Re: jA Comment on the Lenses

[Ron Wickersham <rjw@alembic.com>]

hi Tom,

while i can't claim to be the optics expert you requested, i can account
for the effect you're seeing.  there are two reasons for light to fall off
at the corners of the image.  since the CCD is flat, the corners are farther
from the lens than the center is.  this first effect occurs even with a 
pinhole camera and is obviously greater when you make the system cover a
wider field.  this fall-off is the fourth power of the cosine of the angle
of incidence of the light on the CCD surface.   (there are some lens
designs which are better or worse than the pinhole camera but most are
fairly close).

then there is a second effect, mechanical vignetting.  the lens designer
had a stop (which may be an actual diaphram, or it may be the smallest
lens opening in the system) which is called the aperture stop.  in the
Mark IV lens, the small drop-off indicates that the entrance lens is
large enough so that off-axis light isn't cut off very much by the entrance
lens opening.  but if you add an entrance stop, and effectively make the
entrance lens smaller, then on-axis rays will fully illuminate the internal
aperture stop, but off-axis rays going to the corners will only partially
illuminate the internal aperture stop.   

another way to see this is imagine that you made a glare shield forward of 
the entrance lens that was the same diameter as the lens itself.  for on-axis 
rays (going to the center of the CCD) it wouldn't make any difference how 
far you extend the glare shield, all the light would still make it into the 
lens.   but for off-axis rays some will be cut-off because the front opening
of the tube obscures part of the rays.  if the tube is made long enough then
you can cut off all the rays that would otherwise go to the corner of the CCD.

so the solution would be to make the real aperture stop that was designed
into the system smaller.  then you'd get the benefit of equalizing the
exposures and reducing the lens abberations without the extra fall-off of
light in the corners that's now being caused by mechanical vignetting.

-ron



On Mon, 4 Nov 2002, Tom Droege wrote:

> As noted earlier, the wide open I lens is down 5% at the edge (on a circle 
> tangent to the sides of the CCD) and 10% in the corner.
> 
> I note that now I have stopped down the I lens, it is down more like 20% in 
> the corner.  This surprised my feeble knowledge of optics.  I thought 
> everything got better as you stopped down lenses.
> 
> Possibly an optics expert can respond?  In any case, we are still 
> considerably superior to camera lenses which are often 50% down in the 
> corners of images of our size.
> 
> Tom Droege 
> 
> 
>