Polarizers or not - Test Images

[Geoff F]

Following on from Brian's post where the use of polarizers was discussed, I have done some tests and here are my findings.

I had stopped using polarizers some time ago after similar tests which produced the same results. Shot with a 7D and Canon 24-105 or Sigma 180 macro lenses. On a tripod with aperture priority at F 11 and Iso 400. Manual focus. Lens Pro polarizers which aren't the very best but seem to get reasonable reviews for those middle of the road filters which most people would find affordable.

Straight from camera shots except for a little bit of unsharp mask after resizing for internet use. Evaluative Metering.

Using a polarizer which took quite a bit of the available shutter speed. I couldn't get a really sharp focus. Shutter Speed 1/100



The same scene without the polarizer but with one stop of negative exposure compensation.Shutter speed 1/640.



With a polarizer and Sigma 180 macro lens. Sharp focus was impossible to achieve even with manual focusing. Shutter speed 1/100



The same scene without a polarizer but with one stop of negative exposure compensation. Shutter speed 1/800



Some hand held landscape scenes where the difference isn't so great. With a polarizer. Shutter speed 1/500



And without a polarizer but with a little bit of exposure compensation. Shutter speed 1/1250



Based on these results, I will not be using polarizers in the future.

[Manfred M]

Geoff - you have clearly demonstrated why I DO use a polarizer when I do these types of shots.

I prefer to cut the glare and reflections. I generally have a tripod along, so getting sharp focus with low shutter speeds are not an issue.

[Stagecoach]

Geoff, my first question would be, why as these were taken as tests did you overexpose the example where the polariser was used? For a comparison test of the 'effect' of using a polariser on these subjects should your aim have been to simply expose them to get an equal result and not one image clearly brighter than the other?

My second question would be, why did you have problems getting focus just because a polariser was used? A polariser will take 1.5/2 stops at worse but is your macro lens not f/2.8?

[John 2]

Not so sure about your findings Geoff. The polarizer cuts reflections and by definition, mostly acts on reflected highlights. If the with and without polarizer images have been correctly exposed, I would expect the shadows to be about the same. However, the polarized images seem to have lighter shadows which makes me suspect that they are somewhat overexposed. 1/100th sec might have been a bit low. I might also have expected a decent polarizer to cut the glare a bit more than that but of course I can't be sure.

[AlwaysOnAuto]

These are my best examples of a polarizer in action:





YMMV, as they say.

[NorthernFocus]

All I can say is that you must be doing something wrong, Geoff. When shooting foliage I find a polarizer will completely remove the glare on leaves. And not sure why you're having focus problems. Are you using a circular or linear polarizer. AF won't work with a linear polarizer. And if AF doesn't work neither does the focus indication when focusing manually.

[Cogito]

Geoff, what are your images meant to "prove"? All reflected light is in a single plane - that's O level physics. Your inability to focus has nothing to do with using a polariser unless as Dan says it's not a circular polariser... tho' I'm not entirely sure of that.
Polarisers allow light in a single plane to be unaffected and so will cut out reflections that are not in the "required" plane. Hence turning a circular polariser affects what you see in the viewfinder. It depends how much reflection you inhibit. It shouldn't affect you ability to focus and as Manfred states they do allow you to reduce glare.

[Manfred M]

unless as Dan says it's not a circular polariser... tho' I'm not entirely sure of that.

In a DSLR, the main mirror is a pellicle mirror, commonly referred to as a semi-silvered mirror, which has a hinged mirror behind it that reflects the light onto the autofocus module. The semi-silver is created by etching very fine parallel lines onto the mirror, i.e. a polarizing grate. A linear polarizer will variably impact the amount of light passing through the pellicle mirror, but a circular polarizer will not. This is why we need to use a circular polarizer when shooting with a DSLR.

If a beam splitter is used on the metering element, then a linear polarizer will affect the camera's light meter as well; giving an incorrect light reading. With the metering issues that Geoff describes, I suspect this might be why he has to dial in exposure compensation.

[Stagecoach]

With the metering issues that Geoff describes, I suspect this might be why he has to dial in exposure compensation.

Manfred, if you take an image without a polariser and manually expose using matrix metering to centre bar position then use a circular polariser and again manually expose to the centre bar the image with the polariser will be brighter overall than the first. I suspect this is because the matrix metering with the polarizer is seeing less 'bright areas' to use in its averaging calculation.

Geoff was using negative exposure comp on shots without the polariser which I think was simply a way of him attempting to reduce highlights.

[Stagecoach]

For comparison with Geoff's findings I did a quick 'controlled test' to show that a polariser is beneficial in certain conditions especially with respect to some foliage. Note, I'm not saying it is necessary or the answer for all occasions. The plant is a Sansevieria which has waxy leaves.

No 1 - Without Polariser


No 2 - With Polariser


Note, the predominant light was coming from the right and the leaves are all at different angles to the camera. There is a 2EV difference between exposures and in PP I adjusted one image by just 0.2EV to equalise the shade tone.

Also note the leaf at the right and how it has assisted in the lower edge definition.

Conclusion, make your own

[Rent]

Geoff F. I think you will find that the angle that the light falls on the subject is critical as well, and I can't see why you should have a problem focussing
Roy

[billtils]

Geoff, I've never encountered any of the problems that you describe, and have a circular polarising filter fitted as default on a couple of my lenses.

One feature that has not been mentioned in this thread is their ability to intensify blue skies, as in this shot of Puffins on the Isle of May:

[Manfred M]

Manfred, if you take an image without a polariser and manually expose using matrix metering to centre bar position then use a circular polariser and again manually expose to the centre bar the image with the polariser will be brighter overall than the first. I suspect this is because the matrix metering with the polarizer is seeing less 'bright areas' to use in its averaging calculation.

Geoff was using negative exposure comp on shots without the polariser which I think was simply a way of him attempting to reduce highlights.

Could be. I found that the metering system generally compensates well for this without having to resort to exposure compensation. As I shoot to the histogram, I will fix issues like highlights and shadows in PP, I only resort to exposure compensation when I see clipping or blocked shadows.

The main reason I mention the metering too is that my old Leica R3 required a circular polarizer because a beam splitter was used to direct light onto the metering cell, so that approach has been around in cameras since at least the 1970s.

Like Bill, for certain types of photography, I leave polarizers on my lenses when I head out to shoot as a default and only remove them when the exposures get ridiculously long as the light levels drop.

[pendennis]

Like others have mentioned, they're invaluable for removing glare from certain images. I learned a long time ago, though, that they're very specialized in their usage, and actually a negative value when using wide angle lenses. BILLTILS image of the puffins is a great example of separating the subject from a blue sky.

ALWAYSONAUTO also shows a great example of glare elimination, especially when glass is involved. BTW- Beautiful Porsche!

I like the subtlety of the image improvement that STAGECOACH demonstrates with his images. At first glance there's not a great deal of difference, but the polarizer makes critical differences in capturing the image.

I've never had problems with focusing when using a polarizer, even with more primitive autofocus film systems. I have, however, learned that they're not a panacea.

PS - I wasn't aware of the Leica R3 problems, but it's a good bit of information. Thanks, MANFRED M!

[Manfred M]

I learned a long time ago, though, that they're very specialized in their usage, and actually a negative value when using wide angle lenses.

I assume you are referring to "banding" that results from having too much sky in a wide angle shot. That is only true if there is a broad swath of sky in the shot. No sky, no problem. A narrow view of the sky; for instance shooting in vertical / portrait orientation, or having something obstructing part of the sky, also no problem.

A polarizer is a tool like many other tools we use in photography; we have to understand how to use the tool and what its strengths and weaknesses are. For example, in a shot like Bill's I will tend to back off a bit with a shot like that in order that the sky does not darken too much.

[pendennis]

I assume you are referring to "banding" that results from having too much sky in a wide angle shot. That is only true if there is a broad swath of sky in the shot. No sky, no problem. A narrow view of the sky; for instance shooting in vertical / portrait orientation, or having something obstructing part of the sky, also no problem.

A polarizer is a tool like many other tools we use in photography; we have to understand how to use the tool and what its strengths and weaknesses are. For example, in a shot like Bill's I will tend to back off a bit with a shot like that in order that the sky does not darken too much.

Correct. Sky with clouds is often used to denote the expansiveness of a scene, and I've seen any number of images in 50 years that have misused the polarizer in an attempt to dramatize the effect.

As you mention, it's another tool in the bag, to be used with discretion and propriety as needed.

[Geoff F]

Firstly, it was circular polarizers which was set to the darkest effect in all cases.

Secondly, the first shots with a polarizer were not deliberately over exposed. The camera was set to Evaluative Metering and that is how the shots came out when using a polarizer and no exposure compensation. I was using the 24-105 lens because I had a polarizer of that size. It definitely isn't one of my favourite lenses and I rarely use it now.

Can a polarizer change the minimum focusing distance for a close up shot? The landscape scenes do not show any difference in sharpness between using a polarizer and not using one. The closer shots were manual focus while the landscapes were AF.

I haven't tried moving back a little to see if that makes any difference for closer work.

[Stagecoach]

Could be. I found that the metering system generally compensates well for this without having to resort to exposure compensation. As I shoot to the histogram, I will fix issues like highlights and shadows in PP, I only resort to exposure compensation when I see clipping or blocked shadows.

Manfred, for info I agree that the matrix metering with a polariser compensates well and was not suggesting that any compensation was necessary due specifically to its use, other than what we would normally do when taking any shot.[/QUOTE]

[Stagecoach]

Can a polarizer change the minimum focusing distance for a close up shot? The landscape scenes do not show any difference in sharpness between using a polarizer and not using one. The closer shots were manual focus while the landscapes were AF.

From the tests I did, which could be considered 'a close up shot' it appears not, but, it may be that the DoF disguises things. I did not alter focus between the shots, just screwed the polariser on and adjusted it very carefully.

[Geoff F]

That is exactly what I did, but I was shooting from a distance of around 18 inches.

With these various examples, I wonder what sort of light was involved, particularly the car shots which seem rather 'flat' to me. It seems like it was a rather diffused somewhat dull light while I was shooting in full sun.