On Tue, 02 May 2006 19:19:34 +0200, Vic Mortelmans
<[EMAIL PROTECTED]> wrote:
A thought that puzzled me lately may be related to this question.
It's about the focal plane and the effective distance to the camera lens.
When you focus a lens to 1m, it means that all subjects are in focus
that are in a plane positioned at 1m from the lens and perpendicular to
the lens axis (the focal plane).
At one meter from the sensor or film plane, not from 'the lens'. When
talking about a meter it might not matter too much, but when you focus a
50mm marco as close as possible it starts to count: Close focussing
distance of the D FA 50/2.8 is supposed to be 20cm. The lens itself is
60mm, the flange distance of K-mount is 45,46mm, so the object is only
10cm from the front of the lens...
If a subject is at the outer edge of the field of view and in the focal
plane, the distance between camera and subject will be more than 1m. In
case of wide angle, the difference can be significant like 1.4m e.g., if
your angle of view is 90degrees). The subject will be in focus at 1.4m
'real' distance, while the lens is set to 1m.
Now, if you want to focus your lens on this outer-edge-object, you'll
have to tilt your camera and point it to the subject (assuming you use
the split-screen-focus at the center of the viewfinder). OOPS! You will
set the lens to 1.4m! If you now tilt the camera back to it's original
position (for optimal composition), the outer-edge-object will be out of
focus, because it's in the 1m-focal plane and not in the 1.4m focal
plane!
Do you follow?
Is this a 'real' problem, or only theoretically?
Just tested this: I took a macro lens on an MX, focussed as close as
possible (that's 1:2, focus distance 0,39m) and focussed on my PC screen
using a matte gridscreen by moving the camera. Tilted to the edge (trying
to tilt around a vertical axis running though the film plane) refocussed
(to 0,395m), tilted back. The screen is now distinctly out of focus.
So, yes, focus-recompose introduces real errors. My guess would be that
ninety-five percent of the time it is not noticable. Extreme wideangle
and/or extreme close focussing could be the exceptions...
--
Regards, Lucas
