On Thu, 01 Oct 1998 10:58:35 -0400, Viken Karaguesian wrote:

> Hi guys. Maybe I'm just a little confused about DOF. Say I want to
>photograph a subject thats 200 feet away. If I open up my lens to f/2.8
>or 3.5, will that subject be in focus even though the DOF for that small
>a stop is very limited? Ex. My family has a summer cottage, and there's
>a house across the lake (about 200-300 feet) that I want to photograph.
>Knowing that f/2.8 has a DOF of only a few feet, will the house be in
>focus in the picture (I'm focusing on my subject ) even though it's far
>out of DOF range? Or will my DOF be within a few feet of my SUBJECT, no
>matter how far the subject is?
> Also, does nighttime photography follow the same "rules" of DOF
>(that may be a silly question, but one that I must ask)? For instance,
>that same example of the house across the lake: That house has a bright
>spotlight that at night, casts and interesting shadow over the house.
>Many times (especially around this time of year) fog will be coming off
>the lake after dark and the spotlight makes some interesting patterns in
>the fog. I'm assuming that I would want to use a large apereture, since
>it's dark and I need the maximum amount of light, but the DOF is so
>shallow at 2.8, that I wonder if the house will come out during a time
>exposure. Or should I use a small aperture since my subject is so far
>away.
> So I guess this long winded question boils down to one question: is
>the DOF aound my subject, or does the DOF start from me? A 20 foot DOF
>is up to 20 feet from me or 20 feet around my subject, no matter the
>distance of my subject. Thanks for any replies.

Lets start all over......;-)

If you replace your lens with a pinhole of the ideal size for the
"focal length" you are using, and imagine the subject as made up
of an infinite number of infinitesimal points ("subject points")
reflecting straight light rays in all directions, including toward
your pinhole (which will pass only the ones that are not obstructed
by something in the way [the edge of the pinhole material, or
something "out there"...]), you can see that each subject point
can be represented by a cone of light rays (with the subject point
at its tip and the diameter of the cone at the pinhole being equal
to the diameter of the pinhole itself) intersecting the film plane
in a small circle of light (called a "circle of confusion", oddly
enough...;-). (I could add that light rays are bent at the very
edge of the pinhole, slightly softening the edges of the circle
on the film plane [called "diffraction"], but let's ignore that for
now - though this determines the ideal size of the pinhole for best
sharpness [and of the lens diaphragm chosen later on for best
sharpness...;-) Subject points are thus rendered as image circles
on the film plane of the pinhole camera, and the "lens" is both
very slow (to maintain some sharpness), and not especially sharp
at best.

If I add a lens of the right focal length, it can accept a larger
diameter cone of light rays from each subject point, and bend
those rays into a reversed cone, with a point at its peak that
just intersects the film plane (assuming a really good lens...;-).
Holding all this in mind, imagine moving either the film plane or
lens forward or back relative to the other. You can see that the
lens cone is either intersected ahead of the point (resulting in
a circle), or behind it (resulting in a circle from intersecting
the again-diverging rays). The point of correct focus is made
a circle instead of a point, and the image made from all the
circles together will be less sharp than from all the points
together.

Now, hold in mind the image of the de-focused image (with
circles representing each subject point, instead of each
subject point being represented on the film plane by a point).
If you change nothing but the diameter of the lens (or a
shadow-casting variable hole ["diaphragm"] in the lens...), you can
see that the size of each image circle (the image circle rather
than point results from poor focus...) changes. Smaller hole
equals smaller image circles - and larger, larger; smaller hole
equals sharper-looking image - and larger, softer. Since the
subject may not all be at one distance from the lens, you
can see that for a given set of conditions, parts of the
subject with varying depth may actually be in correct focus
(rendered as points on the film plane), and some may not
(rendered as various sized circles on the film plane), resulting
in varying sharpness depending on focus. So you can see that the
appearance of the depth of subject sharpness rendering in the image
("Depth of Field") changes both with focus, and the size of the
diaphragm that lets in light.

To go further, "DOF" includes both the sharply focused and rendered
subject parts, and parts that are not sharply focused and rendered,
but which are "good enough" to look sharp. The acceptable
circle of confusion size can be specified for a given set of conditions,
and that can be the basis for setting up "DOF tables" for showing what
depth of field in the subject should look sharp with a given FL lens at a
given focus distance and aperture size, on a given format, for a given
image size (film format...) or angle of view. Added to the mix, though,
is visual perception - which often makes DOF tables not quite correct
(especially when very distant subjects are included, or if the lens is
especially sharp...). And lens performance, which is often far from
optimum at wide stops (though the DOF may be OK), and which declines
at small stops due to diffraction, should also be considered.

So, getting back to the house...
The greater the focus distance, the greater the range of apparent DOF
(so a wide stop is less of a problem at 200' than at 2' for getting good
DOF with a non-flat subject). The flatter the subject (the house), the
wider the aperture can be while still giving good DOF (assuming the
surface photographed is nearly parallel with the film plane...). But good
results depend also on the image quality of the lens at f2.8, and the
depth of the illuminated fog... A help is understanding DOF, but
experimentation is good, too...;-)