One of the beautiful things about photography is that there is always something to learn. Recently, I was talking with a talented young photographer who introduced me to the process of stereoscopic photography, otherwise known as 3-D photography.
The idea behind it is simple. Take two images of a subject from two slightly different perspectives. Then in Photoshop, edit the images to filter each perspective to a different color field (seriously a one click process). Line up the images. Then, using a pair of red/blue anaglyph 3-D glasses, the image has amazing depth.
So let’s start at the beginning. My first stab at 3-D photography was done on a standard tripod in my living room. I set up my girlfriend’s view camera. I took two images, shifting my position ever so slightly, attempting to match the distance between my eyes.
I know the frames look identical, but I promise, they’re not. Look at the slight gap between the top left of the view camera and the cameras on the shelf in the background. A little yellow shows through on the image on the right.
Once you have your two images, bring them into photoshop. Copy the second image and paste it on the first image, creating a new layer on top. For many of the images, I desaturate them at this point. The process works in color but sometimes things look sharper in black and white.
Once you have your two images stacked in layers, double click the top layer on the layers window so it opens the layer style menu. That is where you make the one click. Uncheck the “R” box where it has separate RGB boxes. Click ok.
At this point, your image will have the look of an anaglyph 3-D image with the red and cyan edges showing around objects. However, the image isn’t aligned yet. You can see the separate elements on the view camera don’t match up at all.
Simply move the top layer to match up your focus point. On this image, I focused on the lens. I’ll adjust the top layer so the view camera’s lens lines up perfectly between the two layers. Also, this is a good time to crop out the parts of the frame that no longer overlap after you’ve repositioned your layers.
Viola. There you have it. A finished anaglyph image. Dig through your basement of stuff from the 80’s and 90’s until you find a pair of anaglyph glasses, or pick up a pair on Amazon for super cheap.
Here are a few more frames I made that night, just eyeballing the distance between the frames. Some turned out better than others.
A day or so later, I decided to try this technique outdoors, hoping to find places where I could play off the depth. Again, some worked better than others. I was still eyeballing the distance between the frames at this point.
That’s when I got to dreaming. What if I could do this with action? I had been using one camera and moving it slightly to capture the two different perspectives. What if I could capture the two perspectives simultaneously by using two cameras in perfect sync? I knew it could be done. Legendary Sports Illustrated photographer Dave Klutho has been shooting 3-D images for SI for kids and Time for a while now. I’ve seen pictures of his dual camera rig. If you look closely, you’ll notice that one of the cameras is mounted upside-down. It’s set up this way to get the lenses at the optimal distance apart from each other. So I set out to build my own.
Before I got to start my DIY stereoscopic SLR camera mount, I had the opportunity to participate in a photo show where photographers went around town taking portraits that were shown in a gallery that same evening. I decided to try to shoot all my portraits in 3-D. These were all shot completely handheld.
The next day, I had time to get back to the DIY 3-D rig. Every camera will vary in size so you’ll need to measure before you make your own. I cut two 13″ pieces of aluminum for the top and bottom parts of the frame and used 7″ pieces for the sides. They are held together with L brackets on the corners with 3/4 inch 1/4 20 nuts and bolts. There is a hole drilled on the bottom center of the bracket to mount it to a tripod. There are also holes drilled, one on the bottom and one on the top, to match the location of the tripod mount on the bottom of my camera.
So this big aluminum rectangle with a few holes changes everything. Now, with wiring, I’m able to trigger the two cameras at the same time, capturing the same moment from two perspectives. I had it put it to the test.
I called on my co-worker, Brent Lewis, for his help. We spent some time getting the wiring rigged and then we began to test. We shot it the studio at the newspaper for a controlled location.
The first shot, we just cranked the camera’s ISO and fired to see if they were shooting in sync.
We then got a little more creative and tried to light it. We turned off all the lights in the room and did a bulb exposure, firing the flash at the point of peak action.
I then wanted to see what things moving toward the camera would look like. I threw some fake flowers toward the lens while Brent tripped the shutters and lights.
I then got in a fierce, mid-air battle using a duster as my weapon.
The depth and level of detail is persistent event as you crop way in.
Lastly, for my friends that don’t have anaglyph glasses handy, I made that last image into a wigglegram. It’s an animated GIF that alternates between the two frames quickly enough to make it look 3-D
I still have things to figure out. One thing we noticed is that things coming toward the camera look strange and are hard to focus on. It might have to do with using a wider focal length. It may have something to do with the lighting. There is much experimentation to be done.
Be sure to pick up some anaglyph glasses and stay tuned. I have some exciting ideas that I hope to capture soon.
Objects moving toward the camera have a double problem in that the parallax is changing and the magnification is changing. (The latter-foreshortening- is definitely related to using a wide angle lens)
It is impossible to adequately fuse or focus on something that is changing in these two ways. One is bad enough but the combination is insurmountable.
To carry the thought further, an object that has depth and is moving toward you is that the “Fixation Disparity” of the proximal (nearest) portion is different than that of the distal (more remote) portion and the object cannot be fused completely and this lack of fusion while moving makes it look blurred.
Thanks for the science, Mike! I’ll have to find the ideal situations to play with this technique. Have any thoughts about what would work best or what wouldn’t work well at all?