On DIFFRACTION in Lenses...
OK, *in general*, stopping a lens down tends to
progressively minimize lens optical problems related
to resolution, but ALL optics are subject to the
reverse-direction effect of diffraction, which
progressively limits the maximum resolution possible
as the lens is stopped down. In other words, if you
put a simple lens (a magnifying glass, for instance)
in a tube and mount it on your camera with proper
focus established, and add opaque covers on the front
(preferably close to the glass...) with ever-smaller
openings in the middle (reducing the lens aperture,
or the area that light passes through), the image
quality will go from terrible to somewhat better
(or even decent, at a small enough aperture). If
you continue the process, though, you will notice
that the image resolution will stop improving at
one point, and will then begin to decline with
smaller stops. This is due to diffraction, and is
caused by the smaller size of the opening forcing
a larger proportion of the image to be formed by
the diffuse, undirected light rays bent around the
edges of the diaphragm, compared with the light rays
that go straight through (away from the diaphragm
edges). This effect is absolute and predictable, in
that for a given FL and aperture, you can give the
resolution limit due to diffraction, regardless of
the lens design and quality. In practice, most lenses
are design compromises, and VERY few (as in virtually
none...) are designed to be diffraction-limited
wide-open. The better lenses are diffraction-limited
at a fairly wide stop (around f2.8-4 for video,
around f5.6-8 for 35mm still lenses, around f8-11
for 2 1/4 cameras, about f16-22 for 4x5, etc.) in
the center of the image. Since corrections (made
by using multiple elements in the lens design) are
generally worse as you look toward the edges of
the image (or the corners of the rectangular frame
cropped from the circular coverage of the lens...),
diffraction-limiting there usually occurs at a
smaller stop. As a result of all this, lens
resolution generally improves with stopping down,
but when the diffraction limit has been reached,
further stopping down degrades the image resolution,
and this second effect is independent of lens
design quality, which can only place the point at
which diffraction effects begin. If all lenses
of a given FL are diffraction-limited at a given
f-stop, then all of these will likely have the same
resolution at this stop (and cannot have higher
resolution). If you go to:
www.David-Ruether-Photography.com/diffraction.htm
and compare the images carefully (shot at the
stops available with the lens on the VX2000), you
will see that past about f4, fine detail begins to
diminish due to diffraction (you can see the
diffraction effect in the point reflection on the
van - as the diaphragm is stopped down, the points
on the "star" grow [with each point caused by
diffraction around the straight edge of a diaphragm
blade - with a vertical line being a shutter-CCD
effect] - this is what happens with all the "image
points" in the picture...). As you go towards wider
stops, the resolution also declines, though not
much, in this very good lens... As with all imaging
systems, though, the final image resolution is
determined by a combination of medium and lens
resolutions, and it cannot be greater than either...
DR
On Tue, 16 Jul 2002 22:55:18 +1000, "Hughy"
>OK - I'll 'fess up.
>
>Tell me about "diffraction limiting" ..... in *very* simple terms if possible,
>please David.
>
>Regards,
>
>Hughy.
>
>
>"Neuman - Ruether"
>news:3d37147d.6264468@newsstand.cit.cornell.edu...
>> On Mon, 15 Jul 2002 06:18:45 GMT, "UrbanVoyeur"
>>
>> >"Neuman - Ruether"
>> >news:3d32bccd.4546121@newsstand.cit.cornell.edu...
>> >> On Fri, 12 Jul 2002 18:35:29 GMT, "UrbanVoyeur"
>> >>
>>
>> >> >I'd like:
>> >> >a better lens
>>
>> >> Well, the VX2000 lens appears to be diffraction-limited
>> >> by about f4, and good to the corners and close to the
>> >> f4 performance even wide open (see:
>> >> www.David-Ruether-Photography.com/diffraction.htm), with
>> >> excellent color, contrast, and freedom from flare, so
>> >> I assume you mean better zoom control? Wider zooming
>> >> range? Or...?
>>
>> >No. I really meant a better lens - sharper, higher resolution, more
>> >contrast. Lens quality is the bigest limiting factor on 1/3 chip cameras.
>> >The Fuji & Canon lenses used bradcast camersa (1/2 & 2/3 chips) are vastly
>> >superior.
>>
>> But, but, but..., by definition, a diffraction-limited
>> lens *is* as high-resolution as it can be, from the
>> diffraction-limited aperture. Few lenses are
>> diffraction-limited at stops wider than f4 (this is the
>> only way it could be better, and it would be better only
>> at these wider stops for resolution - and, as I pointed
>> out, the VX2000 lens even at f1.6 is about as sharp
>> as the lens is at its diffraction-limited f4 stop,
>> so there is not a heck of a lot of room for resolution
>> improvement here... As for contrast, this, too, is close
>> to being maximized in the VX2000 lens. The rest is
>> myth, since it doesn't follow simple optics...
>>
>> >> >24 fps switchable
>> >> >longer tape capacity.
>> >>
>> >> You can get up to two hours now on an 80-minute
>> >> tape recorded in LP mode
>>
>> >LP looks teriible.
>>
>> It looks identical to SP-mode video in Mini-DV...
>>
>> David Ruether