For the cost of today's lenses, should they be diffraction-limited, wide open?

Joe Kotroczo wrote in news:ag45hrFlkp0U1
@mid.individual.net:

On 08/11/2012 23:32, RichA wrote:

(...)
There is a difference between "diffraction limited" and "sharpest at
full aperture" too... A lens that is sharpest at full aperture is not
necessarily diffraction limited.

What difference? What would cause a lens wide open and diffraction
limited (across the visual spectrum) to not be sharpest when wide
open? Resolution laws would argue otherwise.

Who says that a lens has to reach it's diffraction limit wide open? Can
it not be diffraction limited at f/11 or something, and less sharp at
any other stop, smaller or larger?

And who says that a lens which is sharpest wide open has also reached
it's diffraction limit? Can it not be sharpest wide open, but still a
long way off it's diffraction limit?

Yes.

After all, what "diffraction limited" really means is "reaching it's
theoretical maximum resolution".


At a given aperture. Aperture determines resolution provided a lens is
diffraction-limited. Camera lenses, by their normal nature are not
diffraction limited, cannot reach their maximum resolution because they
do not provide it wide open. Sharpness and resolution also are not
inextricably linked, a large optic can have higher resolution than a
smaller one with the image appearing to be less sharp. This is because
the diffraction disk produced by a large optic is smaller by proportion
(double optic diameter equals double the resolution). By the formula for
resolution, if camera lenses met the challenge of being diffraction
limited at their maximum aperture, an f/2.8 lens would outresolve an
f/4.0 lens of the same focal length simply because its clear aperture was
wider.

On 09/11/2012 23:39, Chris Malcolm wrote:
RichA wrote:
On Nov 8, 5:49 am, Martin Brown
wrote:
On 08/11/2012 00:27, Rich wrote:

$1800 for an 85mm f1.4 from Nikon. That's about 2x what the old one cost.
Is the lens $1000 better or should it be as good at f/1.4 as f/4? I'm not
sure. I know that some optics made as f/4.0 have been diffraction limited.
Pentax had some, but they weren't camera lenses. Some have claimed certain
telephotos in the pro bracket have been diffraction-limited at f/2.8, but
I've never seen it demonstrated. So, the question is, is it possible to
make say a 35mm, 85mm diffraction-limited at f/1.4 and if so, at what
price? Likely it is, but they haven't done it.

f1.6 and full achromatic mirror telescope has been done by Cambridge
University.

http://www.ast.cam.ac.uk/about/three-mirror.telescope

20" aperture prototype was built. That is one of the fastest diffraction
limited wide field instruments I know of.

You are hampered in real cameras by simultaneously wanting diffraction
limited and a flat film plane when the lens is fast and the small angle
approximations no longer hold. There is always a trade off.

Anything can be done in principle but the cost to manufacture it and
difficulties in assembly make it prohibitive. You could get the on axis
sharpness truly diffraction limited by sacrificing edge resolution but
never all at the same time and a flat film plane. Something has to give.

At around f4 or f5 things are a lot easier. Most real lenses tend to
have their resolution sweet spot at about that working aperture.

Only problem, with a central obstruction like it has, contrast would
suffer.

It should be possible with today's lens and mirror making technology
to devise a mirror which instead of reflecting straight back, offset
the folded beam offset to one side, thus avoiding the need for the
obstruction.

It can be done but the folded mirror designs are a bit exotic and only
valid for slow focal ratios like f10. An example:

http://bhs.broo.k12.wv.us/homepage/a...tevick/fsp.htm


--
Regards,
Martin Brown