AN ALGORITHM TO CLEAN CLOSE STELLAR COMPANIONS

In our effort to study the post-main-sequence evolution of the globula r cluster stars we developed some procedures which were designed to pe rform stellar photometry in moderately crowded fields by using a perso nal computer. This paper presents a description of an algorithm for cl eaning close stellar companions which is included in our data processi ng. The main goal is to establish how the cleaning procedure biases th e accuracy of the derived photometry. As a test for our data reduction , BV photographic photometry for nearly 500 stars, down to 0.6 arcmin from the center of globular cluster M 5 = NGC 5904 = C 1516 + 022 has been performed. In addition, we present a comparison of our photometri c results with those obtained with the DAOPHOT package for the same cl uster region. We also discuss the ability of three kinds of magnitudes - based on Gaussian fits. the use of synthetic apertures and the outp ut of the DAOPHOT package - to reveal morphological features in the co lor-magnitude diagram.