On the concept of absorption for a Fabry–Perot interferometer