I applied the method of Swanepoel for a non-homogeneous film in thickness. The problem is that for some thin films, the experimental spectrum is reproducible, but for others it is not possible, despite the presence of fringe. Can anyone offer an explanation or clarification?