q is set equal to zero. A sequence of framing lenses and a stationary focal surface (recording arc) ideally concentric with the arc of framing lenses are, in effect, sequential box cameras individually aimed and focused at the rotating mirror as shown in Fig. 17. Ideally, we would like the two-dimensional image that is proximate to the rotating mirror to appear motionless to each box camera. Unfortunately, while the mirror rotates to complete an exposure, the image moves as we look at it from a framing lens perspective. The magnitude of this motion is the key to much of this camera's recording integrity. Proper pupil control is the other cardinal design issue but not one that demands exactitude. Nevertheless, successful design and patent claim require the shape of the exit pupil of the objective lens system be imaged on and matched in size to the entrance pupil of the relay lenses. This requirement assures minimum and truly successive (noncrosstalk) exposures as the mirror revolves; however, this requirement does not impact static image resolution which could be misleading per se. Estimating the correct size for the exit pupil as judged by eye is an adequate procedure, because the residual motions of the primary image at the mirror are not unduly sensitive to small errors in pupil size matching. To initiate a camera design, we may for a first approximation locate the rotating mirror such that its reflective face is proximate to the origin of the coordinate axis as shown in Fig. 17. Intuitively, we would place the primary image, i.e., the image that should be somewhere in the vicinity of the mirror's reflective surface, at the origin of the coordinate system by setting Q = 0 in Eqs. (7) and (8). With this beginning, we then seek a position for the axis of the rotating mirror that will introduce the smallest collective image motion (aberration) as seen and summed from each framing relay lens perspective.