Stereoscopic Scanning

A little history
The technique of shifting a camera to get left and right views for stereo still lifes has been known for over 150 years. For over 50 years it has been known that photocopiers could be used to take pictures of more than just documents, keys and other small objects as well as hands and other body parts soon began appearing in photocopies. When flatbed scanners were introduced with similar optical properties it wasn't surprising that various objects were soon scanned. It seemed inevitable that people familiar with the "slide" technique would soon apply it to flatbed scanners.

The first documented use of sereoscopic scanning found so far was by Bob Wier on Dec 14, 1995, when flatbed scanners were still mainly an office item and few home computer users could afford them. It is interesting to note that he refers to previous experiements by others. These early users didn't know enough about scanners to explain why stereoscopic scanning worked, they just had a hunch that it would and found out by trying it that it did.

I first discovered stereoscopic scanning in late 1997, soon after getting my first flatbed scanner. By march 1998, I was regularly posting the results to alt.binaries.pictures.stereo. Like many people, I came up with the idea independantly by just trying it, unaware, but not surprised, that many others had done stereoscopic scanning years earlier.

In the summer of 1998 Richard Schubert wrote about how the properties of flatbed scanners made stereoscopic scanning possible and referred to it as "recently discovered" apparently unaware that stereoscopic scanning had already been used for years. He soon got a German patent on a system that automatically shifted the scanner to produce multiple views which were then used for lenticular printing. His website offering it for sale was terminated in 2003 and now he just sells artwork produced by it.

How is stereoscopic scanning done?
It's really quite simple. You place an object on the glass, scan it, move it sideways and scan it again. This produces pairs that can then be edited with a program such as Stereo Photo Maker to make stereo pairs, anaglyphs (red/blue), JPS files, etc.

first view second view

The first experiments in stereoscopic scanning were done by just placing an object on the scanner and moving it over by hand, but it was soon discovered that better results could be obtained by using more sophisticated techniques such as the glass front box show above. This limits undesired shifting and rotation of the object. Objects that are prone to moving or which need to be arranged a certain way can be glued to plain index cards which are then trimmed as necessary before being placed in the glass front box.

method for smaller objects, this slide was severl years old and so the label has yellowed slightly

Smaller objects such as seeds can be secured to a microscope slide with white adhesive labels. When the brightness is adjusted for darker objects the background gets washed out and the objects appear to float free in space.

Colonial (tabulate) coral
visits since February 25, 2012:
visits to stereoscopic scanning page
A few tips:
  • Be sure you clean the scanner and any other glass before scanning, dust, smudges and small fibers can be a big problem, especially at high resolution!
  • You must move objects sideways across the scanner, no parallax is produced by sliding them up and down!
  • The best focus is close to the glass and near the center, so do the first scan slightly to the left of the center and the second slightly to the right
  • The amount of shift needed will vary depending on the object and the scanner but 2.5 inches usually works best. I suggest experimenting with various distances see demo here
  • Higher resolutions have less depth of field. Because of the mangification often achieved, shallower objects often show amazing depth, and are usually in better focus
  • Most scanners don't scan all the way to the top, so an object such as a small ruler with 2 straight edges can be placed at the top of the scanner and the box or slide can be pressed against the other edge.
  • You can also remove the scanner lid and turn the scanner upside down, then move the scanner itself mechanically. This works well for extremely flexible objects and objects such as small plants that can't be turned upside down.

You should now have a pretty clear idea of how stereoscopic scanning works. I hope to write additional pages with a more detailed treatment of specifc aspects of stereoscopic scanning and more specialized applictions.

Some fine examples of the use of stereoscopic scanning can be found in my Stereoscopic fossil gallery, stereo seed gallery, Grade F amber gallery and others. Some cruder examples originally scanned in late 1997/early 1998 can be found here.

3d galleries
$6.95/mo? What a deal!!!!

3Dham's homepage