The science of it
[How to] [References]

          Stereoscopic vision is the result of having two eyes. Because our eyes are set slightly apart (typically 6.5 cm from pupil to pupil), each eye sees a slightly different image. It is for this reason, as seen in the [Cross-Eyed] instructions, that your finger will seem to double when held between your eyes and the screen, provided you focus on the screen while doing so. In fact, you can “bounce” your finger back and forth by allowing it to double, then alternately winking one eye and the other.
          In this detail from a [Gallery] picture, we see at least two differences between the images representing the left- and right-eye views:

1. On the left side, a light shadow line from the railing is visible next to Bruce’s head; on the right, his head has covered it up.
2. Because of this, there is little room on the left side for the “pocket” of concrete plainly visible on the right.

          The information from each eye is processed and combined in the brain to present a single image of depth. To reproduce this phenomenon artificially, one has simply to provide the necessary left- and right-eye images and a means for the viewer to process those images as usual. There are a few ways to accomplish this.

How to
[The science of it] [References]

          A single camera may be used without any special lenses, but the process can be painstaking. Essentially, a given scene is photographed once, the camera is carefully slid 6.5 cm to the left or right (recreating the natural distance between the eyes), and then the scene is photographed again. [footnote] The camera must remain horizontal at all times. Because time passes between shots, this technique is recommended for still lifes or other scenes where absolutely no other movement occurs. Other than the intentional camera slide, if either the left- or right-eye image differs even slightly — because someone walks by or the clouds move — the 3-D effect is diminished. You can see an example of this problem in one of my own pictures (Chess portrait), although I didn’t use this method. It happened because I was very close, and the shadows of other pieces didn’t quite match up.
          One way around the movement hazard is to use a beam splitter attachment. This is a relatively inexpensive apparatus which uses mirrors to split the image in two. The benefit is that action shots are allowed, since the camera doesn’t move. The down side is, naturally each picture is only half as wide as usual, so the images tend to be narrow. I used the Konex “Stereo Photo Adaptor,” available from Porter’s Camera Store, for each of my shots.
          Once you have a pair of images, they may be viewed side by side using either the [Parallel] or [Cross-Eyed] method. In the case of [Anaglyph]s, the required occular dexterity is minimal. The left- and right-eye images are overlapped from the start, but presented in opposite colors. The left eye, while wearing a red filter, can’t see anything red. It can, however, see other colors (especially green, red’s opposite), and since we want the left eye to see only the left-eye picture, that image is presented in green. The opposite occurs for the right-eye picture.

* For extremely large or extremely small (macro) shots, the rule of thumb for this distance is “one in thirty” — that is, a 1:30 ratio of slide distance to subject distance. If the subject is 15 cm from the lens, the camera is moved 0.5 cm, because 0.5 x 30 = 15. [Back]

References
[The science of it] [How to]

          If you care to reference this website, you may do so as follows:

Stiller, David. XYZ 3-D Photography. Internet URL: http://www.quip.net/xyz (Originally Posted 10/26/97)

I’ve personally gained insight from these books:

Burder, David, FRPS, and Pat Whitehouse, FRPS. Photographing in 3-D. Rev. 3rd Ed. Chicago: Black Box Collotype, 1992.

Photography. 4th Ed. Dubuque: Wm. C. Brown Company Publishers, 1982.

The Print. New York: Time-Life Books, 1975.