At 18:51 10/3/99 , Jan wrote:
>So if good wide angle lenses (and fisheyes) don't "distort," what do
>they do that people call "distortion?" I don't know what to call it,
>except something unwieldy like "perspective shift." Suggestions?
Perspective shift might be confused with a PC or "shift" lens, which shifts
a larger image circle around the film gate. These are inherently
rectilinear lenses. The Zuiko 35mm shift actually has an image circle
equivalent to a 24mm lens. The image circle is big enough to move it about
12mm up, down, left or right before the corners of the film gate hit the
edge of the image circle. The film gate only captures a field of view of
the image circle equating to a 35mm lens.
>From the literature I've read, there are two basic types of lenses used in
photography:
- rectilinear
- spherical (this may not be the exact term for it; equidistant maybe?)
The rectilinear lens is what we are most accustomed to. It maps flat
fields in space to a flat field of film. If you were to aim a perfect
rectilinear lens at a brick wall so that its axis is perpendicular to the
wall, you will get a true rendition of the brick wall with all mortar lines
straight. This is what most grids for lens tests are set up for. Barrel
and pincushion distortion on these types of lenses describes how much
straight lines will diverge from being straight in a flat subject
perpendicular to the lens axis.
The "spherical" (equidistant?) lens is the theoretical fisheye. It maps
spherical fields in space to a flat field of film. A round polar map is an
example of this (albeit convex sphere to flat surface, but the end result
is the same). Thus if you aim one of these perpendicular to a flat brick
wall, the mortar lines will converge as you move towards the edges because
the edges are farther from the lens. This is not distortion, unless how
well it maps a concave spherical field to the flat film is inaccurate. I
guess I would call it an equidistant image.
I too get disturbed when fisheye images are referred to as inherently
distorted. Actually they're not provided the lens accurately maps a
concave spherical field to the flat film properly.
-- John
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|