How can metasurfaces be used reliably for imaging?
I’ve seen a few papers where metasurfaces are used for some imaging task, and it seems like a few nuances are being glossed over.
The main problem I have is that most metasurfaces are incredibly sensitive to the angle of incident light. Take the attachment for example (taken from the supplementary info in https://www.science.org/doi/10.1126/sciadv.adp5192, describing a metasurface used for hyperspectral imaging)—the light from the pulsed laser should be more or less normal to the optical axes, but then it transmits through the sample and passes through lenses, certainly no longer normal to the optical axis. This wasn’t mentioned at all, despite the fact it should make a significant difference in how the metasurface transmits the light.
This isn’t really specific to this application though—in general it doesn’t seem like controlling the angle of incident light is easy to do. To my knowledge certain samples (like something rough) can transmit/reflect light at several angles over a tiny area which makes controlling the incident angle of light on the metasurface seem impossible.
I am wondering if I am mislead, or if there’s a reason why it’s okay to overlook this. Thank you for any help!