Re: fish eye lens test

[ADavidhazy <ANDPPH@xxxxxxxxxxxxxxxxxx>]

Alan,

<snip>
> ... As predicted there is no way to get the full 180 vertical coverage. 
> I think the lens will be best with ambiguous space without straight lines. 

This has been my mischievous way of saying that whenever the angle of view of a
rotating panoramic camera extends beyond the axis of rotation of the camera one
might be illustarting how things look "beyond infinity" ... infinity being
defined as that location in the photograph where there is an infinite amount of
distortion ... where a "point" ... the extension of the axis of rotation to a
location above the camera ... is reproduced as a line. 

I had placed an illustration of this effect in a webpage devoted to my
experiments with making a digital camera out of a hand scanner and not only
does the photograph approach the point directly above the camera but
significantly exceeds it. I was trying to make a point about this and so did
not just point the camera straight ahead and included both the North and South
"poles of rotation" but decided to only include the North and go beyond it.

This, of course illustrates that subjects beyond the infinity line are
reproduced as being 180 degrees out of phase and upside down with respect to
their "real" counterparts. It is easiest to see this effect by example. The 
photograph is at: http://www.rit.edu/~andpph/text-better-scanner-cam.html

A side effect to this discussion is that raising the angle of view of the
camera by tilting it backwards would cause converging verticals in a normal
photograph but in a rotating panoramic camera, as long as you keep the axis of
rotation vertical, the verticals in the scene will be reproduced as verticals.
Sure, there will be more and more unsharpness due to image/film slippage but
otherwise the lines will not converge. This unsharpness can be reduced by small
slit sizes ... the smallest ones being those that are not slits at all but
photosites in a linear CCD array. And, BTW, with these variants on our film
camera conterparts, rotation direction does not matter as there is no moving
photosensitive material in the camera! The only thing that matters is TIME.
                       
happy rotat'n

andy

Andrew  o o  0 0 o . o  Davidhazy, Imaging and Photo Tech
         \/\/\/\/\/\/          http://www.rit.edu/~andpph
__________|        |_____________________________________