John_F Loc: Minneapolis, MN

If the front and rear filter surfaces are oerfectly parallel and flat (meaning infinite radius of curvature) and the glass is of uniform composition and its index of refraction is nil and the surfaces are perpendicular to the lens axis, then optically the filter can not alter any property of the lens. Of course, all of these factors are subject to acheivable manufacturing standards; therefore, we have the difference between quality and cheapie.

When it comes to polarization, there is the question as to the degree of source polarization being wavelength dependent which could include reflection surface variations. I'll have look this up to make sure.

On the other hand, if what you see in the viewfinder is sharp enough, then the 'out of focus' might have nothing to do with the filter itself.

imagemeister Loc: mid east Florida

I agree with this .......sadly, NO filter meets these criteria !

imagemeister Loc: mid east Florida

I will qualify this, by saying no affordable commercially produced filter ...........

rmalarz Loc: Tempe, Arizona

Not necessarily flawed logic as much as flawed interpretation. One is not going to get a superb product, such as B+W, for 29.95. (I did see a filter kit at Staples the other day for 29.95 and it included 2 filters, a UV and a polarizer).

High price does not imply good quality, but high quality is going to be priced accordingly.
--Bob

JCam Loc: MD Eastern Shore

So because some of my shots are not quite tack sharp, I should remove the daylight filter that I keep on my 18-135mm Canon kit lens for protection????

I don't use it on my 10- 300 mm Tamron zoom because the sun shield is so deep that it seems to provide adequate protection without the filter.

imagemeister Loc: mid east Florida

There is some logic that if you are using a filter - especially on 200mm or longer - that you focus cal the lens with the filter in place and then of course never remove it .....

Rich1939 Loc: Pike County Penna.

Based on comments here on the UHH I decided to test my camera/lens/CPL.
I set up and used the autofocus, then added the CPL without changing the focus and then finally refocused using the af with the filter in place.
Then the raw files were processed in ACR using the same preset for each.
To my eyes there was no discernable difference in sharpness between the first two shots but refocusing with the polarizing lens in place the third did reveal a small loss of sharpness. My camera is a D600 and the filer is a Hoya “Pro digital series”

Armadillo Loc: Ventura, CA

dugole,

This problem is something you may want to consider.

Your description and replies throughout this thread sound more like a vendor problem in supplying you with the correct polar filter. The problem reads more like a fixed polar, or a fixed Neutral Density filter. Both filters will inhibit the camera TTL auto focusing.

A variable Polar Filter (CPL) will usually have two rings and possibly two glass plates. Ring 1 and plate 1 thread onto the objective lens threaded mount, ring 2 and plate 2 rotate around ring 1. This is what gives you the circular adjustment in the polar effect.

A correct CPL filter will normally not affect a camera TTL auto-focus. The greater majority of digital cameras, when the shutter is pressed half way down have the shutter and aperture wide open to capture the most accurate light measurement available. When the shutter is pressed all the way down the shutter adjusts its speed in combination with the aperture f/stop setting.

Most quality CPL filters have a maximum of -2/stops for the exposure, this is not enough to affect capturing auto-focus, considering the aperture of the lens is wide open.

Michael G

imagemeister Loc: mid east Florida

If your lens is f5.6 wide open this means you are attempting to auto focus at f8-11 - and this is "not enough" to affect it ?? !! :shock:

amfoto1 Loc: San Jose, Calif. USA

Most filter manufacturers offer various grades of filters, different qualities at differnt prices. Hoya has some 5 or 6 grades of C-Pole. And, there are cheaper, uncoated or single-coated B+W filters in aluminum frames, too. So beyond the brand name, you also have to look at the specs and grade of any filters... and be a little careful that they are "as labelled". S

Some years ago here was an infamous CPL test, comparing about a dozen of them. The supposedly "top-of-the-line" B+W filter performed very poorly, compared to the others... turned out it was actually an incorrectly packaged Linear Polarizer, not a Circular (as required by AF cameras and even some camera's TTL metering systems). Probably that filter was a "return" that some unscrupulous buyer had substituted a cheaper filter the put back in the packaging (Amazon buyers have has seen this happen more than a few times)... or maybe it was incorrectly packaged from new. Regardless... it was not the filter indicated on the packaging and didn't perform as expected.

Personally I wouldn't worry about the 88% versus 86% difference someone else cited in an earlier post. That's not really much difference... Plus I've often seen B+W offered at lower cost than some of the top-of-the-line Hoya filters, anyway.

Some lenses simply do not work well with filters. The original Canon 100-400 (push/pull version) is one of those lenses and is widely known to "go soft" when ANY filter is fitted to it. More than a few people have been really surprised just how good and sharp their lens is, AFTER they remove the expensive clear (UV, etc.) filter they bought and installed hoping to "protect" the lens from some unknown risk. Even the highest quality clear filters cause issues, on that lens. I suspect a CPL would be even worse, with it's multiple layers of glass.

However, I have not heard similar complaint about the 100-400 Mark II (and don't use that lens myself... yet).

A CPL "cuts" between 1 and 2 stops of light... so with it in place, an f5.6 lens becomes an "effective" f8 to f11 (depending upon how the filter is set). Your 5D Mark III (assuming that's what you're using) is f8-capable... but only the five center AF points.

If you attach a 1.4X teleconverter to an f5.6 lens (making for an effective f8), the camera "knows" the TC is there and restricts itself to using only those center AF points. If you fit a 2X TC to the same lens so that it becomes an effective f11, the camera detects it and turns off AF completely (a work-around sometimes used is to tape over a couple of the electronic contacts on the TC, sort of "fooling" the camera into at least trying to focus). With a filer instead of a TC, maybe the camera doesn't "recognize" the low light condition and "know" to restrict itself to use only those "better" AF points at the center, or in worst case turn its AF off.

In other words, the CPL is a fairly strong filter, and the 100-400mm is f5.6 at the long end, so maybe the camera's AF system sensors simply isn't getting enough light to focus... or is slowing down significantly.

Some things you might try with the filter on the lens are:

1. Restrict the camera to using only the "better" center AF points. Set the camera to Single Point and manually select the AF point.

2. If using AI Servo, be sure that you have Custom Functions governing 1st image and 2nd/subsequent image priority set to "focus priority" versus "shutter release priority". This may slow shutter release slightly (how much depends upon lighting conditions, subjet movement and more), but what's the point of releasing it to take shots that aren't in focus?

3. Also experiment with One Shot and a stationary subject. Maybe set up an angled target to see if there is focus error being caused when the filter is in place. Also test Live View Focus (which uses different AF sensors on the image sensor itself).

I would not Micro Adjust a lens with the filter in place, because there are just too many times I wouldn't want a CPL on there, and MFA with it on the lens may make for focus issues whenever the filter is NOT being used. Personally I think I use a CPL about 10 or 15% of the time, at most. Even less with a telephoto like the 100-400. I have a drop-in CPL for my 300/2.8 and 500/4, but have only used it a very infrequently with them. I almost never use screw-in CPLs on my 300/4 and 70-200s either. I may use one a little more often, but still not a lot on 135mm and 85mm "portrait" teles, but still not anywhere near as often as on normal to wide lenses that I'm more often using for scenic shots.



Actually, yes. Try it without it. ANY filter costs some image quality. A really good, multi-coated filter will minimize the loss, perhaps to the point where it's undetectable in less difficult lighting conditions.

It is more likely your techniques and camera setups that are causing missed focus, instead of the filter... but it doesn't hurt to try without it and see if things improve. You'll never know, unless you try.

Besides.... What do you think some thin piece of glass will "protect" against? Most things severe enough to actually damage the lens wouldn't even be slowed down by a filter. A lens hood gives better protection when shooting.... and a lens cap is better when storing the lens. (In fact, the hood and cap are needed to protect the protection filter from damage, too!)

JohnFrim Loc: Somewhere in the Great White North.

Definitely try NOT stacking filters.

Here's another possibility that might explain an interaction with the auto-focus. CPLs are essentially a linear polarizer followed by a quarter wave plate in a sandwich. The linear polarizer is what creates the effect you are looking for, but its output (linearly polarized light) can cause problems with the metering system because of the mirrors involved, and maybe with the AF system depending how it is implemented. The quarter wave plate turns the linearly polarized light into circularly polarized light, which restores the proper metering, and maybe AF, capability in the camera.

There have been instances in which the CPL sandwich has been mounted backwards in the ring at the factory. In that case you are still just getting linearly polarized light entering the camera.

So here is how to tell if your CPL is mounted correctly in the ring.

Hold the CPL in front of your eye with the "front" pointing away from you (i.e., pretend you are the camera). Look through the filter at your image in the mirror, and look straight through the reflected filter at your eye behind it. If you can see your eye the filter is properly assembled. If you can't then flip the filter around and you should be able to see your eye in the mirror. If it's the latter case, the filter elements need to be turned around in the mounting ring.

As I said, this may not solve the AF problem which is what you are really out to do, but it doesn't hurt to check the filter assembly.

houdel Loc: Chase, Michigan USA

No, he is saying just the opposite. What he said was a cheap filter is always cheap, but an expensive filter is not always better. Case in point - Lenstip did a study on filters a while back and found that while increased price USUALLY resulted in a better product, such was not always the case. Without going into names, one well known brand "A" had significantly poorer performance than another well known brand "B" despite the fact filter "B" cost 4 times what filter "A" cost. Go to lenstip.com and look around if you want specifics.

imagemeister Loc: mid east Florida

Something to consider ! - thanks for bringing to our attention ....

RWR Loc: La Mesa, CA

You might want to re-read and correct what you wrote!

romanticf16 Loc: Commerce Twp, MI

A friend of mine shoots for the military. His German made 77mm polarizer cost over $390- I know because I priced it and that was the average from 3 US vendors. Sorry I can't remember the brand- I know I can't afford it.