DIYer ISO expert-ish help with Optics & Mirrors in a Custom Lens Assembly

Hi All!
I'm seeking some expert-ish input on a project I've started working on for my 20d. I have a budding interest in Stereo Photograph and my goal is to have a 'lens box' that would bayonet onto my camera like any other lens. (no electronics unless it's possible to easily just activate the AF indicators in the 20d reusing an old Sigma Zoom Mount.) On the front would be two spots at the appropriate seperation to screw-in C-Mount Lenses.
This is the mount of 16mm movie cameras, early TV cameras, and Currently used in Security and Industrial Sensor Cameras. These lenses vary in quality between disgusting and magnificence. I luckily picked up two modern and very expensive Pentax lenses in mint condition for almost nothing. They have approximately the field of view of a 40mm lens on a 35mm SLR. Their image circle covers the approximate dimensions of half an aps-c sensor.
I am kind of inspired be the 'idea' of the Loreo lens in a cap's simplicity, but am trying to make a quantum leap past it in image quality, focusing ability, and aperture control - Especially in Maximum Aperture. My goal is something to easily put on the camera then meter, focus, compose, and shoot - even handheld and subjects with motion.
My difficulty is getting a lenses' image properly routed back to the imaging sensor. My initial 'sketched out' design failed miserably in a 'rough' table top optical test. If anyone or many someones thinks they might be interested in helping me. I would be terribly grateful. My budget is very very very small!!!!
Sincerely,
Richard in Snowy West Michigan
Brrrrrrrrrrrrrrr

Just thinking about what you propose brings up what to me would be an insoluble problem. Maybe not to you?

Never mind the mechanics of mirrors and housing for the moment. The lenses you are proposing to use will have been designed to bring the image to a certain point of focus (the film plane). You are having to install additional elements between the lens and the sensor cum film plane, so you cannot bring the image to a focus as things stand.

Surely, any lens such as you are trying to construct would need to be designed fron scratch to include the far longer light path? Even if you could somehow rig up two sets of glasses (Hubble telescope?), the design and manufacture of these would be very expensive while the optical results would probably be dreadful, with all sorts of aberrations.

Hubble only had a very slight problem, measured in microns, where you are talking millimetres or even inches.

It would be far easier and cheaper to mount two identical cameras and lenses side by side on a plate arrangement, link the two with a double shutter release cable and you're in business.

Richard.

Hello,
I am fearing that you are going to turn out to be quite correct in regards to my initial concept needing extra optics and such to account for the increased lens to focal plane distance. I've discovered that 'facts' about SLR design and construction I'd been laboring under were anything BUT facts! I'm completely stymied at this point.
I tried a new experiment this evening. I set up a measured layout with a 35mm lens at the appropriate flange to focal plane distance from the mid point of a mirror on a 45 degree angle to the lens. Another mirror on a parallel angle then bounced that mirror's image back to my camera whose lens was focused on the midpoint of the second mirror. My results were less than stellar. I ended up with a fairly crisp image of the rear of the primary lens.
I was hoping that a 50mm lens on a lens reversal mount would gather the image the primary imaging lens had focused upon the mirrors and give me an optical fix to the blur of light I'd received initially. I was working from the thought that ordinarily a 35mm slr lenses' image is bounced off a mirror into the prism or mirror arrangement (porro - prism?) for composition and focusing. I neglected to account for that little focusing screen at the bottom of the prism when thinking about that optical pathway and replicating a similar optical route to bounce my lenses image back to the focal plane.
I'm keenly aware of my lack of optical physics education at this point.
I'll keep people informed
Richard

A couple of things to think about:

1. Fixing the lenses together means you cannot vary the distance between them... meaning you're stuck at a single stereo focal distance. If you want to make a stereo image of something far away, you need to move more than making a stereo of something up close. If you want to make a stereo of a thin, elongated object, you need a different angle than a flatter object. I think by linking these two lenses, you'd be stuck at a certain subject shape at a certain distance.

2. Assuming you'd be using two distinct lenses, you're going to have to "join" their streams of light somehow. Since they both produce circles, you're going to have a black concave lens shape in the middle. So, when looking through your viewfinder, you'll most likely see two half-circles turned outwards. So, your camera's metering will be all thrown off, because there will be a large section in the center that gets no light. Most likely, it would take this into consideration when it tries to meter, and I bet you'd get some very inconsistent results.

3. Your lens setup will only be as good as the weakest element. So, even if you managed to get all of the focus points to line up (which I think would be very hard in and of itself), you're going to have to have a set of very precise, and very sharp silvered mirrors. Since these are very fragile, you'll also need a protective housing to protect these expensive lenses.

4. Any tiny difference in lens positioning will throw your focus off. Lenses that are mounted even a few millimeters in front of or behind their optimum length lose infinity focus. Millimeters. So, you're going to have to either

a) File the lenses down and pres them up against the camera, taking into account and fabricating a mount that keeps them EXACTLY, to the MILLIMETER, their optimum distance from the camera. If not, you'll again lose the ability to focus on anything.

or

b) Make your own lens that corrects the difference, which will most likely cost a fortune and look like crap.

So... it doesn't seem all that possible. But, maybe you'll surprise me! Good luck.

Me Again! Original Poster.
I've come up with a few options I'm testing out. My mobility is a 'bit' limited (read that to mean I can only work until the pain gets to be to much), so I even though my days are long, my experiments are abbreviated.
To address Air-brontosaur's thoughts, you've kicked out some important variables to be addressed. The first is that In order to cover the full sensor with the image circles of each lens would mean the return mirror would be covered completely or nearly so. If it doesn't completely do so, I plan to conduct exposure tests to compute what it is I need to dial in for compensation. If the edges of the image circles vary to much from aperture to aperture, I'll mask things off to give complete darkness to unused areas and remove a metering and imaging variable.
The difficulty of bringing the image from the imaging lens back to the sensor turned out to be a brass biddy. My current testing is revolving around a zeiss 50mm lens reverse mounted on my Kodak DCS 520 (a 2mp dSLR made out of an EOS 1n by 'The Big Yellow' ages ago). In a pain killer/muscle relaxer fog, I left my 20d at my brothers last week! I HATE those pills . They make me stupid and don't really help hardly at all. It makes for yet another 'jimdandy" variable I'm coping with - Yippee
Reversing the lens, via a 49 to 52mm filter adapter taped to black matte board taped to the camera, gives me about a 1.2:1 macro lens (10mm = 12mm in the viewfinder aprox.). I positioned the midpoint of my second mirror so it was in focus in the camera. I then bounced the image from a 2nd 50mm lens back to the primary imaging mirror. Moderate success was achieved this way!
there wasn't much in the way of contrast or great sharpness in what was received, but my setup was hardly light tight. Also there was a distinct lacking in the matching of the angles between the primary lens and the secondary lens. The DCS-520 is positively huge compared to my 20d (it's a bit taller than an EOS1 with the battery booster! & it's heavy as Heck! Too). Well it makes it very hard to line stuff up, I have to balance the camera on the lens on the table top. The mirror's I'm testing with are cheap 2nd surface craft mirrors as well. My pricey 1st surface mirrors are a very thin and fragile - read won't stay flat! - and are going to wait for mounting on a smooth panel permanently which can be adjusted for convergence, etc, in the final design.
I think I might be on to some thing at this point.
I'll update later.
Thanks
Rich
PS.. how do I post a pic or two - simply cut and paste?