I had attended Bittorrent reunion recently at Ashwin Navin’s restaurant at The Summit in SF. Ran into some old friends, and had an interesting discussion with David Harrison and Jens, and Arvid. So here’s the idea that I pitched and we discussed:
Say I have a room, well, say I’m on the inside of a sphere, and say there’s a glass floor that I am standing on and that all of the interior of the sphere are covered with little projection devices that can very quickly beam a light at any direction inside the sphere. Individually, these are miniature projectors rotating on two axes.
Now, each of these pivoting projectors pivots very fast, and is able to scan through the entire space in front of it and alter color as it does so, if one puts a piece of paper in front of one of this device, one would find that its able to project any image like a typical projector.
The spherical room’s wall are completely covered with these micro projection systems, each the size of a TV screen pixel, or smaller. Each of these wall-pixels are connected to a central computer system.
The goal is to render a 3-D image for a person standing on the glass floor anywhere in the room, simultaneously.
This means, the computer system computing the output for each of these woxel (wall-pixel instead of voxel or pixel), does not have to take into account where the person is in the room or what he is doing. All view points/angles are simultaneously rendered correctly.
The room will be able to accommodate two or more persons and still produce faithful images on/behind or in front of real objects–again without respect to where each person is. Think simply, each woxel is producing rays that a ray-tracing program would produce using parallel projection.
The genius of this discussion is not necessarily the concept of a woxel, which I just named but my friend David Garmire told me about some years ago, but the fact that the computer behind this system is a massively interesting CG problem that has not been worked. The optimizations there-in has not been written yet. AFAIK.
A second related thought is one of specular reflection. So one of the things that current cellphone manufacturers does not allow us to do, actually also screen/TV, is that it does not allow us to set the apparent reflective property of the glass infront of each pixel. I can set RGB, but I can’t make that point more glossy by increasing the specular reflection.
It would be very cool for CG effects if we could set these properties of the material on top. It would look and reflect like paper when it’s displaying a book, and the iPhone 16 could turn into a mirror, and there’d be an app for that.
Anyways, some ideas and description of systems/implementation/algorithms for posterity to snicker about… poor old Huan, didn’t even think of the possibility that the iPhone 11 render mirrors completely useless (obviously not by having pixels that can turn into reflective mirror as I’ve discussed here)….
sigh…… talk about a set of lonely-friday-night-typo’s.