These look ok to me. The brightness profile (PSF) of a typical star should be a Gaussian distribution(bell-shaped curve) not something ending in a sharp boundary.Geoff