Re: Dr. George's querry

[P3D John W Roberts <roberts@xxxxxxxxxxxxxxxxx>]

>Date: Tue, 29 Oct 1996 14:02:59 -0600
>From: P3D  <Al_Razutis@xxxxxxxxxxxxxxxx>
>Subject: Re: Dr. George's querry

>Since photo-3D appears in an "edu" site, one can presume that the
>administrators are educators
>and that site activities are monitored with that preference in mind.  

Careful about drawing inferences - when the list and its charter were
created, it was a "gov" site. (I don't think that means we're primarily
concerned with governance. :-)

>...So, in answer to your querry, and without invoking (that is,
>citing) the specifics of
>"saccadic eye movement" and "feature rings" (well known to physiologists of
>perception), I offer
>a quote from Paul Virilio's essay "The Vision Machine" (Indiana Univ. Press,
>1994):

>"The veracity of the work therefore depends, in part on (the) solicitation of
>eye (and possibly
>body) movement in the witness (viewer) who, in order to sense an object with
>maximum clarity,
>must accomplish an enormous number of tiny, rapid movements from one part of
>the object to
>the another.  

I agree - as far as detailed vision is concerned, the eye is primarily a
scanning device. In fact, if the scanning is suppressed, the image starts
to fade out.

>I am also unaware of any
>research that has
>proven the non-existence of saccadic eye movement and its role in graphic and
>spatial
>perception and cognition.  Looking through a tunnel (viewer), at a
>fixed-distance screen (SVGA
>monitor or TV) obviously denies 'the eye's motility'.  

Incorrect. The fovea (the area of sharpest vision, which is scanned over
a scene to build up a visual model in the brain) has only about a 1 degree
field of vision. I haven't found a reference to the magnitude of the saccades,
except that the motions are "tiny". Even a Viewmaster view is many times
the field of view of the fovea, so it's still possible to allow the eyes to
roam around within the field of view.

Most of the photographers on the list like to mount their images so as to
create a "stereo window", with most (usually all) of the scene "behind the
window". This can make the scene appear more "natural", since it recreates
what a person would see when looking through a physical window in a house.
(Though unaided human vision has the "stereo window" set at infinity.)

>My personal experiences in ongoing screenings of 3D video to a variety of
>individuals confirms
>that there is no 'typical' viewer, no predictable perceptual response, no
>collective subjectivity, no
>standard receptor (physiological or psychological) by which one can guage the
>results.  

I agree that there's tremendous variability in how people perceive depth.

>I have yet to hear that
>fixed-gaze viewing of stereo-pairs or stereoscopic viewing of motion-picture
>scenes would
>conform to the viewer's previous experience of perceiving the world of
>objects and things(!)  It
>would only be possible in a world where peripheral vision and body motion was
>'outlawed'.

Ferwerda emphasized that freezing motion would spoil the realism of a stereo
view, and recommended that any people included in stereo photos be in
passive poses (i.e. sitting). I seem to be less sensitive to "temporal
realism" than that - a 3D snapshot of an action scene doesn't bother me
any more than a 2D snapshot.

Does anyone know the field of view of the LEEP system offhand?

>In the making of stereo-3D art there is a compromise enforced by the viewing
>apparatus which
>becomes as important as the selection of subject matter, a compromise further
>compromised by
>the range of perceptual abilities on the part of the viewers.  In related
>mediums like holography, I
>noticed that viewers exposed for the first time to holographic images in the
>early 70's had less
>problems 'seeing' a virtual image (behind the plate) of a concrete-looking
>thing than they did a
>pseudoscopic 'real' image (appearing in front of the plate) of the 'thing'. 

What people expect to see, and their familiarity with the medium both
play major roles. There's an old story that in a 19th century war, soldiers
had an opportunity to send a photograph home to their families. Rather than
go to the trouble to photograph each soldier, copies were made of just two
photographs: "soldier with beard" and "soldier without beard" - it was
hoped that the recipients would not be able to detect the counterfeit because
of their unfamiliarity with the medium! (I have no idea whether that story's
true or not.) Nowadays, there's a certain amount of material available which
is advertised as 3D (stereoscopic), but isn't. (I'm *not* referring to the
2.5D vs 3D debate - I mean things like television programs which purport to
use the Pulfrich effect but don't.)

>The
>science of perception can add to our understanding of 'how we see' but it has
>little to say about
>'what it means to see' and 'why we interpret what we see as meaning such and
>such...'.  These
>issues and their philosophical counterparts are part of the mystery of
>artistic expression.

But as the physical structure and working of the eye and brain are better
understood (we're still a long way from being able to state that we actually
understand how vision and perception work), many of these issues are gradually
moving from philosophy and psychology to physiology (including neurology). 
That's not to say that *all* the issues will make the transition.

John R


------------------------------