>Normal lenses introduce light in a sort of cone shaped projection against
>the film plane, while digital lenses should and now do project the light in
>parallel almost a rectangular shaped prjoection that fall upon the chip.
Hmmmm ... I think what is more the case is that given any light receptor in the
CCD image area the cone of illumination from the lens arrives at the sensors
located close to the edges of the chip in a more normal direction than with a
conventional lens. The reason that this is desirable with electronic sensors is
that they are not exactly on the surface but are located in a (slight)
depression (might simplyfy this to read: at the bottom of a short tube) and
this makes them susceptible to collecting less light when a certain amount
(controlled by the aperture) is arriving at an angle than straight down the
tube.
If I recall, film is also sensitive to the direction of incident illumination
(actually the distribution of energy) and several effects contribute to the
fall-off (quite severe in extreme cases) in illumination at the edges of an
image plane.
I think (I have no factual basis for this and am only speculating) that in a
"digital lens" the diaphragm, when looked at from a pixel located near the
edges of a CCD array, appears to be located more directly in front of the pixel
than when a conventional lens is used. This would make the apparent location of
the diaphragm appear to move up-down-sideways depending on where on the image
plane one looks at the rear of the lens from. Quite an optical "trick" if you
ask me!
I think this would be most desirable with short focal length (well, wide angle)
lenses but would not be of major significance with long focus or most telephoto
lenses.
ok ... now 'tis time to wake up and make breakfast.
andy
