Stereo Depth Cues and Stereopsis
Not everyone has the ability to perceive depth through stereopsis. About 10% of the population is at least partially "stereo-blind". Stereo-blindness may be related to the malfunction of one of the eyes (muscular disorder, cataracts, etc.), or may be an entirely neurologic failing.
However, stereo-blind people can perceive some depth using non-stereoscopic cues. This can be proven by looking at any standard photograph - usually one can make some judgments as to the size and distance of objects in the picture. An understanding of shapes and volumes gives many cues about depth. Haze and atmospheric effects also provide depth cues on larger scales. More depth cues can be derived from motion parallax, mostly used when driving a car.
The Value of other Depth Cues:
DOF and Motion Parallax
Stereo photography does not mimic the limited depth of field of the human eye - and therefore does not perfectly reproduce the human visual experience. With the stereo photography on this page, I propose that the missing DOF (or focus range) information, as would ordinarily be available to the human eye, can adversely impact the spatial interpretation of the scene. Other missing depth cues, such as motion parallax, contribute to the special failure of this stereograph.
As you can see from the images, I went out on a recent snowy day, and, pursuant to my "policy" of seeking out subjects with spatial complexity and specular and transmissive qualities, found a wonderful snow laden section of chain link fence around a neighbor's yard. It was pretty, and it looked like an ideal "complex" three dimensional subject.
Imagine my surprise, when after receiving the slides, I discovered that the chain link fence has the property of completely scrambling all spatial information behind it! Apparently, the brain gets confused because the fence is rythmic and symmetrical - this allows for interference effects between the left and right view, which enable the fence to appear at several different depths simultaneously! In effect, in trying to look beyond the fence, it is difficult to avoid parallel free-viewing different segments of the fence, thus placing it at different depths.
A complex but random structure, such as the branches in the background above, are spatially interpreted without problem. But the fence with a horizontally repeating rythmic structure, becomes spatially ambiguous.
It is remarkable how this demolishes the spatial scene behind the fence. On the other side of the yard, about 20ft. away, is more fence. And you can see how it interferes perfectly (it is a "harmonic") with the fence close to the camera. The result is that the far fence looks to be the same distance as the near fence - it appears as a miniturized fence nearby.
Then it is also nearly impossible to get a spatial sense of other objects behind the fence, because they may also be aligned with similar segments in the foreground rythm. For example, the utility pole in the distance aligns with the same part of chain link (not the same link, however) in both left and right views. This makes its placement in depth space practically impossible.
Why did I not recognize this while I was examining my subject, composing the picture? I believe that both motion parallax depth cues and limited depth of field in human vision overcome the problems that are seen in the static slide. Motion parallax depth cues are constantly available, as you move your head slightly and continuously. Additionally, your eyes would also provide depth cues based on focus. At least for this scene, where the near fence is quite close, I don't think most persons would be able to focus sharply on the near fence and more distant objects simultaneously. In the slide this depth cue is removed by the greater depth of field of the camera lenses.
Maybe this very curious discovery will save some of you from wasting film shooting through chain link fences, or other horizontally rythmic screens. I imagine any kind of picket or iron grate fence would cause similar problems, as would shooting through rythmic patterns in sheer fabric. The key is to avoid rythmic screens that are regular in the horizontal axis, which is used by the brain for analysing stereo disparity. Then again, it makes an interesting subject for experimentation
Please mail comments and suggestions to 3dshow@starosta.com
For technical problems with this website, mail webmaster@starosta.com
Unless otherwise noted, all contents of this page, individual or aggregate, are copyright 1998 Boris Starosta. All rights reserved. Unauthorized duplication is a violation of applicable laws, and is destructive of free enterprise, creative expression, and the human spirit. All other product names mentioned in these pages are used for identification purposes only, and may be trademarks or registered trademarks of their respective companies, and the exclusive property of their respective owners.