Polarising filter with UV filter?

Discussion in 'Digital Photography' started by Stimp, Nov 15, 2006.

  1. Stimp

    Stimp Guest

    I recently bought a UV filter for the 17-85mm lens I bought, and I'm
    considering whether or not to buy a polarising filter too.

    When using a polarising filter is it normal to remove the UV filter
    first, or do you mix the two?

    When would you generally use a polarising filter? Would you leave the
    filter on all the time, or only in specific lighting?

    Also what features should I make sure the filter has before I buy it?
    (e.g. circular polarisers are apparently better etc)
    Stimp, Nov 15, 2006
    1. Advertisements

  2. Stimp

    Ken Lucke Guest

    My opinion (and that of my 28-year veteran mentor): Don't. Just use
    the polarizer - even when it's not doing a lot of polarizing, it still
    functions just as well to "cut haze" which, besides "protecting the
    lens" [which, BBTW, a polarizer does just as well] is the main thing
    that people get sucked into buying a UV filter for.
    Yes. The more glass you stack on the lens, no matter how good it is,
    does degrade your image.
    Mine is there at all times, unless I really need the 2/3 stop of extra
    light that it cuts out. I like to shoot underexposed by one stop
    normally anyway [so I have more exposure play room in post and better
    saturation], so it about evens things out, shutter-speed-wise.

    You'd be surprised how much glare it cuts out on just standard shots
    (like wet leaves, damping down highlights to a manageable level on
    rippled water, etc., even on a cloudy day reflections off of interior
    objects, the list goes on and on...), and how much it brings up the
    saturation of color in many shots.

    Circular is not "better" than linear. [Actually, IMO, linear is
    considerably better than circular in side-by-side comparisons I have
    made]. Circulars have to polarize the light, then re-orient it again
    before passing it on - linears don't have that extra step. They were
    designed differently because auto-focus film cameras use a different
    method of sharing the image between the focus/viewfinder/meter than
    non-autofocus cameras. There's all sorts of techincal information on
    this issue if you Google for it.

    Most digital cameras can use either linear or circular. Circular is
    what you are likely to find more of, because most recent [auto-focus]
    film cameras needed them.

    DON'T skimp and buy a cheapie. It's going to be one of the MOST
    important additions you can put on your camera/lens. Buy one
    commensureate with the best glass you are going to be using it on -
    it's high crime to use, say a Canon L series lens, then stick a cheapie
    polarizer out in front of it. If you are going to be using several
    sizes of lenses, buy the Cokin system and adapter rings, that way you
    only have to have one polarizer to be shared between them.
    Ken Lucke, Nov 15, 2006
    1. Advertisements

  3. Stimp

    Ken Lucke Guest

    When I said "Don't", I mean't "Don't buy just the UV", of course, as
    you should have been able to determine from the rest of my soap box
    exposition. Following up on myself here just to clarify, because some
    usenet-nitpicky-ninny will no doubt point out that that one word was
    contradictory to the rest of the post.
    Ken Lucke, Nov 15, 2006
  4. Stimp

    acl Guest

    It also cuts two stops of light, though. If you don't find this
    important, I suppose you live in a very sunny place and never shoot
    except in bright light or with a tripod.
    2/3 stop? Are you sure about this? I mean, did you check? What kind of
    polariser is this?
    Well, if he has an SLR, he'll probably have problems with AF
    (beamsplitters are used for this and a linear polariser will mess it up).

    And circular polarisers don't "re-orient" the light before passing it
    on, but since you dislike nitpicking, forget I ever wrote that!
    Maybe cameras with contrast detection AF will work, I don't know. But
    all SLRs I know of use phase detection AF and this uses mirrors; you'll
    need a circular polariser for them. But maybe some have no problem with
    linear polarisers, I can't say for sure.

    Perhaps it would be best if the OP tested AF and metering with a linear
    polariser should he decide to get one instead of circular.
    That's also good advice.
    acl, Nov 15, 2006
  5. Stimp

    acl Guest

    OK, it's impossible for it to only cut 2/3 of a stop. It will cut 1 if
    it's a linear polariser with perfect transmission for fully polarised light.

    A perfect linear polariser would transmit a fraction cos(theta)^2 of
    light which is incident on it and polarised at an angle theta to the
    angle which it passes. So for unpolarised light it'll be the integral of
    this from 0 to 2 pi, divided by 2 pi, ie 1/2. This is one stop. And
    circular polarisers will cut more.

    I suppose a linear polariser could also pass light polarised at other
    angles, and this would cut less light.
    acl, Nov 15, 2006
  6. You hit the right point when you said "a perfect linear polariser" will
    absorb one stop - real life ones also have an absorption factor, which
    takes the attenuation to 1.5 to 2 stops.

    The quarter wave plate which circular polarisers have after the
    polariser absorb no significant amount of light, so the difference in
    absorption between the two is negligible.

    All this assumes the incident light is unpolarised (or, to be
    nit-picking, randomly polarised). If it is somewhat polarised, then it
    is true that a polariser may be oriented to selectively pass is, and
    thus have a lower effective absorption factor. However, the usual need
    is to orient the polariser to *reduce* the polarised light component
    (which usually arises from reflections from non-metallic surfaces), so
    in use the factor will normally be increased.

    David Littlewood, Nov 15, 2006
  7. In addition to all the other good points made about the redundancy of
    using both, you need to beware of vignetting. If you put two filters on
    a wide angle lens, there is a serious risk that the edges of the picture
    will be darkened by light cut-off. This can be desirable under some
    circumstances if you choose to do it, but to have it built in to every
    picture is a bad idea.

    As for leaving a polariser on the lens all the time, I would advise
    against doing this if it means using too slow a shutter speed, or too
    wide an aperture. 1.5+ stops is quite a lot to loose unless you need the
    polarising effect.

    David Littlewood, Nov 15, 2006
  8. Stimp

    Paul Mitchum Guest

    If the filter is on your lens, and you have through-the-lens metering,
    then you're not underexposing anything, it seems to me.
    Paul Mitchum, Nov 15, 2006
  9. UV filter on a digital camera has essentially no effect on the image
    other than the addition of some flair and reduced sharpness. With a good UV
    filter the losses will be minimal and seldom noticed. It may under some
    conditions help protect the lens. It certainly will improve the bottom line
    of the retailer.

    Many years ago I was a photo retailer and I did advise against UV
    filters for lens protection. Even then I sold quite a few of them because
    the customers had been told by someone that they needed them. I can
    guarantee you that I sold far more dollars of those "protective" filters
    than repairs or lens replacements due to the kind of damage such a filter
    might provide. In short, it is expensive insurance.
    Joseph Meehan, Nov 16, 2006
  10. Stimp

    Jim Guest

    Exactly once in more than 60 years of photography has a UV filter actually
    saved me money.
    However, the filter did save me a lot of trouble because I was in
    Yellowstone National Park at the time.
    I dropped my F3 with 100-300 zoom attached to the ground, and the only thing
    that was damaged was the filter.
    Without the filter, the impact may have damaged the lens. As it happens,
    there are lots and lots of camera shops in Jackson Hole, but that town was
    more than 2 hours away.

    So, I was lucky to have ruined only that filter.

    Also, you don't need to filter the UV light when using a polarizer because
    the polarizer filters UV.
    Jim, Nov 16, 2006
  11. z
    Oh, just stack as many pieces of glass in front of the lens as you can.
    Randall Ainsworth, Nov 16, 2006
  12. Stimp

    acl Guest

    Now that you mention it, introducing a phase difference shouldn't
    necessarily cost anything in terms of transmission. I had assumed it
    does. So, do you know why there is extra loss (most polarisers I've
    seen lose closer to two stops)?

    acl, Nov 16, 2006
  13. Stimp

    Phil Wheeler Guest

    Only issue may be vignetting.

    Phil Wheeler, Nov 16, 2006
  14. Stimp

    Ken Lucke Guest

    Tripod. 99.999% of the time. I think I also said "unless I really
    need the 2/3 [sic] stops"
    My fat fingering: I was writing in a hurry. Let that read 2-3 instead
    of 2/3. Mea culpa.
    _Most_ digitals don't use beamsplitting AF, AFAIK. He was talking
    about a digital. I've been able to use a linear on every digital I've
    ever tried. Googling the subject [as you made me doubt my experiences
    as semi-unique] seems to find almost universal agreement to that
    Uhm, a circular polarizer nothing more than a linear polarizer with the
    addition of a followingy 1/4 wave plate - I'd call a 1/4 wave plate
    "reorienting", but then I guess I'm funny that way.
    See above.
    Ken Lucke, Nov 16, 2006
  15. Stimp

    Ken Lucke Guest

    Uhm... I think I said "...evens things out, _shutter-speed-wise_".
    Ken Lucke, Nov 16, 2006
  16. Stimp

    Ken Lucke Guest

    As I just noted in another article, I meant "2-3" not 2/3. I was in a
    hurry and didn't proofread my own stuff like I normally do. My
    apologies for starting this furor.
    Yeah, well. OK. You guys can worry about the math. I'll just go take
    pictures. =:^)
    Singh-Ray has a circular polarizer they call "LB" (Lighter and
    Brighter) that is rated at only 1 1/3 stops. Only trouble is (in my
    view), they also combine it with a built-in warming filter, and I don't
    necessarily want that. I suppose I could compensate in post, but I'd
    rather not.
    Ken Lucke, Nov 16, 2006
  17. The polarising material does not have a perfect transmission of light
    whose polarisation is aligned with its maximum transmission axis. Thus
    it absorbs something like 99% of the light transverse to this axis and
    say 25-50% of the light aligned with it. The better quality the filter,
    the lower this additional (unwanted) absorption is likely to be.

    In addition, there is the factor I mentioned before - one deliberately
    aligns the filter to block polarised light from non-specular reflections
    (which is polarised. IOW, the incoming light is partially polarised, and
    one seeks to preferentially block it.

    David Littlewood, Nov 16, 2006
  18. Stimp

    acl Guest

    Well, a digital for which he bought a 17-85mm lens. So I assumed an
    SLR, all of which use phase detection AF, as far as I know. Which may
    or may not work with linear polarisers (I never tried, but would expect
    them not to; but maybe I am wrong).
    No, you're wrong, not funny. It introduces a phase difference between
    two directions. Sorry to pick nits.

    Anyway, not much point in going on and on about this.
    acl, Nov 16, 2006
  19. Stimp

    acl Guest

    OK I see. I thought they could be made perfect.
    acl, Nov 16, 2006
    John McWilliams, Nov 16, 2006
    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.