At 9:23 PM +0000 5/28/02, olympus-digest wrote:
>Date: Tue, 28 May 2002 00:27:05 -0400
>From: "Jim L'Hommedieu" <lamadoo@xxxxxxxx>
>Subject: Re: [OM] Help! Light theory, cp-l
>
>I thought the pattern in a circular polarizer is circular, that is,
>concentric circles, sharing a common center. So the inner most circle has a
>very small diameter, and each successive circle, surrounding its
>predecessor, is slightly bigger. Rotating such a circular polarizer has no
>meaning, unless the centers don't match the center of the lens axis.
The name is a bit confusing, but that isn't how it works. If one breaks a
circular polarizer in pieces, one gets some little circular polarizers of equal
effect, just as with linear polarizers. I'll try to explain it:
Linear polarizers require the electric field vector of the lightwave to vibrate
all in one plane, conventionally called vertical (up and down) or horizontal
(side to side). This convention is a bit confusing: one converts from vertical
to horizontal by physically rotating the polarizer, which is itself unchanged.
Nor does vertically polarized light differ from horizontally polarized light
except that the plane in which the electric vector vibrates is rotated. When
unpolarized light falls on an ideal linear polarizer, half the light is
absorbed, and half passes through as linearly polarized light. (Practical
linear polarizers absorb more than half and pass less than half.) If this
polarized light encounters a second linear polarizer, all the light passes if
the second polarizer is the same as the first (v-v or h-h), and no light passes
if the second polarizer is the other kind (v-h or h-v).
Metallic mirrors have no effect on linearly polarized light other than a change
in direction. However, half-silvered mirrors act as linear polarizers, so if
one has a linear polarizer on the camera lens, more and less light will either
be reflected or pass through as one rotates the polarizing filter for the best
visual effect, thus undermining the lightmeter function of cameras with
half-silvered mirrors.
Circular polarizers require the electric field to spin in a circle, with the
electric vector tracing out a helix (corkscrew) as the lightwave progresses.
Circular polarizers come in two kinds (senses), left-handed and right-handed,
and pairs of circular polarizers work as described above for pairs of linear
polarizers. (Elliptical polarization, a mixture of linear and circular, is
also possible, but not an issue in ordinary cameras.)
The key is what happens when one combines a circular polarizer (on the lens)
with a linear polarizer (the half-silvered mirror): The same proportion of
light is reflected or passed as for unpolarized light, so the lightmeter
function is no longer undermined. It doesn't matter which way the corkscrew
turns or how the linear polarizer in oriented, the same amount of light gets
through.
>I would guess by extension that combining two circular polarizers of
>identical spacing between the rings would, in theory, overlap perfectly,
>resulting in a darker image. In reality, I imagine that the centers of BOTH
>polarizers would be slightly askew from the center axis, resulting in "hot
>spots" in light arcs.
See above. There won't be a hotspot.
>Considering the horrific cost of circular polarizers, it seems like the most
>expensive possible way to get neutral density. (I bought my 55mm CPL in a
>shop instead of mail order and it was $40 USD. Ouch! Worst of all, I've
>opened it.)
Yeah. I never bought one, because I used the linear polarizers so little.
Only used it for crossed-polarizer reduction of surface reflections. One can
actually use circular polarizers for this, with one rotation sense on the
camera and the other rotation sense on the lights, but it would be real
expensive. The OM-1 with linear polarizers is just what the doctor ordered.
Joe Gwinn
< 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 >
|