Shooting with a CPL filter affecting metering?

[BrianA61]

I rarely, if ever, shoot with a CPL on. But, on occasion, I like the enhance colors and contrast they provide. The last couple of times I've shot with one, the photos have come out dark. I was out earlier this week and decided to shoot with a +1/3 exposure setting which seemed to correct the exposure issue. What I'm wondering is why doesn't the camera sensor correct for the difference in exposure? Maybe there is a reason I don't shoot with anything other than a UV filter on.

[Mike Buckley]

I almost always use a polarizer when shooting outside and I use it occasionally inside my makeshift studio. The camera meter has always automatically taken into account that the filter reduces the amount of light entering the lens. You compensated only 1/3 of a stop. Considering that a polarizer typically reduces the amount of light by about 1 1/2 to 2 stops, your adjustment of such a small amount is proof to me that your meter automatically adjusted.

[dje]

Brian I can only speculate - it may depend on the nature of the scene and the metering mode used. A polariser could reduce the brightness of different parts of the scene by different amounts (eg the sky). If the metering mode gives emphasis to one part of the scene, this could result in different exposure with the polariser on and off. eg spot or centre-weighted metering in particular.

Dave

[BrianA61]

Brian I can only speculate - it may depend on the nature of the scene and the metering mode used. A polariser could reduce the brightness of different parts of the scene by different amounts (eg the sky). If the metering mode gives emphasis to one part of the scene, this could result in different exposure with the polariser on and off. eg spot or centre-weighted metering in particular.

Dave

I had my metering mode set to 'evaluative'. It measures the overall scene equally. I also did not have the full effect set for the filter. Color/contrast was too strong for my taste.

[Manfred M]

Color/contrast was too strong for my taste.

Without seeing the image you are writing about, it is hard to comment. That being said, vivid, contrasty skies are definitely a possible outcome of using a polarizer. The sky can darken / saturate considerably and the clouds can be quite contrasty. The solution is not to dial in maximum polarization and to back off a bit when adjusting the filter for the shot.

[BrianA61]

Manfred, I did not capture a shot with the filter set at full effect. I backed off somewhat. My main concern was the shot came out dark as if the camera's meter wasn't reading the correct exposure and compensating for it. I'll go back through my shots and post a couple so folks can see what I'm talking about.

[Mike Buckley]

Brian,

Is your issue that you had to compensate by 1/3 stop? If so, I'm at a complete loss as to why that's an issue.

[rpcrowe]

Shooting with Auto Exposure Bracketing gives you the best of all worlds. It is extremely simple with Canon gear - simply set the AEB how you want it and put the camera in burst mode. Then every time you press the shutter button, the camera shoots 3-bracketed shots and stops until the next time you press the shutter.

With the high speed burst mode of my 7D2, the three shots are cooked off in only slightly more time than a single shot. You can (and I often do) combine AEB with Exposure Compensation...

AEB shooting does eat up 3x the memory but, memory is fairly cheap on today's market.

If you use a tripod or monopod or hand hold steadily or brace your camera on something steady, you can often use the three bracketed exposures as a basis for an HDR image...

[Black Pearl]

A third of a stop seems like a tiny amount of compensation to 'correct' a meter by, certainly in situations where you're manually tweaking a filters effectiveness from shot to shot and the lighting is both variable and uncontrolled.

[xpatUSA]

I had my metering mode set to 'evaluative'. It measures the overall scene equally.

Maybe that's it right there . . . speaking as one who has a horror of letting the camera decide anything for me . . .

[pnodrog]

If a CPL filter strongly darkens the sky in comparison to the rest of the photograph with "evaluative metering" the camera should increase the exposure to compensate therefor the other area's should brighten slightly. What you are observing seems odd.

[DanK]

I had my metering mode set to 'evaluative'. It measures the overall scene equally.

No, it doesn't. That's why it has that name: it uses an algorithm to evaluate the entire scene. E.g., from http://www.learn.usa.canon.com/resou...etering.shtml:

This commonly used metering system analyzes surroundings to compare the brightness or darkness of the picture area to the metered subject, then adjusts exposure accordingly.

My hunch is the same as Dave's: a polarizer can change the relative brightness of different parts of the image, and depending on the evaluative algorithm, that could throw off exposure. For that matter, the algorithm is often wrong anyway. You could check this out by shooting a uniform, glare-free, and uniformly lit surface with and without a CPL, using either spot or partial metering, both of which take a simple reading from one area. The two should be exposed similarly regardless of the filter.

[rpcrowe]

Just a thought... Although a third of a stop is very little correction needed and could be individual taste in PP, I wonder if the filter type itself might or might not have any impact on the exposure reading.

I know that a linear polarizing filter may throw the exposure meter off and that the circular polarizer is meant to correct that. But, is it possible that circular polarization from one company to another might be off a tad? Or even off from filter copy to filter copy? Is it a standard 1/3 off over all your shooting or does the type of shot make any difference

[pendennis]

Just a thought... Although a third of a stop is very little correction needed and could be individual taste in PP, I wonder if the filter type itself might or might not have any impact on the exposure reading.

I know that a linear polarizing filter may throw the exposure meter off and that the circular polarizer is meant to correct that. But, is it possible that circular polarization from one company to another might be off a tad? Or even off from filter copy to filter copy? Is it a standard 1/3 off over all your shooting or does the type of shot make any difference

There's no real difference in linear and circular polarizers for exposure. The circular polarizer was designed so the lines in the linear polarizing filter wouldn't interfere with auto-focusing. You can use circular polarizers on manual focus cameras, but not vice versa.

In addressing the OP's error, he should try center-weighted, or even average metering. He doesn't tell us what lens was used, and if he was using his 17mm-40mm f4, the sky at wider than 24mm can be very uneven. When metering at very wide angles, it gets dicey with a lot of sky maybe overwhelming the total scene.

Finally, if unsure, and you have the time to do it - Bracket, bracket, bracket. Set the exposure manualy, and shoot over, spot-on, and under.

[pendennis]

I rarely, if ever, shoot with a CPL on. But, on occasion, I like the enhance colors and contrast they provide. The last couple of times I've shot with one, the photos have come out dark. I was out earlier this week and decided to shoot with a +1/3 exposure setting which seemed to correct the exposure issue. What I'm wondering is why doesn't the camera sensor correct for the difference in exposure? Maybe there is a reason I don't shoot with anything other than a UV filter on.

You have a Canon lens (17mm-40mm). At the wider angles (24mm and wider) all polarizers tend to cause problems with the sky, and they'll provide very uneven results. Even at 24mm, I've seen uneven skies. At the edges of the polarizer, the angle of light is different than on-axis. The camera's meter can get fooled by this, and it'll go with whatever algorithm was programmed into the metering system.

As I mentioned in my earlier post, you should bracket your exposures, and change your metering to center-weighted, or even average. Finally, it may not be a bad idea to bracket, using manual exposure. Start by overexposing, then what the meter shows as exact exposure, and then underexposing.

[dem]

If you are using a DSLR, which you are, both exposure and metering will be affected by a linear polarizer but neither of them will be affected by a CPL. That's why CPLs were introduced in the first place. I expect the small difference in exposure is due to the metering mode being too clever. If you switch to average metering, this should produce a more consistent exposure between the shots.