Re: 3d Rocket Cam

[P3D John W Roberts <roberts@xxxxxxxxxxxxxxxxx>]

>Date: Wed, 12 Mar 1997 09:35:13 -0600
>From: P3D Paul Albers <PAlbers@xxxxxxxxxxx>
>Subject: 3d Rocket Cam (Was:Newbie with questions)

>The story so far:

>A 4 ft separation can be done inside a single rocket. With altitude ranging
>from 100 to 400 ft, that gives ratios of 1/25 to 1/100 so that should
>work out to at least some good 3D shots along the way, maybe all of them.

>Current questions:

>1.  Can a decent beam splitter be home built?

The key question: how big can the outbound mirrors be? Do they have to
fit in the rocket tube? If you're thinking of something like the "Mean
Machine", I don't think you can fit anything over 2 inches. With 4-foot
separation, I believe that means you would have to use extremely long
focal length (high magnification) for it to work with such small mirrors.
Long focal length requires extremely precise alignment (which may not be
possible in a cardboard tube), and 3D pictures taken with long focal length
lenses are usually so highly distorted (when viewed in 3D) that the stereo
effect can hardly be seen. Also, with small mirrors and long focal length,
your aperture ratio would be extremely small - you'd need extremely fast film
and long exposure to properly expose the film. But long focal lengths are
very susceptible to motion-induced blur, which would call for a very short
exposure time, and fast film is grainy, which is especially noticeable
when viewed in 3D. If you can provide exact dimensions, etc., there are
people on this list who could calculate the constraints.

I think there are ways you can get acceptable 3D photos from a model rocket,
but I don't think a beamsplitter running the length of a rocket is practical.
I still favor the two cameras falling separately. If you just *have* to get
the two views from one rocket, two cameras would be much better than a
beamsplitter.

>2.  If I use twin cameras in the rocket, and rotate them to point
>    outwards slightly, (reverse keystoning?) would that simulate
>    a larger ratio?  

Don't think so.

>Would the large altitude cancel out or reduce any keystone effect?
 
If you have an extremely long focal length that would help, but it would
introduce all the problems described above.

John R


------------------------------