Hi everyone, I'm trying to study reflection on the rear side of an optical device. I wrote a transfer-matrix code in matlab to compute the reflectance. I checked my code by comparing some basic cases with other kind of optical simulations (FTDT with Lumerical). When the incident medium is Air or Dielectric materials, I got relevant results (both transfer matrix or FDTD give the same). When I only have one interface (like Silicon-aluminum), My code gives the same solution than the one obtains with Fresnel coefficients. The problem appears when I introduce thin films between two absorbing medium. For example, Silicon-MgF2-Aluminum interfaces, where silicon is the incident medium and Aluminum the substrate. For some wavelength, the reflectance (so the amount of light reflected in the silicon) is higher than 1, which is of course wrong (no luminescence in transfer matrix code). I don't know what's wrong with transfer matrix. This problem appears only when the incident and substrate medium are absorbing and a dielectric layer are inserted between them.