salukikev
Mechanical
- May 14, 2008
- 110
Hi! You might remember me from the last two threads in this group..(not all that active I see). Anyway, I'm still dabbling with that same project where I'm trying to capture an image from inside a transparent sphere. I've got the hyperhemisphere, and I've got a few ideas for how to manage the wiring but still brainstorming ways to reap more benefits without all the shortcomings. To that end, this weeks idea was to have a stationary camera point down into the sphere. At the origin of that sphere there would be a reflector mounted on an axis. That axis would be mounted to a single sided swingarm that could also rotate. (this detail omitted in attached drawing for clarity) In this way, I could get a full view around the equator easily and a pretty decent angle view up and down. (I'm even considering a flip action to selectively view straight down).
So of course this is pretty cool in that I can skip all the hassle of mounting the whole camera assembly and I don't have to worry about wiring anything except the optics.
What I DO have to worry about is distortion and range of motion now that my hemisphere is acting as an optic too. (which it always would but now it varies per angle).
I still use Optical Ray tracer for stuff like this, but thought there might be a more elegant solution or some unforseen hangups with something like this. I surely like the benefits but as always these complications are cause for concern so here I am. Fundamentally though, I feel like a system like this is a really good idea, but I can't seem to find any existing examples or even useful keywords to describe it. I'm sure there's some out there and I'd love to see something similar that's been fleshed out a bit more before starting toward an experiment.
**I guess what I'm really trying to accomplish can be described as a rotary prism, but I want to try to place the axes in a way that delivers as uniform a viewpoint and consequent distortion as possible.
So of course this is pretty cool in that I can skip all the hassle of mounting the whole camera assembly and I don't have to worry about wiring anything except the optics.
What I DO have to worry about is distortion and range of motion now that my hemisphere is acting as an optic too. (which it always would but now it varies per angle).
I still use Optical Ray tracer for stuff like this, but thought there might be a more elegant solution or some unforseen hangups with something like this. I surely like the benefits but as always these complications are cause for concern so here I am. Fundamentally though, I feel like a system like this is a really good idea, but I can't seem to find any existing examples or even useful keywords to describe it. I'm sure there's some out there and I'd love to see something similar that's been fleshed out a bit more before starting toward an experiment.
**I guess what I'm really trying to accomplish can be described as a rotary prism, but I want to try to place the axes in a way that delivers as uniform a viewpoint and consequent distortion as possible.
