The effect of aperture is very pronounced on a lens' performance. From f/2
to f/4, resolution is cut in half. So in a theoritically "perfect" lens,
with 500 lp/mm at f/2, that's 250 lp/mm @ f/4, 125 lp/mm @ f/8, and a
disappointing 62 lp/mm @ f/16. And for this we pay hundreds and sometimes
thousands of dollars!!!!!
But, there is more to the story. What actually matters is not the len's
resolution, but rather the **image's** resolution. That is, after all, the
point. The resolution of film also varies with aperture such that the
resolving power is always highest when the light source area is minimized to
limit dispersion. An interesting dichotomy - lens resolution is optimized
when it's wide open but image resolution is best fully stopped down. Once
again, we have a compromise to make. Let's see the camera makers automate
this one!
A side note, this assumes that contrast is fixed. We can also vary contrast
and realize resolution changes. A high quality film such as Fuji Velvia
will display twice the resolution at maximum contrast comaperd to very low
contrast scenes.
>From the book "Image Clarity", by John Williams, the net resolution of an
optical system, including the film itself, can be approximated. First,
convert the respective resolution values to their reciprocals to define the
permissable Airy disc spread radius in mm. Then calculate the net spread
radius (S) of the system as:
S(2)= s(2) [lens] + s(2) [film]
Where (2) is my notation convention to represent a value squared when
superscripts aren't available. To state the above expression in words, the
net spread radius of the system squared equals the spread radius squared of
the lens plus the spread radius squared of the film.
For example, if our film resolution is 120 lp/mm and the lens resolution is
180 lp/mm @ f/8 and contrast is fixed at around 30:1, then s[film] = 0.008mm
and s[lens]=0.006mm. Squaring both parts, adding them and taking the square
root of the result yeilds S=0.01mm. for a system resolution of around 100
lp/mm (99.84 to be more precise)
All of this is particularly applicable to slide shooters since the film *is*
the final image. For print shooters, there is another whole chapter which
introduces enlargement to the equation. We'll save that for another time if
there is interest.
John P
______________________________________
My Grandfather taught me to live by two rules. Rule #1: Don't tell folks
everything you know.
Gary <PCACala@xxxxxxx> pointed out:
>Hopefully he pointed out that you have to shoot at maximum aperture to
>THEORETICALLY get the maximum resolution and minimum loss due to
diffraction.
>The article in Photo Techniques on the Leica 50mm f/1.0 Noctilux shows that
as
>you stop down any lens you progressively reduce the potential resolution.
>Thus, f/22 is often an inferior aperture on our most widely used Zuikos.
To Mark Hammons' report of:
>> He gave
>> us an equation and we worked through an example with
>> a 50mm F2 lens. It turns out that the resolution limit
>> DUE TO DIFFRACTION is on the order of about
>> 500 line pairs per millimeter. Thats right -- 500!!!!
< 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 >
|