At 14:46 10/23/00 , Chris O'Neill wrote:
>On 23 Oct 2000, at 13:10, atk@xxxxxxxxxx wrote:
>
>Hmmm... I didn't think that which one to use was camera dependent?
>So, please allow me to modify or expand your question...
>
>1. What's the difference between a linear and circular polarizer?
>
>2. Why (when?) would you use one over the other?
>
>3. Is one better than the other? If so, which one?
>
>Chris
>(Who uses a Cokin P160 linear polarizer)
A polarizer tutorial:
Most light is randomly polarized. However, light reflecting from a smooth,
shiny, non-conductive surface _will_ be linearly polarized if observed at
"Brewer's Angle" which is dependent on the index of refraction for the
reflecting material. As you start to deviate from this angle, the amount
of linearly polarized glare will decrease. Far enough and it goes to zero.
By using a linear polarizer, which passes only light linearly polarized in
a particular direction, and aligning its polarization at 90 degrees to the
glare polarization, you block it out. No practical linear polarizer is a
perfect one, so there is a "null" range around the 90 degree point where
you can control how much gets blocked. This is why, even when rotated out
of the "null" you still lose light through one; in fact it's roughly half
the light, a full stop, perhaps a little more.
So far, so good, but here's the rub. Modern cameras, including the
OM-2S/sp, OM-3[ti], and OM-4[T,ti] have TTL light sensors behind the
mirror. How does the light get through the mirror to them? The mirrors
are half-silvered, a form of linear polarizer, and it works based on the
assumption that the light arriving at the mirror will be randomly polarized
in all directions. Some passes through to the sensor in a particular
polarization, and the rest gets reflected to the prism. That's why the
OM-2S/sp, OM-3[ti] and OM-4[T,ti] have slightly dimmer viewfinders, and
users of them often put in a "2 series" screen.
Most of the time a plain linear polarizer won't be much of a problem for
you metering. However, if perchance the glare you are blocking out just
happens to be the same or 90 degrees to the linear polarization as the
light passing through the half-silvered mirror to the metering sensors,
then you have a problem. In the first case, you will end up overexposing
as not enough light is getting through the mirror because you blocked it.
In the second case, you will end up underexposing, because the light you
blocked is only light reflected to the prism, and none of it would have
passed through to the TTL sensor.
How does the circular polarizer fix this problem? A linear polarizer has
approximately the same effect on circularly polarized light as it does on
randomly polarized light. If, after using a linear to block the glare you
can circularly polarize what's left before it hits the mirror, then the
risk of over or under exposure goes away. A "quarter-wave retarder" does
exactly that; it circularly polarizes light. So, if you put one of these
behind the linear, then the problem is fixed. This is exactly what a
circular polarizing filter is: a linear polarizer with a 1/4-wave retarder
behind it.
There are some diagrams showing how this works on these two sites:
http://www.doehrer-elektrooptik.de/datenblaetter/ip/iplinpol.htm
http://www.opticalfilters.co.uk/products/standardfilters/polarizers.htm
Which should you use? In those cameras that have half-silvered
beam-splitter mirrors, you should use a circular. In all others with
full-silvered mirrors, you can use either. One is not inherently better
than the other, except purists will tell you to get that extra piece of
unnecessary glass out of the light path to the film if you don't need it,
such as with an OM-1[n] or OM-2[n].
-- John
< 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 >
|