3-D is a really hot topic these days, and for those of you with lots of cash, Panasonic just released a new 3-D video camera that will connect to the new 3-D TVs, so with several thousand bucks you can get in on the 3-D action with home movies.
Step 1: History
But these antique viewers and sets of common slides are often available on ebay for around $100, and I also saw some very cute wooden "Holmes Stereoscope Kits" for sale there recently for $39.95. These antiques are admittedly primitive, but fun to pass around in the parlor after dinner.
At one time in the late '70s I made a rig using two Polaroid cameras mounted side by side that could more-or-less fire the two cameras simultaneously. These could be stapled or taped to a card and viewed in the Stereopticon viewer quite well. But very often there would be a miniscule delay with one camera or a slight difference in color or exposure, so the slides were never all that great. Today, the same thing could be done using two digital cameras, but the same disadvantages would exist. Plus, video would be challenging and require special software, etc.
Step 2: Idea Development
I was thinking about looking for one of the original double lens cameras that I assume were used to produce the original slides, but then had an idea. Why not make a device that will shoot the two images side-by side with almost any single camera and easily produce the 3-D prints and/or video images? Of course, many still cameras shoot both stills and video, not to mention new small HD video cameras that also shoot still great still pictures.
The 3-D photos would be easy to reproduce on any color printer as long as you have Photoshop or any software to size the prints at about 6 or 7 inches wide. When you shoot video (ideally 16 x 9 HD) with this device, you should be able to open the video file on your computer or iPad, size the image width to about six or seven inches on the screen, and watch it with a Stereopitcon viewer for astounding 3D home movies. I was so sure this would work that I decided to give it a try.
Step 3: Basic Layout
My son the SpaceClaim, Inc. CAD software founder (for real) sketched a basic drawing with the mirror, angle, and size lay-out, and based on this drawing I ordered two sets of front surface mirrors -- the small ones are 2" X 3" and the large ones are 5" X 7". These are available from several suppliers, including FrontSurfaceMirrors.com, where I paid about $45 with shipping for the four mirrors.
Step 4: Materials and Decisions
I happen to like working with acrylic and made most of the parts from scraps lying around the workshop. But wood could easily fill the bill for most of the elements. The top and bottom plates are 1/4" material and measure 4 1/2" X 18". The posts that support the top are 1/2" material, 5 1/8" high -- just enough space to clear the large mirrors. I used a few pieces of 1 inch thick acrylic for the mirror supports -- but all of these could be made from wood as well. Frankly, the top is probably not really needed, but I decided one would help shade the lens and protect the mirrors. The top also provides an upper bearing hole for the convergence adjuster that I eventually added, but that whole system is not really necessary for you to start shooting.
Step 5: Mounting the Mirrors
For the mirror support blocks, I carefully checked the 90 degree angle of my jointer and made one long piece for cutting these to length to be sure all of the mirrors would be true vertical when mounted. The small mirror mounts have both the 1" base plus a 1/4" thick extension to raise them to the necessary height. The mounting blocks are just glued to the glass side of the mirrors with a dab of Gorilla Glue. I advise drilling the mounting holes prior to attaching the mirrors, using a drill press so the holes are reasonably accurate as well. The large mirrors are mounted with the bottom edge of the mirror sitting directly on the base, and the small mirrors are mounted at the proper vertical height to be centered in the 5 inch high large mirrors.
For the mirror positioning, I printed the CAD drawing full-size, taped it to the bottom plate, and used a center punch to mark the exact locations of the mirrors. Please note that the drawing shown is my original and is a bit battered plus has some random lines that are to be ignored.
Step 6: Mounting a Camera
You will have to build the rear camera mount to match your camera, both in terms of length and height -- the vertical distance from the base of the camera to the center of the lens should be designed so the lens will line-up with the center of the small (and large) mirrors. In this case, I used a 1 1/2" wide piece of 1/2" thick material that is glued to the bottom of the base plate and extends 5 1/2" beyond the rear edge of the base. I put a slotted hole in mine so I could fool around with the camera to mirror distance. For the EOS 7D, a 1" elevation block brought the center of the lens up to the center of the mirrors. This does not have to be anything fancy, but it is very important that the camera end up level from left to right -- the eyes have an amazing ability to converge on the horizontal plane, but they don't do very well when there is a vertical miss-match between the left and right.
Ideally, you will want to be able to slightly swing the large mirrors on their single mounting screws to make the two images converge correctly for the distance to the selected elements of your photo. This is what your eyes do naturally at all times so that your left and right images converge properly, and of course, proper convergence is a key element of professional 3-D shooting.
Step 7: About the Convergence System
When doing my initial tests, I simply left the screws on the large mirror mounts slightly loose and manually set the two sides so the left and right images appeared to be similarly composed. I could compare the positions of the two mirrors to the base plate to keep them more-or less equal, and found that this was accurate enough. But the process was clumsy and I eventually added a more convenient system to quickly adjust the two large mirrors for convergence. This is the most complicated part of what otherwise is a very simple project, actually.
There are probably many ways this convergence control could be designed, but I used a small cam on a shaft that pushes a 1/4" aluminum rod rearward. At the end of the rod there is a piece of plastic that pushes on a pair of aluminum rods -- one that is attached to each of the large mirror mounts. Holes for these rods are also best made prior to gluing the mirrors onto the mounts, and with my system there are also small holes for pins that hook onto the return springs. The return springs attach to the outside ends of the large mirror mounts to pull against the cam, push-rod, and mirror rods. I also used little push springs on the large mirror mounting screws to allow them to move quite freely but stay flat against the base plate.
Step 8: Summary
The photos are admittedly a bit difficult because of the unintended reflections of various surroundings that invariably show in the mirrors when you try to photograph this device, plus trying to see the shiny and/or clear acrylic parts in this prototype.