Using Polarisation Filters

Using Polarisation Filters

Page Type Page Type: Article

Introduction

About two years ago I stumbled into a discussion about photography on one of the SP message boards where several people recommended using polarisation filters in order to enhance the quality of the resulting image. In 2003 I started using polarizers myself and since then I rarely venture out without one. During that time I was able to shoot a number of photos which without polarizer would have been utterly unimpressive while I also did a lot of mistakes since I had to find out about the use of the filters on my own.

During that time I have also seen photos down voted because the voter thought them photoshopped (“At that elevation you don’t have a sky that blue!”) or - the other extreme – heard a member claim that the polarizer reveals the real colours while our own eyes see only a grey-shaded version of reality.

Since – as always – the truth lies somewhere in between, I thought it might be helpful to publicly muse about my own experiences as an amateur photographer while also discussing the physical principle of the filters. Don’t be afraid that it gets too technical – but also don’t look for the actual truth. That would probably involve nasty physical formula and after ten years out of the physics business I don’t claim to understand them myself anymore.

Why Use Polarizers Anyway?

So here we are and the first question generally is:

What does a polarizer do?

The answer – as anybody who has used them will tell you

They enhance colours, making the picture crisp and clear – under certain conditions.

In order not to bore you with words, let’s have a look at the following two pairs of pictures.
Große and Kleine...Gamswiesenspitzen without polarizer
Große and Kleine...Gamswiesenspitzen with polarizer

In the first pair you see the pair of Große and Kleine Gamswiesenspitze as seen from Zochenpass in the Lienz Dolomites in September 2003. I shot the first picture as part of a multi picture panorama, for which I generally don’t use filters (you’ll find the reason in the drawbacks section below). However I liked the view good enough to also shoot a second shot with polarizers in place – the shot to the right.

The effect is quite apparent – a good but somehow dull picture suddenly turns into a colourful one, the clouds in the background move much closer so that, as hgrapid put it “The clouds above seem like a sea below.” The structure of the rock is visible and it seems like a grey veil has been lifted from the picture.
Rocca Busambra main summit   unpolarised Rocca Busambra without polarizer
Rocca Busambra main summit   polarised Rocca Busambra with polarizer

The same effect is even more pointed in the second pair of shots, while the object itself might not be that impressive. While hopping along the ridge of Rocca Busambra in May 2006 I turned round after having downclimbed from the main summit to the next saddle. I first wanted to take a vertical shot, adjusted the polarizer accordingly, but then decided to settle for the horizontal alignment. While pressing the shutter I realized that now the polarizer was off and redid the shot with correct settings.

Again, the picture on the right is much more colourful and if not put side by side you would probably not imagine that these pictures show the same crest and were taken just seconds after each other. The lifted-veil-character is not as pronounced here as in the other pair, which is caused by the short distance in the second and the far distance in the first pair.

Now one last example:


Colàc (far right) in front of the Sella-, Geisler/Odle and Saslonch/Langkofel/Sassolungo Groups.


Here you see the western Dolomites as seen from the Buffaure ridge on a perfect day with decorative clouds giving just the right amount of shadow to the foreground. Also colours are intense here and there is no hint of a grey veil here. You can even see the faraway mountains quite clearly like the Geisler / Odle Group in the back centre of the panorama.

Working Principle

So – how is it done? Is it magic?

Certainly not, but it is a physical effect, which is much harder to explain than to master the use of the filters themselves. Therefore there is a short explanation below and a physical one. Let’s start with

The Short Explanation


When rotated to the correct angle, polarization filters reduce stray light by reducing reflections. Have a look at the pair below
Demo Polarisation FilterDemo without polarizer
Demonstration Polarisation FilterDemo with polarizer

To the left on the unfiltered image you can clearly see the reflection on the table which is missing almost completely on the filtered one to the right. The same happens when you take a shot in plain air. While it might appear that the air between your camera and the object you are taking a picture of is completely transparent, there are millions of tiny particles, aerosols. These aerosols might be simple water vapour or smog and dust particles – but all reflect additional light into your camera lens – an effect which you see as a grey veil on your photos.

The polarizer cuts away this additional light – and the grey veil is gone. But how?

The Physical Explanation (Unscientific)


(No mention will be made of circular polarisation though this is an important effect when using polarizers – you have to be able to think and speak in mathematical formulae to get that right ;-))

We have to start with the question: What is polarisation in the first place?
Most of you will know that you can describe light as propagating in waves. When discussing colours for instance we often speak of wavelengths or frequencies – red light has a long wavelength, blue light a short. Now have a look at the light wave below:
What is Polarisation?Light Polarisation

If you pick out one special point on the wave (like the one indicated in green) and if you look at this point from the right hand side you will see it bounce up and down, following the movement of the wave. It would always keep moving in that direction since the wave itself moves up and down along the same pane. In geek speak this wave is linearly polarized, vertically in this example.

However, the light emitted from any light source consists of gazillions of lightwaves, all of which are emitted randomly in all directions. Moreover the panes in which the waves are swinging are also distributed along all angles. When viewed from one side you would find the points of the waves swinging in all directions, every single one bouncing up and down along one line but the multitude of them distributed along all angles. The image b) in the above picture tries to give you an idea.

Now you put the polarisation filter in place and the effect is immediate: only one pane is selected while all the light swinging in another pane is suppressed. As a consequence the polarisation filter greatly reduces light intensity – you’ll have to use greater apertures or longer exposure times – but we’ll be getting back to this in the “Risks” section. If you now choose to use a second polarisation filter you will find that – as you rotate it – you’ll be able to suppress all the light. Both filters have to be at 90° angles between each other.

But how does that help in reducing stray light...


...or in lifting the grey veil?

To answer this question you have to know that every particle, even the tiniest one, while it reflects light changes the polarisation of it. Exceptions are metals which bounce back light as it hits them but our aerosols behave differently. Depending on the angle of the reflection the reflected light is more or less polarised with nil effect at 0° and 180° (light from front or back) and maximum effect near 90°.

In real life this means that if you have the sun at your back or in front – though there is still stray light, the stray light won’t be polarised. Using a polarizer won’t help a bit. However, if you have the sun to one of your sides, rotate the polarizer and you’ll get to a point where the veil has lifted – colours get much more intense (and darker) while details suddenly come to the foreground. Turning the filter will give you the impression that parts of the frame come closer or retreat farther to the back.
Polfilter Working PrinciplePolarisation Filter Working Principle

Now that we have (hopefully) understood the mechanism, let’s have a final look and try to see what really happens: the object (mountain) emits non-polarised light in full colours. On the way to your camera stray polarised light gets added so that everything gets hidden behind our veil. With the polarizer you now choose exactly one polarisation – the one which diminishes the stray light. In geek speak that is the direction ,,tilted at 90° to the polarisation pane of the stray light”. In the end you’ll get a similar picture as the one that was emitted by the object in the first place.

Polarizer vs. Photoshopping and Other Filters

Photoshopping


We have all heard that one of the great advantages of digital photography is the possibility to edit and enhance pictures afterwards. Can I achieve the same results with photoshop?

The short answer is no – at least not with the standard routines. An automatic level or curve adjustment is a move in the right direction but don’t be fooled: it takes an expert photoshopper to obtain results that mimic the effects of a polarizer. And I doubt that anyone would be able to adjust one of the pictures on top so that it resembles its counterpart on the right. You may try and convince me…

UV-Filter


A different kind of filter – the UV or UV-haze filter promises similar results as polarizers. By stripping away light at lower wavelengths (i.e. below a wavelength of 380nm = UV light) they claim to get rid of stray light without the cumbersome aspect of having to rotate the filter to obtain the best effects. However the stray light is not primarily emitted at these wavelengths – only very high up in the mountains you will obtain visible results. If you plan to climb above 4000 or 5000m a UV-filter might be all that you need.

Limitations

There are a number of conditions in which a polarizer is useless or gives unexpected results. I wouldn’t say these are drawbacks but you have to be aware of them.

Back light / Front Light


As that in the theory section the polarisation of the stray light is virtually nil at 0° and 180°. You won’t get results with polarizers under these lighting conditions. However – if the sun is very high up in the sky – polarizers can still help because now there is an angle of almost 90° again between the sun and the direction you are looking.

Portraits and Macro Shots


Close-up shots are not affected by polarizers. There is simply not enough stray light between the object / subject and the camera. However, if the background (like blue sky) is of importance or if there is an reflecting object nearby (like spectacles), you might still enhance the shot. Be aware that the filter reduced light intensity!

Flash Light Photos


Don’t do this with a polarizer! The automatic settings between flash and camera won’t recognise the filter and underexpose the shot severely.

Telephoto Pictures


The opposite is true for tele photo pictures. Here the effect can be seen though you might want to adjust the filter while looking at a wide angle an zoom in later. Depending on the amount of haze between you and the faraway object you might have trouble seeing the effect directly.
Western Sicilian HighlandsRocca Ramusa on Sicily

Haze and Smog


Too much haze or smog reduces the polarizer results greatly. Now you have multiple reflections on the aerosols and the polarisation of the stray light gets blurry. The hazier it is the less impressive the results. Polarizers are useless in fog!

As for smog – the same holds true as for simple haze. However, the smog particles are “dirty” and thus change the colour of the stray light. As a result faraway shots with polarizer can look brownish.


Overcast Skies


There is absolutely no use in a polarisation filter when the sky is overcast. All the light – direct and stray – will be polarized in all directions so that a filter won’t do any good at all.

Risks

Blurring


I has been mentioned somewhere above that polarisation filters greatly reduce the light which falls on the chip or film. In everyday application this means that you either have to open the aperture or – if that is already open all the way – you have to use longer exposure times. Especially if you are using the automatic exposure mode of your camera you might not even be aware of it. Even on a bright day a telephoto shot can get blurred if the exposure time is longer than the reverse of the focal length (f=200mm -> t<1/200s). With a polarizer which cuts off additional light you might easily fall below that threshold.
VignettingIsola Alicudi in Sicily

Vignetting


There is another effect which – though it occurs with any filter – is most pronounced with polarisation filters: vignetting. Since chip as well as film are rectangular while the whole optics is circular you might observe circular patterns in the corners of your photos. The effect is pronounced with wide angle lenses together with polarizers. As the opening angle of wide angle lenses are wide they might get obstructed by the filter, which is placed in front of the lense. Imagine it as casting a shadow into the lense. Polarizers are especially thick so that the effect is largest with these kinds of filters.

However, Brenta told me that there are special thin-mount filters which are more expensive and lack the thread in front. With these kind of filters you may be able to eliminate the vignetting effect. Due to the lacking front thread stacking of filters is not possible.

Sky GradationSky gradation in Germany's Soonwald

Sky Gradation


If you plan to use wide angle lenses together with a polarisation filter you might observe that the colour of the sky changes brightness over the width of your photo. Since polarisation is different at different angles the stray light on the left side of a wide angle photo is differently polarised than the one on the right hand. Have a look at the shot to the right and see the colour change from a very deep blue on one side (one you would never achieve at an elevation of 600m without a filter) to much lighter blue on the other. The shot was taken with a 18mm lense with an opening angle of 74°.


Panoramas


Don’t use polarizers for panorama shots! Certainly not for panoramas covering an angle of more than 90°. Since the stray light polarisation differs at different angles and since the polarizer picks out only one polarisation you are bound to get visible “steps” in your panoramas. Even if you leave aperture and exposure settings fixed the “color aspect” of the pol filtered shot changes over the width of the pano, thus switching from blueish (directly against the sun) to yellowish (at 90°). With the individual shots at different of these hues you will get “colour steps” in the pano. With patience these can be treated individually in photoshop but it is a very cumbersome process.

Conclusion

Gamskofel (Mooskofel hidden...Gamskofel in the Carnic Alps

If you can spare the 30 – 50 dollars (or Euros) that a polarisation filter costs and if you are interested in good photos – by all means buy one and start experimenting with it. No number of words can prepare you to what a polarizer can do and you should start observing its effect by simply looking through and turning it. As you do so you will see the sky turn a magnificent blue, the clouds turn a dazzling bright white and objects in the background seem to move closer to the front. The filter does not have to fit to the lense mount – if it is wider you won’t run the risk of vignetting. Be aware of limitations and risks and the filter will give your photos a new life.

External links

Similar article with additional info

Comments

Post a Comment
Viewing: 21-40 of 41
ExploreABitMore

ExploreABitMore - Jul 18, 2006 2:52 pm - Hasn't voted

Thanks

Thanks for posting this ... very helpful.

Klenke

Klenke - Jul 18, 2006 7:31 pm - Hasn't voted

Do our eyes auto-polarize?

Nice page, Gangolf.

Question: do our eyes have natural built-in polarizers? If no, then aren't our naked eyes also subject to the "gray-veiled" scene as non-polarized photographs? That is to say, polarizing a particular shot allows us to see a scene how we'd like to see it (or would if we had polarizing glasses on) not how we really see it with the naked eye. If this is true, then polarizing filters more or less fake the scene. But at least they fake it into something more enjoyable.

Gangolf Haub

Gangolf Haub - Jul 19, 2006 6:40 am - Hasn't voted

Re: Do our eyes auto-polarize?

What is real? I tend to share your view that pol filters enhance the view to something we don't usually see. However, since eyes and brain work together our perception usuallly also enhances what we really see. It's real hard to tell what is real...

Tie-Dye Mike

Tie-Dye Mike - Jul 27, 2006 9:01 am - Voted 10/10

Thanks Dude

I'm looking at getting a polarization filter for my camera,and this actually helps me make some sence of how this works. Thanks!

cho_mik

cho_mik - Jul 28, 2006 3:41 pm - Voted 9/10

UV filter

In digital photograpy UV filter is useles (until you use it as mechanical protection of the lens). I took test pictures on the top of Muztagata (7536 m.) and there was almost no difference.

Gangolf Haub

Gangolf Haub - Jul 29, 2006 8:46 am - Hasn't voted

Re: UV filter

I've noticed much the same thing though I'm not quite sure why. I know that to protect the CCD chips they have to be coated. Maybe this coating also already reduces UV transmission. Another explanation would be that the chips simply don't work in the UV-range. But if I remember correctly (from my time as physics student) CCDs were used to detect soft X-rays and in that case the longer wavelength UV must leave their marks on the chips. However, different cameras - different behaviour :-)

cho_mik

cho_mik - Jul 31, 2006 6:40 am - Voted 9/10

Re: UV filter

To be precise - I used Nikon D70 :)

sich

sich - Sep 1, 2006 2:40 pm - Voted 10/10

Re: UV filter

Yes, CCDs and CMOS in a photocameras are almost insensitive for UV.
A wavelenght function of CCD could be found for example at http://images.google.com/images?hl=en&lr=&c2coff=1&q=ccd+spectral+response.
A relative UV sensitivity is about 1% (relative to its most sensitive wavelength). So, there is no need on the digital cameras for main purpose of UV filters. We can only use it for lens protection.
This is because of CCD or CMOS spectrum, but also we need to have in the mind that any type of a glass are a good apsorber of UV so even the lenses do not allow many UV rays to pass to chip.
Also lowest part of earth atmosphere absorb good amount of UV, so only at high, high altitudes you can see result form UV filters, but all that is just minor difference.
Conclusion is that the chips in the cameras with help of the lenses are killing UV, so we don't need the UV filters.

huckleberry

huckleberry - Aug 1, 2006 7:25 pm - Voted 10/10

Thanks Gangolf!

Great article. I think that this is great because of how much photography there is on SP...Keep it up

chskier

chskier - Aug 23, 2006 6:27 pm - Voted 10/10

Aid

Nice article. I use filters out there, too but I just thought I would comment on something you stated in the begining of your post and address the whole issue of "cheating".

Basically, it does not matter if one uses Photoshop, or filters or both. The point of making an image is that you use whatever you have at your disposal to created an image that is powerful and compelling. Follow the flawed logic that meny people live by and one could make the point that using a camera is cheating! That's what some landscape painters claimed when photography first gained popularity.

There is a fairly well known landscape photographer who insists that he never uses filters or photoshop, yet his prints are stunnily colorful. Perhaps he wants us to believe that somehow his raw skills make their way to the film and create color but what he actually does is use specialized paper and chemical processes. So, one technology is used and the others denounced. This is merely politics.

All technologies are fair game. It does not matter how you arrive at a dynamic image; the point is that you got there. Some climbers scoff at aid climbing and they may have a point if reaching the goal is to be done so in the most challenging way but in the world of photography and art, applying filters and touching with Photoshop are not cheating. They just represent additional tools for us. It does not matter if creating the image is difficult or easy.

I'm sure you probably feel the same way about it but it is rarly stated, so I thought I would throw it out there. Maybe I am stating the obvious, though.

Cheers!

Gangolf Haub

Gangolf Haub - Aug 23, 2006 7:22 pm - Hasn't voted

Re: Aid

Your point is well made and I guess you are right though I must say that there are different levels of "aid" or "cheating" that people find acceptable. From the artistic point of view there can be no question - an artist is allowed to do whatever he wants to produce his art. But even here there are different levels of acceptance for different people. We just all have our own opinions of what can be done and what should be left undone. Which I think is good.

chskier

chskier - Aug 23, 2006 8:07 pm - Voted 10/10

Re: Aid

I think it's good, too, just as long as nobody tries to force their ordinary opinions on us. Which is what I was trying to address. The photograph does not care what we think. "It" is what it is and how it is but it's up to us try to make it the best it can be and if that means using Photoshop or filters, then that is what we must do regardless of our occupation. It is the photograph that should dictate what is acceptible and what is not.

I just like to try to remind folks that in order to improve one must be as objective as possible.

Thanks, again, for your article and your other efforts. I have enjoyed your contributions to SP very much.

oldandslow

oldandslow - May 29, 2007 2:24 pm - Hasn't voted

Vignetting, Sky Gradation and Sunglasses

For many years I experienced vignetting using a polarizer on my film SLR. I now use a digital SLR with a full size lens rather than a lens designed for a digital sensor. No vignetting so far having used two different cameras and lenses.
I recently encountered sky gradation using a polarizer with my new 17-40mm lens. The gradation appears to be inversely proportional to the focal length--not noticeable at 40mm. Is there anything I can do about this problem other than sharpen my clone stamp and smudge techniques?
Two years ago my son and I tried using sunglasses in lieu of a polarizer on a digital point and shoot. The only result was a change in color.

Gangolf Haub

Gangolf Haub - May 30, 2007 5:27 am - Hasn't voted

Re: Vignetting, Sky Gradation and Sunglasses

From a physical point of view you will always have sky gradation if the opening angle of your lense is to large. It is caused by the polarization effect itsel, which is dependent on the angle at which you see the light (speak:sun). The effect goes from nil to max in less than 90° which is why gradation is always visible at wide angle lenses like yours. At 40mm you have an opening angle of some 40° while at 17mm it should be above 90°.

andret

andret - Jul 5, 2007 12:45 pm - Hasn't voted

Thank you

I have a point and shoot right now, but this article was extremely helpful in detailing the polarization effect. The shots with the pol filter look great. I'm looking forward to an SLR...

peninsula

peninsula - Aug 5, 2007 10:08 am - Voted 10/10

Excellent Overview

Very good, thank you for your well written and depicted explanation!

eyedeebee

eyedeebee - Jun 8, 2009 11:28 am - Hasn't voted

Thanks

Very useful article - picked up some great pointers.

Gangolf Haub

Gangolf Haub - Jun 8, 2009 1:45 pm - Hasn't voted

Re: Thanks

You're welcome. Actually this was the intension of the article. I found I understood photography much better after having used a polarizer for the first time.

EricChu

EricChu - Jan 19, 2010 8:40 am - Voted 10/10

Thanks a lot, Gangolf!

This is a great help, as you go quite into detail on this page! It's also great that so many people commented the page, this also provides information.

Gangolf Haub

Gangolf Haub - Jan 19, 2010 9:25 am - Hasn't voted

Re: Thanks a lot, Gangolf!

You're welcome!

Viewing: 21-40 of 41