Microphotometric Errors for Photographically Recorded Spectral Lines of Nonuniform Half-Width

Abstract

Various microphotometric (or densitometric) techniques for generating quantitative intensity information from photographically recorded spectral lines of time or spatially resolved sources are discussed. The impact of various parameters on the accuracy of quantitative densitometry is presented. These parameters include line widths of the calibration spectrum, source line broadening, microphotometer scan slit width and the optical density of the photographic image. Nonrandom errors associated with the use of various microphotometer slit widths for spectral lines of nonuniform half-width are presented. Spectral lines which are uniform and exhibit slit width or diffraction limited resolution can be scanned with any size microphotometer slit width as long as the calibration curve is prepared using the same scan slit width. The use of microphotometer slit widths narrower than the line width produce H and D curves with maximal γ and increased accuracy in the final intensity value. A density-to-intensity conversion accuracy with a 6% average error was determined for SA-1 plates. For sources whose line widths are larger than the spectrometer bandpass, minimal errors are generated by using a narrow line source for calibration and scanning this spectrum with a slit width less than the line width. Scanning of the broadened line of interest is accomplished using a scan slit width equal to approximately twice the spectrometer slit width. Under these conditions an average error of approximately 11% was determined experimentally for SA-1 plates.