This file contains the answers to frequently asked questions about polarizing
and UV/skylight filters. I saw so many questions about these filters on news,
I thought it wouldn't hurt to make it a FAQ.
If anyone has comments on this faq or useful supplements, please E-mail them to
Richardg@morra.et.TUDelft.nl
or RAG@octopus.et.TUDelft.nl
All the information useful to other news-readers will be added to this FAQ,
mentioning the name of the one who sent it to me. You can also E-mail to the
above addresses if you still have unanswered questions about filters. Should
there still be questions that are frequently asked, I'll add them to this FAQ.
This file may be copied and distributed freely except for direct commercial
use, as long as you keep in intact.
So let me get the credit for the time I spent on it and do not just copy parts
of it, for chances are that I will get questions that are already answered in
this FAQ.
This FAQ will be posted regularly, at least once a month, with an expiring time
of approx. 2 weeks.
-------------------------------------------------------------------------------
* THINGS THAT HAVE CHANGED SINCE THE ORIGINAL FAQ *
> update september 28, 1994
- The answer to Q2 of the polarizing filters has been replaced by the FAQ on
circular polarizers written by TOM DAVIS (davis@bedlam.asd.sgi.com)
- DAVID JACOBSON (jacobson@cello.hpl.hp.com) suggested I should go for the
wave theory on light instead of the particles. I tried. He also had some
useful comments I have added to this FAQ, and he pointed out an error in A5
on the polarizers. I was wrong on the amount of light you loose, but it has
been corrected now.
- RICHARD KARASH (rkarash@world.std.com) also had some comments on the circular
polarizer, but they were already in the circular polarizer FAQ by Tom Davis.
- Via news JEFF SPIRER (jffspirer@aol.com) and Dr. GROVER LARKINS
(larkinsg@solix.fiu.edu) also commented on the polarizers, mostly about the
circular vs. linear question, and dr. Larkins commented on my advice of
overexposing film when in doubt of the light. This is corrected. The other
comments on the circular polarizer were already corrected in the circular
polarizer faq by Tom Davis.
> update september 29, 1994
- LUCA DE ALFARO (luca@cs.stanford.edu) sent in a test about determining if you
really need a circular polarizer. Jeff Spirer sent a comment on this, so I
decided to unite both of them in Q6 and A6 on polarizers.
> update october 3, 1994
- JOE CALI (Joe.Cali@anu.edu.au) sent in an explanation of the so called
"Brewsters Angle". This is added to the FAQ in Q7 and A7.
> update october 4, 1994
- Rev. Dr. PHIL HERRING (revdoc@wumpus.uow.edu.au) pointed out that even if you
use a manual camera, a linear polarizer might disturb your light meter. This
is added to A2 on polarizers
> update november 28, 1994
- A.T. Young (aty@mintaka.sdsu.edu) suggested that instead of shooting a roll
of film just to see the effects of polarizers, we should take a look at the
book "Polarized Light in Nature", by G.P. Ko"nnen (sorry, the computer won't
allow alt-148), which gives you a lot of information and shows you a lot
without having to shoot even one picture. Of course, for those who cannot
find the book (me, for example), keep trying yourself, or try to find another
book on this subject.
Also, he gave the comment that at a refractive index of 1 there's no
reflection, so 'air' should be removed from the list in Q7/A7. I did, but
keep in mind all the refractive indices are relative to air (n=1).
Last, he pointed out that the sky is not reflected light but scattered light.
This point is corrected in Q3/A3 on polarizers.
> update april 19, 1995
- I saw a thread about the advantages/disadvantages of using multiple filters at
the same time, so I added Q6/A6 to the UV filters part. I tried to use all the
comments I saw in the newsgroups, leaving the ones that were really wrong. You
can see for yourself how well I managed doing this. Because the answer is a
compilation of comments from different people, I can't give credits for it to
one person.
> update july 14, 1995
- I noticed a question about the use of polarizers and non-uniformly coloured
skies when a polarizer was used. I also read some comments on this that it
might have been caused by vignetting, when you are using a wide-angle lens.
I combined this question and comments in Q8/A8 on polarizers. Again, because
the answers came from a number of people, no single person can get the credit
for it.
* I want to thank all people who commented for their contribution to this FAQ. *
-------------------------------------------------------------------------------
- Polarizing Filters -
Q1: What does a polarizing filter do?
A1: In order to answer this question, you'll first have to know something about
light. Light can be regarded as a flow of particles (called photons), or as
an electromagnetic wave. I'll try to stick to the wave theory. So, light
can be regarded as a waveform, oscillating in an arbitrary direction
perpendicular to its direction of motion. There will be waves oscillating
up/down, oscillating left/right and all that's in between those two.
A polaroid filter only allows to pass the component of the oscillation that
is directed in the polarizing direction of the filter.
So only the waves that are oscillating in the polarizing direction can pass
the filter unattenuated. All other waves will attenuate according to the
formula:
Amplitude after polarizing = Amplitude before polarizing * cos (theta),
where theta is the angle between the oscillating direction of the wave and
the polarizing direction of the filter.
The overall effect is blocking out half of the available light, and
'directing' the other half.
Q2: What is the difference between a linear and a circular polarizer?
A2: The main points are:
(1) If you use a standard linear polarizer with cameras that use auto-focus
and/or auto-exposure, you may have problems.
This filter may also cause trouble with manual cameras, if you're using
TTL light measurement. A little dependent on the camera, your meter
might go crazy.
(It is not said you _will_ get problems, only that you might get them.)
(2) A circular polarizer can be used on all cameras, and will work the same
as a linear polarizer with respect to darkening the sky, eliminating
glare, et cetera -- circular polarizers are just more expensive.
A circular polarizer is just a linear polarizer followed by a quarter-wave
plate set at 45 degrees to the axis of polarization.
A quarter-wave plate is made of a material in which light polarized in one
particular direction travels more slowly than light polarized in the
perpendicular direction. A quarter-wave plate is just thick enough that
after passing through it, light polarized in one direction is delayed 90
degrees (or one-quarter wavelength) relative to light polarized in the
other direction.
Since the quarter-wave plate is set at 45 degrees to the polarization, you
can think of the incoming light as having two equal components in the
principal directions of the quarter-wave plate. After passing through the
plate, one component is delayed 90 degrees, and the resulting light is
circularly polarized.
The idea is to use a linear polarizer up front to get rid of some linearly
polarized light you don't want (glare off shiny surfaces, for example, will
have a large linearly polarized component), and then it "stirs up" the
result so you don't have linearly polarized light bouncing around in the
camera.
A problem with linearly polarized light in your camera, for example, is
that when you bounce it off a mirror at (near) Brewster's angle, it may be
(nearly) completely eliminated. If the light meter measures the light
after it bounces off a mirror, the amount of light arriving at the meter
may be drastically different than the amount of light that will arrive at
the film with no bounce, since the mirror has flipped out of the way.
Of course, a quarter-wave plate is only exactly a quarter wave for one
frequency of light. That frequency is usually chosen to be a yellow in
about the middle of the visible spectrum so that on the average, the light
will be circularly polarized with various degrees of elliptical
polarization mixed in. I suppose if you were photographing something that
was primarily red, or primarily violet, your metering might be slightly
off, even using a circular polarizer.
And of course, since there's another chunk of material in the way (the
quarter-wave plate), there will be slighly more degradation of the image
with a circular than with a linear polarizer.
Another nice way to think of circular polarization is to imagine a wave
travelling down a rope where you hold one end and the other end is tied to
a wall. If you shake your end back and forth along a line, the waves will
all lie in a plane. You can shake your end in any direction perpendicular
to the rope, and the only change will be in the direction of the
polarization. Now start moving your end around in a circle, and circular
waves will move down the rope. This corresponds to circular polarization.
If you move your hand in an ellipse with various eccentricities, you'll get
the equivalent of elliptical polarization (with various eccentricities).
If you're wondering whether your polarizer is circular or not, look through
your polarizer at a mirror and look at how dark the polarizer is that the
guy in the mirror is holding. Reverse the polarizer in your hand so the
other side of the glass is pointing toward the mirror. With a circular
polarizer, one direction will be significantly darker than the other. With
a linear polarizer, both sould be the same. The reason is that linearly
polarized light will still be linearly polarized in the same direction
after bouncing off the mirror. Clockwise circularly polarized will be
counter-clockwise after bouncing off a mirror, and will be cancelled when
it comes back.
So if you hold a circular polarizer as if your eye is the camera (with the
side that's normally screwed into the camera nearest your eye), it'll
appear light in the mirror. If you flip it over it should appear almost
black.
Some manufacturers (B+W and Heliopan, for example) sell a so-called
Kaesemann polarizer which is even more expensive. A Kaesmann type has the
foil stretched and held under constant tension in all directions. To do
this it is necessary to totally edge seal the filter in glass rather than
just bind the glasses and foil with an adhesive. This type of polarizer is
available in linear, circular and in warmtone types.
Its advantages are that the polarizing effect is slightly greater, the
filter is "tropicalized" so it is immune to moisture, fungus, etc and
it is very, very flat. So it will not adversely effect the sharpness of
longer lenses. For this reason Heliopan only supplies Kaesmann type
polarizers in sizes from 82mm up.
Q3: What can I use a polarizing filter for?
A3: The manufacturers will have us believe that you can block out any unwanted
reflection in glass, water etc. You can *not* block out the reflections in
metallic surfaces, since they do not polarize the light.
But, although the manufacturers are right for the greater part, you will
have to use the polarizer in the right way to get the above effect. This
means, you will have to take your picture in a direction perpendicular to
the sun (i.e. the line sun-reflecting surface has to be perpendicular to
the line camera-surface), as is illustrated below:
O sun
\
\
\ o you
\ /
\ /
\/
--------- reflecting surface
You will be able to block out the unwanted reflections this way, dependent
on the direction of the filter. When you're standing perpendicular to the
sun, the effect will be maximum, slowly decreasing as you move in line
with the reflecting surface and the sun. Then the effect will become zero.
You can also use a polarizer to control the colour of the sky, ranging from
light blue to dark blue/grey. Since the sky is scattered light and hence
polarized, you can deepen the blue by removing light scattered by dust and
molecules of, for example, water and hydrogen in the atmosphere(haze). In
this way, you are able to let the clouds almost disappear or make them
better visible. This also works best when you are standing on a line
perpendicular to the line sun-earth.
Q4: What is the best way to work with a polarizer?
A4: It depends on what you are planning to do. When taking pictures of
reflecting surfaces, it will give you the possibility to remove the
reflections, thus creating a 'better' picture than without the filter. You
can also use the polarizer to create more contrast in your pictures. The
best way to find out what you can do with a polarizer is just try it. Use a
roll of slide film (can't be corrected or ruined during printing) and take
pictures of the things you normally take pictures of, but now use the
following system: Take four or five pictures in a row of the same object,
preferably with just a short period of time between them. The first picture
should be taken without filter, just for reference. Then start with the
filter in an arbitrary position, take a picture, rotate the filter a little
(about 15 to 20 degrees ) and take the next picture, rotate the filter
again etc. until you have four or five pictures. Then move on to another
situation and repeat the above sequence. After developing the film, you
will see quite remarkable differences between the various positions of the
filter. Do try to start with the filter in the same starting position each
time you start on a new series of four or five pictures.
David Jacobson suggested you just look through the viewfinder of your
camera, but that won't give you information on what influence the polarizer
has on the autoexposure or autofocus of your camera (All cameras are equal,
but some are more equal than others - after George Orwell). Also, there are
compact cameras and TLR's with a possibility to append filters to them, and
it is not much use looking through the viewfinder in those situations,
since you're not looking through the filter. I admit, this is a situation
that will not occur very often, but I wouldn't say it's impossible.
Q5: What effect does a polarizer have on taking pictures?
A5: As said before, a polarizer can influence the colours in your picture by
darkening them, it can block out unwanted reflections and it can disturb
your AF measuring beam or autoexposure (only linear polarizers). Also,
because it will block about at least half the available light, it will slow
your film down 1.5 to 2 stops, so if you are using a separate light meter,
set your ISO dial 1.5 to 2 stops lower to correct for the loss of light.
(You can also try measuring the amount of light through the filter with
your light meter, but this is not a very accurate way of calibrating it).
In this case, just try a few pictures, you'll soon find out what correction
to use in your particular case.
In case of doubt: A little overexposure is not as bad as underexposure, so
if you want to be on the safe side, use the 2 stops correction. Warning:
this is, if you're using normal film. Slides like to be underexposed a
little if you're not sure on the amount of light.
One other comment: the 1.5 to 2 stops is *not* valid for every polarizer.
Most of them will indeed take 1.5 or 2 stops, but it may happen that your
polarizer only takes 1 stop, or takes as much as 3 to 3.5 stops. It is
totally dependent on the brand and kind of equipment you use.
Q6: How can I find out if I really have to buy a circular polarizer?
A6: It depends on the camera you're using if you need a circular polarizer or
not. Most autofocus cameras have a semi-silvered mirror, and this can
cause a significant difference in the amount of light reaching the
photocell when using a linear or circular polarizer. Most of the manual-
only cameras have their photocell in the prism, and they will not see the
difference between linear and circular polarizers. So, in case of doubt,
try a linear filter and -looking through the viewfinder- see if the reading
of the light meter changes when rotating the polarizer. If it changes more
than 1/2 stop, use a circular polarizer. This test has the most validity
if you do it in daylight, looking at a grey wall.
Q7: Why do I keep getting reflections, even if I use a polarizer?
A7: Light which reflects off any surface is polarised to some extent. The
degree of polarisation is related to the angle of incidence of light and
the refractive indices of the two materials. At a certain angle known as
"Brewsters angle", light is 100% polarised. At other angles of incidence
the light is partly polarised.
Brewsters angle is given by
Brewsters angle = arctan( n'/n)
where
n' is the refractive index of the material giving off the
reflection (eg glass, water).
n is the refractive index of the material through which the
light is incident (eg air).
Refractive indices of common materials (relative to air, n=1)
Material refractive index Brewsters angle
water n=1.333 53
glass n approx 1.5 56
(depends on the glass)
So enough theory,
All the surfaces a photographer wants to control lie in the 50 degree
range. Say you want to take a picture through a glass window. If you have
no filter on you will see a reflection. If you put on a polarizer and take
your picture looking straight through the window the reflection will still
show up. But if you move around so you are looking through the window at
an angle of 50 degrees, the reflected light will be 100% polarized. You
then rotate the polarizing filter on your camera lens until the reflected
image disappears. This is because the direction of polarization can vary
with respect to the camera depending on the angle of incidence of the
light.
Q8: When I use a polarizer and a wide-angle lens, the sky seems darker at the
edges of the picture. What is this effect and what can I do to prevent it?
A8: There are two explanations possible. If you're using a very wide angle lens
(28 mm or less), there's the possibility of vignetting because the lens can
'see' the edge of the filter. When your pictures suffer from vignetting, all
4 corners of the picture will be darker than the rest of the picture. There
is only one good solution to this: work without the filter if you do not
really need it. If you really need the filter, make sure you use a very
thin filter, or a filter so much bigger than the lens it is mounted on, that
the edges will not be visible.
The other possible explanation, when you're sure your pictures are not
suffering from vignetting (when only the 'sky-corners' are darker) is, that
when you use a wide-angle lens, the polarizing effect of the filter when
viewed from the lens is not the same everywhere, because the polarization of
the sky itself is not uniform. So, the more sky on your picture, the better
your chances are for getting a non-uniform coloured sky. The only cure for
this is using a longer lens instead of wide-angle lenses, for there's almost
nothing you can do against this 'polarization-mismatch'.
- UV/Skylight Filters -
Q1: What does an UV or skylight filter do?
A1: Both of the filters filter out the UV light that can cause a blueish haze
on your pictures, since normal film is not only sensitive to visible light,
but also to UV.
A skylight filter is also slightly coloured (pink or yellow), to give your
pictures a 'warm' appearance (not so much blue).
Q2: What is the difference between a normal UV and a skylight filter?
A2: As mentioned above, the skylight filter is slightly coloured to give your
pictures a 'warm' appearance.
Q3: In what situation do I use an UV filter?
A3: UV and skylight filters are useful when you are taking pictures in the
mountains or at sea or any other place where there is a lot of UV light.
It will filter out the blue haze that normally blurres the background of
your picture. It is also very useful when taking pictures in the snow,
since snow is a very good UV reflector.
Q4: Can I always leave the UV filter on the lens?
A4: About half of all the photographers keep an UV filter on their lens perma-
nently, for it prevents your lens against dust, scratches and perhaps
damage due to accidentally dropping the lens. A filter is much cheaper than
a lens, so ruining your filter will not be as bad as ruining the front part
of your lens.
On the other hand, some photograpers (the other half) think it unneccesary
to keep the filter on the lens, since everything between the original
picture and your film, including filters, can cause blurrs or errors in the
image, and that's one of the things we don't want to happen.
It is just a matter of personal preference. I must admit I always have a
filter on my lens, and I take it off only for cleaning.
Q5: What effect does an UV or skylight filter have on taking pictures?
A5: First of all, it will filter out most of the UV light. A skylight filter
will also colour your pictures a little. Because the UV is filtered out,
you may notice the sky in your pictures is not as deep blue as it used to
be. This is because the UV component is now missing, resulting in another
kind of blue on your pictures. Just try a few pictures with and without
filter and see what you like best.
An UV or Skylight filter doesn't have any effect on the amount of visible
light falling through the lens, so you won't have to correct for it.
Q6: Do I have to remove my UV filter when I want to use another filter, such as a
polarizer?
A6: It is not necessary to remove your UV filter when you're going to use any
other filter you may have, but it won't do any harm. Some filters, like
polarizers, do cause image degradation because of their construction. This
degradation is usually more than the degradation alreadycaused by the UV
filter, so in those cases you do not really have to remove the UV filter.
On the other hand, you run the risk of light being 'trapped' between the two
filters, thus causing lighter spots on your picture.
When you're using filters of very high optical quality, it's always a good idea
to remove all other filters. In general, the less filters on your lens, the
better the quality of the image.
Another thing to remember: multiple filters on your lens may cause vignetting,
especially when you're using a wide-angle lens. So decide for yourself whether
the saving of time by not removing your UV filter before mounting another is
worth the trouble you may experience from all kinds of nasty effects on your
pictures.