remember those old stereograph cards from the turn of the century? I'd like 
to do that...except by using my digicam

I only have one digicam (Canon G2), so taking  "action pictures" will be
practically impossible, but it should work out for still-lifes and such.
  
This is the best info i've found so far:

http://www.acsu.buffalo.edu/~kswiatek/StereoIntro.html

  "The stereo base lens separation should be one 30th the distance to the
   nearest object in the scene. Example: if the nearest object is 30 feet
   away, the lens separation should be 1 foot. Most stereo cameras are 
   factory set at the interocular distance of 2.6 inches thus suggesting
   the nearest object should be at (2.6 X 30) 6.5 feet, which is sufficient 
   for average photos of people in the environment." 

This is what I think i'll do:

I think I can find an old stereograph card at the flea markets around here.
I can use that to duplicate the dimensions. Then I need to find a viewer or
somehow construct one....that'll probably be the toughest, as i'm trying to
keep this a lo-budget thing.

I'll keep ya'll updated on my progress.

13 responses total.

You might follow some on ebay. I just checked and the really nice old ones
have gotten pretty expensive, but some later, practical stereoscopes seem
reasonable, and parts of one are even cheaper (e.g., one missing the card
holder, which would be easily made).

Sounds like a really neat project.  Let us know how it turns out.

ok...so I went to the antique market to find an old stereogram card.
The card I got has a picture of King's Chapel in Boston, with people
in the foreground. here is the scanned picture with the two frames 
overlapping:

http://members.triton.net/eprom/stereogram.jpg

although the images look identical, when viewed through a pair of 
stereogram glasses, the boy in the bottom right really jumps out 
from against the building.  there is only a 1 mm difference between
frames.

I'm trying to find a formula to calculate both the interocular
distance the cameras should be spaced and if I need to stay within
a specific range of focal lengths so the images 'jump out'?

I read somewhere that it was 1:30 ratio, for every 30 ft, the
cameras should be spaced 1 ft apart. But that don't sound right,
considering that (in humans) the distance between the eyes don't
change when viewing an object. also the closer an object is, the
eyes tend to cross, I'm wondering if the same thing happens when
viewing a distant object?

I checked out the design of the stereogram viewmaster (or whatever 
its called). It looks pretty simple. I estimated that the card should 
be around 15cm away and the eye pieces are just 20x15 mm x2 magnified 
convex lenses angled slighting inwards.

One of the classes I'm teaching this term uses photography as it's focus for
teaching required first-year curriculum topics.  The Mechanical Engineering
prof who is co-teaching understands all this very well.  Let me see if he will
give us some clues.  

I measued my pupil distance: 61mm
anyone know the standard deviation?

Re #3: please repost the whole card. I've trained myself to view stereograms
without a viewer. It is a matter of relaxing your eyes to have each eye
fix on its own image - and then bringing them to focus. I'd like to see
if this works on a scanned stereogram - that is, whether the screen
resolution interfers with the 3-D illusion or not. 

Ok...its at http://members.triton.net/eprom/original.jpg (131KB)

I scanned it at around 300 dpi, so its larger on screen than in real life,
so with my 19" monitor I can almost make my eyes see the image pop-out.

It works. I couldn't resolve the image at first, which filled my 16"
screen, but grabbed it into Photoshop and reduced it to the usual
stereogram size.  That worked fine. Even more distant objects - the lamp
in front of the building, for example - resolved into 3-D. I think part of
the problem I had in resolving the full-screen image is mental. It takes
an act of will to relax one's eyes and the larger image put me off. 

I'm not sure comparing to the distance between a human's eyes is going
to be helpful for anything except close-ups.  Stereoscopic vision
doesn't actually help you judge depth at longer distances, precisely
because your eyes are too close together.  Beyond a few meters you're
relying on other cues.  The wide camera spacing for a stereogram card
would be to force that effect at longer distances.

hmmm....I see...that makes sense

Yesterday while up in Grand Rapids I saw a IMAX movie "NASCAR 3D". I have to
saw, that was one of the coolest things i've ever seen. I saw a 3D IMAX movie
in Philly but it didn't even come close to comparison.

well, it's only been two years, but anywho....

I was at Barnes and Noble this past weekend and stumbled across a book
called "The Universe in 3-D" (ISBN 0760766088). The images inside the 
book are really crappy, but I bought it for the 3-d glasses mounted in 
the cover flap. it was only $10. So here is my first digital stereo
photo... (don't pay no mind to my messy apartment; it was a test shot).

http://homepages.wmich.edu/~j4castee/stereo01.jpg

Increase the interocular. 

Response not possible - You must register and login before posting.

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss