Simulated red blood cells and microscopy

I enjoyed the writeup on the optics. Very similar to the contrast transfer function in my field (cryo-EM). Except arguably harder because this is an anisotropic technique with a much less coherent source. In this case, refractive index is analogous to coulombic potential in my field. i.e. stuff being in the way (refractive index, electron density) of the scattered wave = weak phase change = things become out of phase and cancel (or reinforce) as a function of spatial frequency.

Of course, you can phase flip in Fourier space (which we do), but that still leaves you gaps at (and approaching) where contrast crossed over zero. To solve that in my field we average across different defocuses (shifts the phase of the oscillations) to fill in the gaps. Nothing analogous here that I can tell but perhaps you could rotate the prism and take other exposures of the same sample to fill in the missing information. Or just live with it.