Re: Circular polarizer question


"Kennedy McEwen" <rkm@xxxxxxxxxxxxxxxxxx> wrote in message
news:JWBssNJcfYhGFwK3@xxxxxxxxxxxxxxxxxxxxxxx
In article <1183140099.254974.19120@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
lphilpot <len@xxxxxxxxxxx> writes
I have a brand new Tiffen 72mm circular polarizer I purchased with my
new Rebel XT. I've not yet shot any images through it, but simply
looking through it outdoors under a partly (cumulus) cloudy but
otherwise blue sky, it appears to have virtually no apparent
polarizing effect. I have two other standard non-circular polarizers
(a 55mm Bower and a 49mm Rolev) and they both make a significant
difference in the sky, in removing reflections from a car window,
etc., etc. as expected. However, a view through the TIffen and either
of the other two will go nearly black as expected when properly
rotated. That tells me the TIffen is indeed polarizing as it should
(maybe?).

This is the first circular polarizer I've owned and Tiffen is a
reasonable brand. Is this normal for a circular polarizer, or is there
something wrong with it?

As Charles said, with an LP is doesn't matter which way you look through
it, you'll see the polarising effect, but with a CP you need to view
through it the way it will be mounted on the camera lens.

The reason for this is that the CP is a sandwich of a normal LP with
something called a quarter wave plate on the back of it. The QWP has
the effect of rotating the angle of polarisation through 360deg as each
wavelength of light passes through it. The high speed and the short
wavelength of light means the polarisation angle coming out of the QWP
spins around an amazing 550,000,000,000,000 times a second. Even
unpolarised, or more accurately randomly polarised light gets its angle
of polarisation spun at this speed.

If you view through the CP filter the wrong way, then the light from the
scene gets its polarisation angle spun by the QWP before it reaches the
LP and then, obviously, half the time it is passed and half the time it
isn't. More importantly, it doesn't matter what the original
polarisation angle of the light is, 50% of it gets through the LP, and
the change is much too fast for your eye, or the sensor in your camera's
AF or exposure meter to detect.

That is why you can't see any polarisation differences through the CP if
you look through it the wrong way - with one interesting and
occasionally useful exception...

View your own reflection in a mirror through the CP. If you look
through it the same way as it mounts on the lens, the way you expect to
see polarisation effects, then it looks just the same as an LP. If you
look through it the wrong way round then the filter itself always
appears completely black, irrespective of which angle you rotate it to.

You could say that the CP is the ultimate one way mirror - you can see
the reflection through it in one direction but it is completely black in
the other direction. ;-)

This is because the light coming from in front of the filter, eg.
reflected from your face or eye, is first linearly polarised by the LP
section. It then passes through the QWP which sets the polarisation
angle spinning at that enormous speed. However, the QWP can cause the
spin to be clockwise or anti-clockwise, depending on its thickness. So
that light now proceeds to the mirror, where it is reflected, which
changes the direction of rotation (actually it keeps the direction of
spin but reverses its direction of travel and it is the relation between
these to that matters). When this reflection returns to the QWP, the
rotation is cancelled out again however, because its direction of
rotation has been reversed, the angle of polarisation is exactly 90 deg
relative to polarisation on its outward trip. Consequently the LP
section of the CP blocks that light and so the image of the filter in
the mirror is always black.

Since the QWP should give a perfect 90deg rotation of the polarisation
angle of the reflected light and the LP should completely block that
polarisation, then how black the image appears is actually a very good
test of the quality of the CP filter itself. With a small flashlight
and a mirror, you can easily test various CPs and pick the best one.

One good photo application of this effect is to eliminate specular
reflections from light sources, especially flash. The common solution
is to place a polariser on the light source which is at 90deg to the
polariser on the lens, thus blocking direct reflections. However,
slight misalignments of the two polarisers let some of the direct
reflection pass. With identical "reversed" CPs, it doesn't matter what
the relative alignment of the two CPs are, direct reflections are always
blocked by the maximum possible amount.


Yes, Kenny, I agree - - thanks for the excellent explanation of the workings
of the C-PL filter.

I thought I'd go whole hog and buy the Targus 77mm C-PL Filter for $17.99 at
Fry's (they really get expensive at this size) for my Canon DSLR. It darkens
the sky just fine when used on the camera, but it completely fails the
mirror test. My understanding is when in front of the mirror with the filter
threads toward the mirror, viewing through the filter, I should not be able
to see through the reflection of the filter - - it should appear black. It
doesn't - - its quite transparent. If I tilt it, maybe 45 degrees and rotate
it in its sliding ring, it cycles through transparent to magenta to
transparent to cyan, but without tilting it, there is no change of color - -
always transparent. Viewing my computer LCD screen through the 77mm filter
with threads toward my eye, the extinction gets quite deep, but turning the
filter around, the screen takes on colors ranging from lemon yellow to a
medium purple. I suspect the QWP isn't very accurate. My question - - is
this level of performance good enough? If it is, why should anyone shell out
$80-$120 for a proper performing filter?

With this C-PL filter in place I took a few pictures with my 30D and an EF-S
17-55 f/2.8 IS USM lens and all were properly exposed with darkened sky and
well focused by my camera. So it looks like it might work okay.

I have another, smaller (48mm) PL-CIR filter by HOYA and it behaves more as
you describe - - with filter threads toward mirror, viewing through the
filter, the filter reflection is black, and tilting it and rotating it in
its sliding ring does not change that. Turning it around with threads toward
my eye, the filter is transparent gray with no changes in transmission or
color as it is tilted and rotated in its ring. Viewing my computer LCD
screen with threads toward my eye, the extinction gets very deep, but
turning the filter around the colors never go deeper than from light tan to
light blue as I rotate it in its ring.. I think I paid a little more for
this filter.

Thanks for your attention,

Chuck





.

Relevant Pages

  • Re: Diode Laser, Quarter Wave Plate and a Linear Polariser
    ... You mean that the reflection off say a lens must change the ... polarisation of incident light? ... mirror along the way is incidental/irrelevant. ...
    (sci.optics)
  • Re: Verschraenkte Gluehbirne
    ... Traeger der alten Theorie nicht ueberzeugt werden, ... auf meine Frage, wie mit seiner Vorstellung von Polarisation ... Nach dem Durchgang durch den 45° Filter ist dadurch der Zustand des ... Alltagserfahrung - und das akzeptieren beide nicht. ...
    (de.sci.physik)
  • Re: UV/Polarised filters
    ... Polaroid filters remove polarized light. ... would accomplish is to remove reflections from the dust itself. ... You need to rotate the filter ... to suite the polarisation of the reflection you are trying to remove. ...
    (alt.photography)
  • Re: Verschraenkte Gluehbirne
    ... polarisiertes Photon durch das Filter kommt (mit der ... Polarisationsfilter die Polarisation des ... Ein zweiter gleich orientierter Pol-Filter 100% der Photonen ... senkrecht orientierten Polfilter von 0% der Photonen ...
    (de.sci.physik)
  • Re: Bras, kets etc.
    ... circular polarisation is the more fundamental of the two, ... Look at the box of beetles through a circular polarisation filter, ... you can't determine if the incident photon was circularly ...
    (sci.physics)