Harward University published PhD Thesis "Vertical Silicon Nanowires for Image Sensor Applications" by Hyunsung Park. The thesis proposes to use Si nanowires of different diameter to create color filters. Oval wires can be used as polarization filters.
So this Theory uses vertical Silicon 'posts' of varying diameter and the 'Foveon Theory' uses horizontal 'sheets' of varying thickness to determine the frequency and intensity of the light.
ReplyDeletehttp://www.foveon.com/files/X3_tech_hero.jpg
Rob calls the Patent, someone invent this:
1. Make posts of various diameters and lengths to determine the frequency and intensity.
2. An ability to wobble these posts at the TOP (like LCD Shuttering, or Sony's rumored APCS (active pixel color sampling) Sensor could be used to: a.) increase dynamic range (increase number of bits per color), b.) increase resolution and c.) provide limited 3D Depth information.
EG: http://www.sonyalpharumors.com/wp-content/uploads/2014/11/structure1-700x529.png
3. An ability to wobble these posts at the BOTTOM to distribute to different "Wells" with differing depths and possibly different characteristics (EG: some Wells are more visible light sensitive whiles other Wells are better in the IR range of frequencies).
That would provide a single Sensor that could exceed any reasonable (and some unreasonable) Specs. It would provide a bit of 3D Depth Info from a single point (rather than 2 Senors at interoccular distance).
With a backend Processor to digest that mass of Data and turn it into an "Animatable 3D Video" (sort of like a Maya or Blender Script) that could have the ability to be re-shot (much like Lytro Photo-Videos can be refocused) we would be able to fix more in Post.
What would be the advantage of this method over the CFAs?
ReplyDelete