I have the following optical setup, in which the goal is to record the diffraction pattern of a sample on an image detector some distance away. In this particular case, the sample is sitting on a standard microscope glass slide. Now, after recording the diffraction pattern, I numerically back-propagate it (using the angular spectrum method) to the sample plane. The distance between the image detector and sample is ~10 mm. However, in order to reproduce the original image of the sample, I need to numerically back-propagate 40 mm. I'm not sure why this might be happening, so I was wondering: Could it be possible that the glass slide is effectively magnifying the diffraction pattern of the sample? If not, why would there be such a large discrepancy between the actual physical distance, and the numerically propagated distance?