Re: [photo-3d] seeing 3-D: innate

["John A. Rupkalvis" <stereoscope@xxxxxxxxxxxxx>]

This was a very fascinating test, that I have often wondered about.  The
problem is, how do you completely eliminate all monocular depth cues?  The
cube and box would have to be designed very carefully, such that from the
observer's distance and perspective the squares of the checkerboard pattern
appeared identical in size and shape at the different distances (they would
appear perfectly square only on-axis; from all other positions they would
appear to be trapezoids).  And lighting.  Any shadows, texture, reflections,
etc. would be very revealing.

The most difficult monocular depth cue to try to eliminate would be that of
relative motion.  It is very difficult for most people to hold their head
perfectly still.  The slightest motion of the head would instantly reveal
depth differences.

You can easily prove this for yourself.  Close one eye, hold as still as you
can (you will probably have to hold your breath), and observe that what you
see in your field of vision indeed does appear flat.  Now move your head
just a little bit (still keeping one eye closed).  You will immediately
perceive depth, in the sense that you can tell which objects are closer to
you, and to a great degree by how much.

Now, while moving your head, open both eyes.  You can certainly tell the
improvement that stereopsis does for your depth judgment.  But, except for
really close items, the difference is not all that dramatic.

Next repeat the experiment.  Only this time try to hold very still when you
open both eyes (perhaps even brace yourself against a wall or something).
When you now open both eyes, the difference may be startling.

This is part of the reason why we get a slight feeling of depth in some flat
motion pictures, where there are a lot of monocular depth cues and the
camera is moving, as a pan or dolly move.  It also has something to do with
why audiences watching stereoscopic (3-D) movies get a more pronounced
reaction out of seeing objects "come off the screen" than those that appear
to be farther away, except when the camera is moving (such as on a roller
coaster).

Anyway, does anyone know how the researchers performing the checkerboard
pattern experiment prevented their human or animal subjects from perceiving
distances through relative motion?  Most animals and babies don't hold their
head perfectly still for very long.

JR

----- Original Message -----
From: "Gabriel Jacob" <gjacob@xxxxxxxxxxx>
To: <photo-3d@xxxxxxxxxxx>
Sent: Tuesday, October 10, 2000 10:31 AM
Subject: RE: [photo-3d] seeing 3-D: innate


> Yes, that was an ingenious test, but it involved aprox. 6 month old
infants.
> By then there visual senses are quite developed. That particular study was
> conducted around 1960 I believe. There have been other experiments done
> with younger subjects (I don't know if that was human or animal
subjects!).
>
> Gabriel
>
> -----Original Message-----
> From: Robert A. Schreiber [mailto:bschreib@xxxxxxxx]
> Sent: Tuesday, October 10, 2000 12:15 PM
> To: photo-3d@xxxxxxxxxxx
> Subject: [photo-3d] seeing 3-D:  innate
>
>     Actually, one of the most famous studies in experimental psychology
> was designed to answer just this question.  Imagine a cube 3' x 3', with
> a checkerboard design on it..Then put a 3' x 6' piece of glass over it,
> so that the first 3x3 is on top of the cube.  The floor is then painted
> with the same checkerboard design...(I'm leaving out all the parts of
> the design to ensure safety)...And various animals, including crawling
> babies,  were put on the 3x3 cube.  If there were no 3-D, then you would
> expect equal time on both sides of the 3x6.  In fact, everything tested
> all have 3-D, and were scared to venture over the edge.
>     I remember this from undergraduate school, over 30 years ago, so
> most of the details are gone, but it's in almost every introductory
> psych book.
>     Our optic nerve has an optic chiasm at the base of the brain - half
> the information from each eye goes to the left hemisphere, and half to
> the right.
> Bob Schreiber
>
>
>
>
>
>
>
>