An overexposed spectrum may be obtained either by design to obtain an adequate signal for weak features or underexposed portions of the spectrum, or by accident, due to the presence of a previously unknown hot companion, or errors in estimated exposure times. Severe overexposures can affect the photometric accuracy of subsequent spectra and potentially damage the cameras. The residual image from an overexposure of a few times the optimal exposure time is effectively erased from the camera target by the standard camera preparation sequence (SPREP). An overexposure of 4 times or greater may require a special overexposure preparation sequence (XSPREP). If the identical portions of several images are expected to be overexposed to a similar degree, the XSPREP may be delayed until the series of exposures is completed. A reliable estimate of the degree of overexposure is helpful in determining the appropriate preparation of the cameras for subsequent exposures.
Overexposures also temporarily increase the phosphorescence in the affected area of the camera, producing a ghost image of the previous spectrum on any subsequent long exposure. This may result in significant photometric errors; false detections, if both spectra were taken at the same dispersion and in the same aperture; or a contaminated spectrum, if the overexposure occurred in the other dispersion mode. The XSPREP usually cannot eliminate this phosphorescence. Moreover, a residual image from a large overexposure may contaminate long exposures up to several days later.
The history of camera preparations preceeding your shift affects the general background level during long exposures. Overexposed tungsten flood lamp exposures (i.e. 200 percent) are performed as part of the normal camera preparation sequence, resulting in a weak phosphorescence across the camera faceplate. Typically, this background accumulates at a rate of 5 to 10 DN/hour, but a recent XSPREP (which contains an 800 percent tungsten flood lamp overexposure) may raise this value. It is possible for the phosphorescence background to contribute 80 DN or more to the signal of an eight-hour exposure. The only method for diminishing camera phosphorescence is to avoid using the camera for a number of hours. Snijders (1983) discusses how to estimate the phosphorescence resulting from a previous overexposure.