BlueSky wrote:
> Hi all,
>
> From many of my pictures, using AF, I observed that the Depth of Field
> zone is always some distance "in-front" of my focus target. Example:
> A.....B......C
> Say C is the camera, A is the subject I autofocus on, the in-focus zone
> (whatever aperture) is always from A to B. Is this the way all modern AF
> cameras work?
Answer 1 (most likely to be the culprit in your case): These AF systems
typically "grab" the nearest subject they can find. Are you sure there
wasn't something closer than "A" in your composition?
Answer 2 (next most likely): The AF sensor in your camera may not be
properly aligned to the film plane. You might want to take it in under
warranty and see what they can do...
Answer 3 (possible but unlikely): From experience, a _small_ amount of
near focus bias appears to be used in some AF camera designs. I suspect
that there are three reasons why this is done:
1. At close shooting distances, there is less front DOF than
back DOF (the infamous[ly inaccurate] 1/3 - 2/3 rule), so
incorporating a small amount of near bias tends to even that
out.
2. The linear servos used for "AI Servo mode" in almost all EOS
bodies tend to lag focus a bit when tracking an approaching
subject, and tend to lead focus when tracking a receding subject.
Either way, the net effect is that the servo tends to focus
slightly behind the subject, so a small amount of near bias
helps once again.
3. There is always manufacturing variation, and based on (1) and
(2) the manufactorers have probably decided that it's better to
err on the side of near-focussing, and have offset their spec
ranges accordingly.
I've observed this bias on both an Elan 7, and on a pair of EOS 3s (one
of which is mine). My 1v does not incorporate ANY detectable AF bias; My
assumption here is that Canon figures that anybody who buys such a camera
can take care of (1) themselves, the 1v uses a nonlinear servo to deal
with (2), and given the 1v's cost there are ways to deal with (3).
Here's the catch: This bias is _very_ minimal (a fraction of the wide-
open DOF), so you shouldn't be able to directly perceive the effect under
normal examination of pictures. The only way I was able to detect it was
by examining exposures of high-detail targets slated away from the camera
(which make it easy to see where the focus is) under a microscope. If the
problem you're seeing is as obvious you suggest (and certainly if it
happens at apertures other than wide-open) then I doubt that this is the
cause.
-- Patrick
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
