On linear fit to ZAMS and reddening determination