=; Lighe transmission, T, of a deweloped file generally 2ecreases with increased exposure. Density, D, of a filw is defined as D = 10815 (l/r). Figure 11.3 shows a typical relatisashia between D and Cae exposure, EL Evi- dent are the toe, a more-or-less straight-line portisn, and the shouller of the curve where che density reaches a liaiting value. The relationshia be- tween D ami E is somewhat sensitive co the wavelength of light used. The teciprocicy law (which states thar density depends only on cotal exposure, E = It, ami not intensity, I, or tine, ©) bolds fairly well for exposure tines of interest, 7 to 200 psec. Fila calibration procedures accounted for the wavelength and reciprocity variations. D-log E calibration curves were obd- tained by exposing the film in modified Fastax streak cameras using the various garrow-band pass filters and several slit widths. The sun was used as a source. Its intensity, after passing through the filter, was seasured with a black-body receiver (Eppley thermopile). All file was from the same emulsion oumber; new calibration films were made at the test site and received the same handling, storage, ani developing as the record files. As a relative processing coatrol, sensitometer step wedges were included at the beginning and end of each roll of film. These were exposed on an ECG ME VI sensitomter at 10 sec. Figure 11.4 shows typical D-log E curves for the four wavelength regions passed by the filters. 1L.4 RESTLTS Curves plotted from control wedge densities indicated that processisg was reasonably constant for any one roll of fils (Fig. 11.5}, but vas not as consistent as would be desired from fila to filo. These differences usde direct comparison of test records and calibration records inadvisable. In addition, the step wedge densities do not extend to as high densities as the actual records. To make the data meaningful, a sethod for obtaining a density-versus~ relative-exposure curve was developed which used only the control step wedge densities, che known exposure difference between the four separate channels on each streax cecord, and the density versus time of these records. scale was obtained from calibration records, tail in Palver's report 12 The absolute This method is described in de- In brief, for the calibratisn fils, che average gtadiext over an exposure interval of 1.5 log E units is tazen so that the line taragh the end points is tangent to the D-log E curve at the toe eod (Fig. 11.2). The intersection of the extension of tais line with rhe exposure