hi--

> Perception fascinates me. that's how I got
> into photog'y. I recent histroy, I shot an
> interior with a swing lens, a fisheye [full],
> and a 14mm [rec], from one spot. Then I
> chose to objects near the edge as ref points
> in the prints, printing so all prints had
> exactly 200mm between the ref points. After
> spending a long time considering how the 3
> prints varied the appearance of the interior,
> I came to the conclusion that the 14 was the
> worst rendering -- rectilinear but with tiny
> imaging toward the center and objects looming
> huge and stretched-looking in the outer area.
>
> The swing lens image can be reasonably
> decribed as the fisheye effect on the x-axis
> with the rectilinear effect on the y-axis.
> Size relationships from center to edge where
> close to the impression the observer gets by
> naked eye viewing the actual scene. But the
> conflict between fish and rect on the two
> axies produces something less credible than
> even the all-axis radial effewct of the
> fisheye. Swing lens pix only view properly
> if you view them as cycloramas, which means
> curving a large-ish print about you head and
> locking your eyes dead-ahead so you must
> turn your whole head rather than scanning by
> moving your eyes.
>
> The fisheye would probably work better in a
> planetarium-dome-like projection viewing but
> the film that recorded the image was flat
> so the projection will be flawed. I think
> a fisheye digital camera and software to
> display images via virtual reality helmet is
> going to be the best "Viewmaster" ever. I
> used a Globuscope for a while [35mm Cirkut
> type, 360 or more] but it just about drove
> me over the edge. Here's visual mental
> gymnastics: describe a projection system
> for cycloramic projection, outward from ctr,
> to project 360 dgr strips of chromes, and
> a segmented appoach [multiple overlapped
> conventional lenses] is taboo. It's sort
> of similar to imagining making a panoramic
> pinholer with no moving parts from a Quaker
> Oats box or coffee can.
>
>Merry,
> - dr David Rosen

There are difficulties...;-)
I have seen fisheye photos projected into a dome,
and so long as you sit very near the projector,
the illusion of proper rendering is quite good.
Short of that, I can still accept the fisheye
image as being most "correct", though when viewed
as a small flat object, it does have a very
unfamiliar look to most people, alas... (I like
'em, though! ;-) Kinda sorry I didn't buy that
Globuscope from you, when I had the chance...
Been into video a lot lately, often with a fisheye
converter firmly attached to the front - "floating"
a super-wide around things sure does show the
advantages of barrel "distortion" when moving
the field (a super-wide "undistorted" image could
drive one nuts with the rapidly shifting edge
magnification...! ;-).