26 May
2007
26 May
'07
2:56 p.m.
On Sat, 26 May 2007, Brent Hilpert wrote:
My main concern for taking close-up (oblique, not just
plan view) photos of
equipment is depth-of-focus. I know/figure it's mainly a function of aperture
size (smaller --> longer depth-of-focus) and (optical) zoom (to minimize the
ratio of focus-depth to lens-to-object distance). (And smaller aperture means
less light which requires longer exposure.)
If I have the above principles about right, the practical question(s) are:
Is your typical $200 / 3*-optical-zoom / 5-MPixel camera good for this or
does one have to go to something higher end with a special lens, etc.?
It's subjective as to what's adequate of course, but I'd be interested in
opinions.
Is there a spec regarding the aperture range that I should be looking for?
If you're not picky, then almost anything should give images adequate for
www. If you are picky, there are some simple formulas to use.
If you are including any photos that show the screen, use a long exposure
to get multiple retraces. (see the article that Dr. Marty and I did for
Coco magazine)
my apologies if the following is too much, or not enough, technical info:
The aperture is stated as in f units, which is the ratio of the
[effective] focal length of the lens to the [effective] diameter.
It is logarithmic - to halve the exposure, multiply the f number by
sqrt(2).
[effective] is because there are optical tricks to make lenses have focal
lengths and diameters other than the actual measured distance (often
needed to make telephotos smaller, or to be able to have a wide angle
whose focal length won't physically fit the camera).
Depth of focus is a function of focal length (shorter has more D.O.F.),
aperture (smaller has more D.O.F.), and how much blur you are willing to
tolerate.
A = (L * B * F) / ((L * F) + C * (B - F))
Z = (L * B * F) / ((L * F) - C * (B - F))
where A is nearest distance "in focus"
Z = farthest point in focus
L = [effective] diameter (aperture) of lens
B = distance nominally focused for
F = [effective] focal length
f = diameter/aperture expressed as ratio F/D
C = circle of confusion (if there is a point, just how big a blur would
you call "in focus")
c = angular size of circle of confusion
the same thing could also be expressed as:
A = (B * B * tan(c)) / (L + (B * tan(c)))
Z = (B * B * tan(c)) / (L - (B * tan(c)))
NOTE: The variable names that I used are old traditional names for them,
including F and f and other potentially confusing choices.
FORTRAN and WATFOR are capitalized, because they are acronyms, not
for emphasis nor shouting.
US f numbers are typically 1.4, 2, 2.8, 4, 5.6, 8, 11, 16, 22, 32, 45, 64,
90, 128; for a while, some european companies used 4.5, 6.3, 9, 13, etc.
Both the number of calibration points, and the endmost numbers can vary.
And the focal distance calibrations will vary in range, choices of
intermediate values, and units of measure.
Slightly more on-topic:
Because of those variations, the first FORTRAN program that I did that
used "runtime calculated/specified formatting" was to printout depth of
focus tables for my lenses. I used WATFOR at George Washington
University, about 1970.
--
Grumpy Ol' Fred cisin at xenosoft.com