exposure is abandoned; each exposure is treated as if it constitutes a separate 
plate (or frame), thereby introducing a fresh set of unknown elements of exterior 
orientation. However, elements of interior orientation x, ,y,,c and coefficients 
of radial and decentering distortion are regarded as being common to all such 
frames. Although each frame contributes a new set of unknowns to the reduction, 
no limits need be placed on the number of frames that can be reduced simulta- 
neously by SMAC, for the overall number of computations increases only linearly 
with the number of frames carried. This is because the general system of normal 
equations assumes a patterned form that can be formed and solved with great 
efficiency by means of the same set of algorithms that were originated in Brown 
(1958) for application to analytical aerotriangulation . 
Stellar SMAC has proven to be so consistently successful that it has totally 
displaced the conventional process of stellar calibration at DBA Systems as well 
as at those governmental organizations having access to SMAC. Because data 
gathering procedures are so vastly simplified (with stability of the camera and 
precise timing of exposures no longer being required), the cost of stellar calibration 
has been drastically reduced. This has permitted DBA to perform SMAC calibrations 
of a sizeable number of commercial mapping cameras over the past five years. It 
has also made the infinity calibration of DBA's various close-range cameras a 
relatively simple and inexpensive undertaking. 
Aerial SMAC, which is also developed in Brown (1968), is the version of 
SMAC applicable to a set of frames exposed over a targeted test range. In a 
typical application the calibration may be derived from the simultaneous reduction 
of a set of 20 to 40 frames each recording from 25 to 50 targets. This not only 
leads to an extremely accurate calibration of radial and decentering distortion but 
also generates such a large composite set of residuals that it becomes possible to 
derive empirical corrections for those residual systematic errors that persist through- 
out the set of frames. 
When the target range is suitably distributed in three dimensions, SMAC 
can also produce an accurate calibration of x, , y; and c from vertical photographs. 
Otherwise, these parameters are so strongly coupled with coordinates of exposure 
stations X^, Y^, Z€ that they must be tightly constrained in the adjustment in order 
to avoid indeterminacy. Alternatively, in those rather special situations where 
Xt, Yc ,Z¢ can be tightly constrained, the recovery of x ,y;,¢c becomes possible . 
A major advantage of Aerial SMAC is that it does not require a photo- 
grammetric range containing a large number of targets in order to be effective. 
Excellent results can be obtained from a range containing as few as 10 to 15 
targets (in which case, the reduction of 50 or so frames would be advisable). A 
drawback of Aerial SMAC in comparison with a more comprehensive process to be 
described later is that it has no provisions for consideration of errors in coordinates 
of targets and thus requires that such coordinates be established to a very high order 
of accuracy. Extensive practical results with both Aerial and Stellar SMAC are 
reported in Brown (1968).