Underexposure with polarizer

[Blake Cook]

I have added a Canon drop-in circular polarizer to my 300 f/2.8, 1D4 combination and have been experiencing behaviour I've not see before. I adjust the polarizer for a particular direction of shot and when I shoot in that direction it is underexposed by about 1 1/2 stops using spot metering and about 1/2 stop using evaluative. Shooting either side of that, produces normally exposed shots.
It's like the metering does not see a reduction of light from the polarizing effect.
Previously I did not see the same effect with a circular polarizer on a 100-400, 5D2. No sun today so I can't try the polarized 100-400 o the 1D4.
Is this effect normal?

Shooting manual seems to be the only way around it.

I've posted the question elsewhere. No replies.

[Christopher C.M. Cooke]

This is a common problem and I address it by upping the exposure steps and I have found that this is best done by trial and error.

I also tend to shoot in Manual mode when using a CPL.

[John Chardine]

Best thing to do is not use a polariser. In the digital age, the value of such a filter has diminished greatly.

[Blake Cook]

Thanks for the comments Chris and John.

I guess I never ran into the problem in the past with my 100-400 because I almost always shoot aircraft in manual mode.

I haven't used it for birds but I do want to try it and see what it does for white feathers.

[Roger Clark]

Blake,

It is possible the circular polarizer is in backwards in the mount. If so, that would put the linear polarizer (a circular polarizer is a linear polarizer plus a quarter-wave plate) at the exit of the filter. The quarter wave plate is supposed to be closest to the camera. Linear polarization will cause havoc with the metering system. This page shows such an example, where the meter was in error by over 5 stops:
http://www.clarkvision.com/articles/...izing_filters/

I have a drop-in polarizer for my 500 f/4 (I haven't tried putting it in my 300 f/2.8), but I rarely use it. I do disagree with John about not needing one in the digital age. No digital filter can do what a polarizer does, which is cut haze, and cut first-surface reflections, e.g. from water and from windows. It also improves color saturation on leaves (again by reducing first-surface reflections). While you can boost contrast and saturation to mimic a polarizer in a photo editor but at the expense of noise. In some cases, the reflection can be so strong, no matter how much processing one does, you can't get the subject behind the reflection to be clear.

Roger

[Blake Cook]

Roger,

I do agree a polarizer can deliver photos that would be completely lost otherwise.

It appears the polarizer is set correctly If I look at an LCD with the drop-in CP in one orientation from the camera side, it will go close to black, while turned around looking at the LCD from the lens front direction it dims only slightly (probably the 1.3 stops Canon claims)
My 77mm CP does exactly the same thing in those orientations.

[Jeroen Wijnands]

Personally I don't think a polarizer is totally obsolete in this day and age. After all, cutting down on reflection still can't be done on the computer.

I tend to dial in +1EV when shooting with a polarizer, works well most of the time.

[John Chardine]

I was a bit flippant with my comment on polarisers! I agree, some polarising filter effects would be hard to do in post processing, however, personally I question the need to do some of these to begin with. I don't much like the effect a polariser has on foliage making it appear like velvet- I find it unreal, and after all, reflections are part of the real world. Perhaps in the case of wanting to reduce reflections on water, then they have a use. If you are looking for a contrast boost, then this can be done at no cost in post processing from a RAW image. Deep blue skies are a bit more of a challenge but still no problem in PP, and they can be made at any angle to the sun, not just at 90°! I didn't realise that polarisers cut through haze Roger- is this simply an overall contrast effect or is it real? I do have a polariser but it's one for my landscape lenses, and frankly it hardly comes out of the case these days. I would never consider using one with my 500 for birds just because I see no benefit to offset the big hit in light of about one stop or more. And then you have the age-old controversy of putting another piece of glass in the light path of your nice and expensive lens.

[Roger Clark]

I was a bit flippant with my comment on polarisers! I agree, some polarising filter effects would be hard to do in post processing, however, personally I question the need to do some of these to begin with. I don't much like the effect a polariser has on foliage making it appear like velvet- I find it unreal, and after all, reflections are part of the real world.

Hi John,
The come-back would, but it's art!;) Also, if you wear polarized sun glasses, it is what you see in the real world.

Perhaps in the case of wanting to reduce reflections on water, then they have a use. If you are looking for a contrast boost, then this can be done at no cost in post processing from a RAW image. Deep blue skies are a bit more of a challenge but still no problem in PP, and they can be made at any angle to the sun, not just at 90°! I didn't realise that polarisers cut through haze Roger- is this simply an overall contrast effect or is it real?

If the haze is very small particles, the haze is polarized too, so has the same effect as blue sky. But larger haze particles, larger than a fraction of a micron, do not Rayleigh scatter so you don't see the polarizing effect. My experience has been that the aerosol haze in the eastern US contains large particles and a polarizer will not help. In more arid regions, like the western North America, a nice clear day that is just slightly hazy often contains very small particles and a polarizer cuts that haze (of course max at 90 degrees from the sun) and can make a so-so landscape into a gorgeous one.

I do have a polariser but it's one for my landscape lenses, and frankly it hardly comes out of the case these days. I would never consider using one with my 500 for birds just because I see no benefit to offset the big hit in light of about one stop or more. And then you have the age-old controversy of putting another piece of glass in the light path of your nice and expensive lens.

In the case of super telephotos with drop-in filters, you pull out a holder with a clear glass filter and put in the polarizer, so there is no increase in glass. In fact if you leave the glass out, then the lens will be soft.

Your experience mirrors mine. I bought the polarizer for the 500 thinking I might use it, but beyond a couple of test shots, I never have used it. I do use polarizers for landscape images, mainly to cut that thin haze,
but I use polarizers far less than when I first started photography.

In general I carry an ND filter (mainly for slowing down shutter speeds on things like waterfalls) and a polarizer.

Back to Blake's problem, I have not had problems with metering with polarizers so I am not sure what the problem is if the filter is in correctly. If the polarizer is not real high quality, that could have some residual linear polarization and affect the metering a little. But I would expect the canon filter to be reasonable quality.

Roger

[Blake Cook]

Roger,

There is no choice in quality of polarizers. The only one available is from Canon. Since it only fits the Super-teles is suspect the quality is quite good. I notice that if I look at an LCD with it in one orientation from the camera side, it will go close to black, while turned around looking at the LCD from the lens front direction it dims only slightly (probably the 1.3 stops Canon claims)
My 77mm CP does exactly the same thing in those orientations. That appears to rule out that it's a linear polarizer.

John,
I agree with your point about foliage, however there are few leaves on the planes I shoot. :)

[Sabyasachi Patra]

I have a polariser for my 300 f4 IS as well as a drop in circular polariser for my 400 f2.8 L IS. The drop in polariser for the 400mm was the most difficult to find. I could neither get it in India, nor singapore. Finally, had to get it from Japan. However, I haven't used it much. :o You don't find too many situations where the light angle is right, and the subject is perfect for use of polariser. However, I have never experienced any problems in metering. It is normal that there would be drop in light, but metering remains the same. So I am unable to know why Blake is having a problem.

Cheers,
Sabyasachi

[Charles Glatzer]

Shooting manual seems to be the only way around it.

I've posted the question elsewhere. No replies.

Blake,

Why would it make an exposure difference shooting manually or auto, assuming you are using the same pattern to determine exposure? If you vary the degree from the null point by physically changing the settings manually or by applying compensation in auto... it is essentially one and the same.

Best,

Chas

[Blake Cook]

Blake,

Why would it make an exposure difference shooting manually or auto, assuming you are using the same pattern to determine exposure? If you vary the degree from the null point by physically changing the settings manually or by applying compensation in auto... it is essentially one and the same.

Best,

Chas

Chas,

While panning aircraft in flight in Av mode the Tv goes from 1/1600th to 1/6400th as I pan through the zone where the filter is set. If manually set at 1/1600th the photos are close to correctly exposed at the effective zone of the polarizer.

Blake

[Charles Glatzer]

Chas,

While panning aircraft in flight in Av mode the Tv goes from 1/1600th to 1/6400th as I pan through the zone where the filter is set. If manually set at 1/1600th the photos are close to correctly exposed at the effective zone of the polarizer.

Blake

Av will always change Tv as the background changes... in this case due to varying degrees of sky polarization. So why use auto exposure when the light on the subject remains constant?

Your needing to open up an additional 1/2 stop tells me the portion of the sky you metered is 1/2 stop brighter than a mid-tone value. Remember... a spot meter will render what is in the pattern as a mid-tone, it does not suggest the metered value is a mid-tone, nor tell us how much to vary exposure to render what is in the pattern correctly.

If shooting the plane in bright sunlight 1/1600 @ f/5.6, ISO 200 (sunny f/16 equiv) should be correct. If using a polarizer I typically open up 2EV, changing either the f/stop and/or increasing ISO to maintain the desired shutter-speed.

I think we are on the same page ;)

Best,

Chas

[Blake Cook]

Chas,

What puzzles me is the AE perceives the zone where the polarizer is most effective as being 2 stops BRIGHTER than the zones either side and sets Tv 2 stops faster.

Blake

[Charles Glatzer]

Chas,

What puzzles me is the AE perceives the zone where the polarizer is most effective as being 2 stops BRIGHTER than the zones either side and sets Tv 2 stops faster.

Blake

Are you saying the same meter pattern sees the darker part of the polarized sky as lighter than the lesser polarized sky area, and this does not occur if you meter off the same spot in manual mode?

Not sure why this would occur.

Chas

[James Shadle]

Have you tried center weighted metering rather than evaluative?
James

[Charles Glatzer]

Have you tried center weighted metering rather than evaluative?
James

First post says he was using spot. It would only make sense if the spot pattern was on the brighter plane for one and the background for another.

Regardless of the meter pattern, manual exposure is the way to go.

Chas

[Blake Cook]

Are you saying the same meter pattern sees the darker part of the polarized sky as lighter than the lesser polarized sky area, and this does not occur if you meter off the same spot in manual mode?

Not sure why this would occur.

Chas

Exactly! Although I didn't try spot metering in manual mode at the time this occurred. (The plane didn't wait around :) ) The effect was more pronounced in "Spot Meterering". Go figure!


Blake

[James Shadle]

First post says he was using spot. It would only make sense if the spot pattern was on the brighter plane for one and the background for another.

Regardless of the meter pattern, manual exposure is the way to go.

Chas

This is what the first post said " I shoot in that direction it is underexposed by about 1 1/2 stops using spot metering and about 1/2 stop using evaluative."

No center weighted metering.
I agree that manual exposure is the best option.