4 ms·
It's a very interesting technology, especially since it holds the promise to directly realize an almost arbitrary phase-mask (analogous to a lens' shape) that h
by wnkrshm 7y ago
It's a very interesting technology, especially since it holds the promise to directly realize an almost arbitrary phase-mask (analogous to a lens' shape) that has been found using already existing optical design technologies. The problem, as in the article is size though.
Apart from the optical (especially on-chip) applications, in theory, it's possible to have a freeform surface and cover it with small meta-atoms (subwavelength structures) so that reflected wave-fronts are planar again, like the incoming one. With millimeter waves such structures are way easier to make than on the tens-of-nanometer scale - these will be used in all kinds of aerospace products. And iirc, the first application (according to Frederico Capasso) was laser optics on a satellite that had to be compacted and have their weight reduced.
- jaclaz 7y ago>The problem, as in the article is size though. I see it more in the amount of light needed, 50% pass-through vs. roughly 90% of conventional lenses seems to me "a lot". Even if the researchers "hope" that they will be able to improve the material/shapes/whatever, it doesn't seem like an easy chore.
- Veedrac 7y agoI would not be surprised if there are many applications where removing 90% of the weight and bulk of the lens, or adding this much flexibility in the lens' functionality, would more than make up for having to double the sensor size or light intensity.