Filter advice needed

Discussion in 'Digital SLR' started by Arild P., Jul 9, 2006.

  1. Sure, but I'd rather get it right the first time.
    And now for something completely ungrounded in science, and based
    solely on gut feeling:

    I find that my exposure tends to be "better" when my circular polarizer
    is on the camera. No, I have not done any tests on this. Nor have I
    ever read or heard anyone else say the same thing. I could offer a
    couple of wild-guess theories as to why this might be the case, or as
    to why I might feel this is the case, but since they're complete
    speculation, I won't inflict them upon you. But I do know that, with my
    C-POL on the camera, I tend to be happier with the exposure (assuming
    I'm letting the camera meter the scene, and I'm not using exposure
    compensation).

    I've got no experience on how well (or poorly) a linear polarizer works
    on an automatic camera. I've heard that, in addition to metering
    problems, they can cause problems for autofocus systems (especially as
    light levels drop). Again, I've never used one, so I can't back that
    up.

    - Darryl
     
    madhobbit.geo, Jul 13, 2006
    #21
    1. Advertisements

  2. Arild P.

    AaronW Guest

    Get the strongest you can find, and it still might not be enough,
    that's why people stack multiple filters to make it stronger.

    If it is only 1 stop, you might as well not use it, but digitally
    brighten your dark scenes 1 stop.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 13, 2006
    #22
    1. Advertisements

  3. Arild P.

    babalooixnay Guest

    Check the tolerance specs on your camera, plus or minus one third of a
    stop for meters is pretty typical. If your meter was reading one third
    too bright the polarizer might bring it right down to zero. My meter
    was one third too dark and I had dialed in a compensation factor. When
    I added a new, brighter focusing screen I got lucky and ended up with
    zero also. Just depends what side of the tolerance you're on.
     
    babalooixnay, Jul 13, 2006
    #23
  4. Arild P.

    AaronW Guest

    It is probably because that the polarizer filtered out the strong
    reflections that would throw off the meter. In this respect, linear
    polarizer would perform as well.

    I did not say that linear polarizer results in a more accurate
    exposure. Depending on the filter angle, sometimes linear gets the same
    exposure as circular, sometimes it is up to 1 stop off. I often adjust
    exposure after the shot, and the meter is not perfect anyway, and I
    might even want a bracket shot at 1 stop off, so it is not a problem
    for me.

    Because of the additional film in the circular polarizer, it can
    degrade picture, flare, color, etc. And because of the interaction of
    the circular polarized light and the mirror/prism of the viewfinder,
    the color you see in the viewfinder may not be the same as the sensor
    gets when the mirror gets out of the way. I saw this problem of
    circular polarizer and switched to linear polarizer.

    If you seldom adjust exposure or do bracketing, if you value
    convenience over slight difference of picture quality, you may prefer
    circular polarizer.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 13, 2006
    #24
  5. Arild P.

    AaronW Guest

    Just get the cheapest multicoated one you can find. If they do not
    cheat you, hopefully they don't waste multicoating on junk.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 14, 2006
    #25
  6. Arild P.

    AaronW Guest

    If you use the tele lens for animals and sports, the f/5.6 may be
    already too dim. You probably don't want to block additional 2 stops of
    light with a filter.

    One alternative is 200/2.8. It is similarly priced. It is up to 2 stops
    brighter, which is significant. And the image quality is better.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 14, 2006
    #26
  7. Arild P.

    AaronW Guest

    Larger filter may be more expensive, so you might not want to spend
    extra on something you do not need. If you do not care about the money,
    maybe you can spend it on a better smaller filter that you can enjoy
    now.

    Besides, a larger graduated filter does not work as well on a smaller
    lens. The density of the filter changes from top to bottem. The density
    close to center is not as strong as the edge. So if only the central
    part of the filter is used, (which is already true for full frame lens
    on sub frame camera), the effect will be weaker than the filter is
    designed to be.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 14, 2006
    #27
  8. Arild P.

    Arild P. Guest

    The wide angle lens I currently have (I assume my Sigma 17-70mm is
    considered having a "wide angle" even though it's a zoom lens) is the
    one with the biggest lens diameter (if I buy the Canon 70-300 IS USM),
    so for the Sigma I'll probably be buying filters that fit directly on
    that lens. Also, the comment in this thread about bigger filters being
    more expensive is something to keep in mind.

    But if I buy a circular polarizing and a graduated neutral-density
    filter with a 72mm diameter (fitting directly on the Sigma 17-70 lens),
    will I still have trouble with vignetting?

    I'm also considering getting a "UV filter" for both lenses. From what
    I've heard the "UV" function is minimal, although they do remove some
    of the "haze" from high altitudes such as an aircraft (turning the hazy
    blue sky into a clear blue sky?). But the main reason for a filter like
    that would be to protect the lens against dirt and damage. I believe a
    protective/UV filter like that has threading in the front as well, so
    that an additional filter (circular polarizing or neutral ND filters in
    my case) can be inserted into that protective filter. Can this cause
    vignetting with my Sigma lens set at the widest angle?
     
    Arild P., Jul 16, 2006
    #28
  9. Arild P.

    Arild P. Guest

    Can you please explain (in layman's terms) what the practical
    differences between a circular and a linear polarizing filter is?
    All I want is to be able to shoot through windows (e.g. remove any
    reflection) and as an added bonus, get that "tropical" effect when
    shooting beaches.
     
    Arild P., Jul 16, 2006
    #29
  10. Arild P.

    Jeremy Nixon Guest

    The practical difference is that, if you use a digital SLR, you must use
    a circular polarizer. A linear one will interfere with the autofocus.

    There is no difference whatsoever in the effect of the filter.
     
    Jeremy Nixon, Jul 16, 2006
    #30
  11. Arild P.

    Arild P. Guest

    So why is AaronW advocating the linear polarizer filter when this is a
    *digital* SLR group and you can't/shouldn't use those, but only a
    *circular* polarizer as you say?
    Perhaps AaronW can reply to this.
     
    Arild P., Jul 16, 2006
    #31
  12. Arild P.

    J. Clarke Guest

    Because this is USENET and if someone says it's a bad idea to put your hand
    in a cage of rabid, starving wildcats someone else will expound at length
    on why he's full of it.

    If you google "autofocus polarizer" you'll find quite a lot of discussion.
    Bottom line is that if you use a meter and manual focus then you can use a
    linear polarizer but any advantage in image quality from doing so will be
    very small.
     
    J. Clarke, Jul 16, 2006
    #32
  13. Arild P.

    Greg \_\ Guest

    I don't know about other peoples experiences or cameras but I have found
    the linear PL I have actually works with my D70 and 18-70mm lens in
    manual mode but with AF on. I was surprised given all the hype and given
    the price of a CP its at least worth a look at the local camera store to
    find out whether one's camera will function properly.
     
    Greg \_\, Jul 16, 2006
    #33
  14. On the simplest level: A linear polarizer consists of a single element:
    a polarizing screen. A circular polarizer has a second element -
    something called a quarter-wave plate. It sits between the polarizing
    screen and the lens.

    The polarizing screen is what gives you the visual effects: blocked
    reflections, darker sky, etc. It's the same in both types of
    polarizers. The difference between a linear and circular polarizer is
    quarter-wave plate exists solely to make sure that your autofocus and
    auto-metering systems work properly.

    The (simplified) physics behind it: All light waves have an
    orientation. A particular light wave coming towards you might be
    oriented up/down, or left/right, or diagonally, or whatever. The
    polarizing screen can be thought of as a bunch of very narrow parallel
    slits, that only let through light waves that have a particular
    orientation. (This is not quite true, but it's Good Enough (tm) for
    now).

    In a room lit by, say, an incandescent light bulb, the light waves are
    oriented in all sorts of directions, so a polarizer blocks some, and
    lets others through, evenly across the whole scene. The effect:
    Everything gets a little darker. In this case, there's not much
    difference between a polarizer and a non-graduated neutral density
    filter.

    However, in an outdoor scene, the light coming from the sky is strongly
    polarized: the light waves tend to be all oriented the same way. By
    rotating your polarizer to be perpendicular to them, you can block a
    lot of them out. This is why the sky gets darker: you're blocking the
    light coming from the sky. This is also why you can block reflections:
    when skylight is reflected off of water or glass, it keeps its
    polarization. The polarizer, if rotated to the right angle, can block
    out these reflections.

    All this comes with a side effect - after going through the polarizer,
    the light is polarized. This means that all the light waves are aligned
    the same way, in nice neat rows. This has no effect on the actual image
    captrued by film or digital sensors. But it can have an effect on the
    TTL metering and autofocus systems in your camera. When the light comes
    through your lens, it goes through a beam splitter that sends it in two
    different directions: some of it goes to the viewfinder, so that you
    can see what you're pointing the camera at, and some of it goes to the
    autofocus and metering sensors, so that the camera can figure out what
    to do. Some beam splitters don't work well with polarized light: they
    may not send enough light over to the sensors (in which case your scene
    will end up overexposed, and your autofocus may not work)...or they may
    send too much light to the sensors (in which case your scene will be
    underexposed).

    So, the circular polarizing filter was invented. The quarter-wave plate
    "de-polarizes" the light...it scrambles it all up into random
    orientations, so that the beam splitter works properly. It's this extra
    element that makes circular polarizers cost more.

    So...do you need a circular polarizer? Well, I'm no expert on beam
    splitters or camera design. It looks like some of them will work fine
    with polarized light, others won't. Maybe there are additional factors
    that I'm not aware of. It is generally advised that, if you use TTL
    metering or autofocus, get a circular polarizer. You could try a linear
    polarizer first, and see if it works with your camera, but it's not
    always easy to tell if the metering is wrong.

    - Darryl
     
    madhobbit.geo, Jul 16, 2006
    #34
  15. Arild P.

    AaronW Guest

    It rotates the light a certain angle. Different colors get rotated
    different amounts.

    You can try this yourself. Stack 2 polarizers together. If the front is
    a linear polarizer, when you rotate the rear polarizer so that they are
    90 degrees to each other, they block out light, which is to be
    expected.

    If the front is a circular polarizer, when you rotate the rear
    polarizer, light will not be totally blocked out. But each color (a
    narrow spectrum) will be filtered out. As you rotate the rear
    polarizer, the scene will dramatically change color to warmer and
    colder. This effect is caused only by the rear polarizer on the front
    circular polarizer, not by the front polarizer on the natural light,
    since the front is not rotating. If you rotate the front instead, what
    you see will be a combination of the polarizing the natural light and
    the interaction of the circular polarizer with the rear polarizer. You
    may not be able to tell which effect is caused by which.

    If the prism in the viewfinder is partially polarizing, it will behave
    the same as the above situation, only to a lesser degree. When you
    rotate the circular polarizer and when the effect is not very strong,
    you may not be sure whether the color effect you see is caused by the
    front polarizer thus will be captured, or whether it is just the
    interaction of the circular polarizer and the viewfinder.

    If the prism is not polarizing at all, then the front polarizer will
    not affect the viewfinder. But it will not affect the exposure meter
    either, so you do not need a circular polarizer in the first place.

    Only if the prism is partially polarizing the meter, but not polarizing
    the viewfinder, then a circular polarizer is all win and no lose. I
    don't know if this is possible.

    BTW, a circular polarizer will not cure the meter problem in all
    situations. It works best if the scene is greyish, a somewhat even mix
    of all colors. If the scene is mostly a single color, e.g., green,
    since the green lights get rotated a similar angle, it would function
    similar to a linear polarizer, that the meter can be off a little bit.
    If this situation is acceptable for circular polarizer, then a linear
    polarizer will be acceptable, too.
    The difference is not large that I never have an exposure disaster
    caused by linear polarizer. And I never have any AF problem caused by
    linear polarizer.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 17, 2006
    #35
  16. Arild P.

    AaronW Guest

    It is equivalent to 27mm, so not ultra wide, as 24, 20, 14, etc.
    If it does, you can always remove the UV filter.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 17, 2006
    #36
  17. Arild P.

    Arild P. Guest

    Several useful comments here!

    So in my case, does this sound like a good solution (quality, function
    and costwise) for my setup?:

    For Sigma 17-70mm f/2-8-4.5 lens (72mm diameter):
    - 72mm Hoya Ultra neutral filter
    (http://www.hoyafilter.com/products/hoya/gf-03.html)
    - 72mm Hoya Super-HMC PL-Cir (circular polarizing) filter
    (http://www.hoyafilter.com/products/hoya/gf-04.html)
    - 72mm Hoya graduated neutral-density (ND) filter

    For Canon 70-300mm f/4-5.6 IS USM lens (58mm diameter):
    - 58mm Hoya Ultra neutral filter
    (http://www.hoyafilter.com/products/hoya/gf-03.html)
    - 58mm-72mm (Hoya?) step-up ring (for use with 72mm polarizing filter)
    - 58mm Hoya graduated neutral-density (ND) filter

    Unfortunately I haven't quite figured out which of the filters at
    Hoya's website (http://www.hoyafilter.com/products/hoya/index.html) are
    graduated ND filters.
    Is it this one: Half NDx4
    (http://www.hoyafilter.com/products/hoya/oef-08.html)?
    if the illustration is anything to judge by, it looks like the border
    is pretty "hard", while I would assume a "soft border" would more
    likely be more versatile for all sorts of situations with high contrast
    differences.
    I had a look at one of the links given here, going to the "graduated
    filter" page, and there are indeed two different kinds; a soft and a
    hard transition: http://www.2filter.com/prices/htpackages.html

    ..... but not Hoya. Does Hoya have soft transition ND filters?
    The above site suggests a strength of 0.6 ND which means 2 stops.


    As for the UV filters ( ) as previously discussed, I found out that
    Hoya has a "neutral" filter, which they claim that "it will not affect
    the color balance or performance of your lenses in the slightest", and
    since UV is apparently removed from lens-coatings (when I don't use any
    filter at all) and filter coatings (when I use a circular-polarizing or
    graduated ND filter) I might as well get one of these and probably save
    some money.
     
    Arild P., Jul 17, 2006
    #37
  18. Arild P.

    AaronW Guest

    B+W's multicoated linear polarizer is priced similar to Hoya's
    multicoated circular polarizer. You can also check whether Heliopan
    makes one.
    You can get the cheapest step up ring.
    If you can find a good 3-stop, it would be useful.
    Coatings are for minimize reflection, not removing UV.

    It seems that SHMC is Hoya's best coating, and it is available on their
    UV filters, not neutral filters.

    And get one filter for one lens first, to see if you like it, before
    buying filters for other lenses.

    http://digitcamera.tripod.com/#slr
     
    AaronW, Jul 18, 2006
    #38
  19. Hell- I never use filters, if i want to degrade the image I just stick
    my hand in front of the lens. Provided your a good enough salesman you
    can probably still get a buyer for the image.
     
    Woo U Flung Poh, Jul 18, 2006
    #39
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.