How will exposure be affected with a Linear Polarizer on EOS Camera

[maxasst]

I don't mind the errors in autofocusing, but I am concerned about

which direction an error in autoexposure goes. Does the linear

polarizer filter result in an over- or under- exposure? And is it a

linear error? Thanks.

[dogbert]

My guess is it should not affect exposure. TTL metering should take care of any filter effects.

[jean_jacques_lemaire]

Exposure will definitely be affected.<BR>

From what I know:<BR>

if you look at the back of your camera mirror, you've got a second smaller one, turned to the bottom. Actually light goes partly through the main mirror (yes there is a semi-transparent part) and get reflected to the bottom, to the TTL measurement cell.<BR>

So you've got one more reflection on metal than when light goes directly to the sensor/film. The amount of reflected light depends on light's polarization and its orientation. This small difference is the responsible for the exposure error.<BR>

Circular polarizers are actually linear polarizers combined with a quarter-wave plate, converting linear to circular polarization. Circular polarized light is reflected normally by metal mirrors, so with these polarizers you won't get the error.

[jbq]

I would expect that at one extreme you'll see underexposure by as much as a stop (the case where all the light is reflected by the primary mirror), in the other extreme you'll see some overexposure (possibly gross overexposure) (the case where all the light goes through the primary mirror).

 

Of course, that would be with "perfect" polarized reflection, which isn't the case in the real world.

 

You can measure the effect yourself, aim at an evenly lit gray card, and see how the exposure changes as the polarizer is rotated. If your polarizer is properly marked, one extreme effect will happen when the mark is at the top (or bottom, same thing), and the other extreme effect will happen when the mark is on either side.

[frank_dzambic]

I don't know absolutely for sure, but I would tend to agree with Geoff's answer. TTL should take care of the exposure. The need of a circular polarizer is for the autofocusing system, not the autoexposure system.

[jimstrutz]

You're wrong Frank. Circular polarisers were required for several pre-AF cameras, Canon among them, and the effect is *not* linear.

 

However, you can still use linear polarizers. Take a meter reading with the filter off, attach & adjust the filter, add 1.5 to 2 stops extra exposure & shoot. Although if you are using slide film it's best to bracket. Using a digital camera, shoot, check the histogram, adjust exposure & re-shoot.

 

Linear polarizers can also be used with a Canon AF system, and most of the time it works, but sometimes (depending on the orientation of the filter) AF simply fails completely. The work around is to turn the filter, lock autofocus, re-adjust the filter & shoot. They will not cause front or rear focusing errors.

[frank_dzambic]

Well, that's why I like this forum. I can still learn things I didn't know.

[steve_dunn2]

<cite>if you look at the back of your camera mirror, you've got a second smaller one, turned to the bottom. Actually light goes partly through the main mirror (yes there is a semi-transparent part) and get reflected to the bottom, to the TTL measurement cell.</cite>

 

<p>The description of the mirrors is correct, but the description of the sensor is not.</p>

 

<p>The device in the bottom of the mirror box is the AF system, not the exposure metering system. Ambient exposure metering is done by a sensor which looks at the focusing screen, above the mirror and immediately below the pentaprism/pentamirror. E-TTL and E-TTL II flash metering (on all Canon-designed EOS DSLRs as well as most EOS film SLRs in the last decade) also uses the ambient exposure metering sensor; A-TTL (film bodies only) and TTL (film bodies plus the early Kodak-designed EOS DSLRs in the 1990s) metering is done by a sensor which looks at the film/sensor during the exposure.</p>

[maxasst]

Thanks for the responses so far. It seems that the opinions are biased towards underexposure with the filter. I'll add some detail. I have two Cokin glass filters, a Pola-blue and a Pola blue/red. I don't know how many stops worth of light they are supposed to decrease the exposure since they don't have any accompanying literature. I'm guessing that they are linear because they were cheap and I tried looking through them and through a mirror but didn't see that things were particularly darker through the mirror.

[dogbert]

FWIW I have used a cheap cokin blue/yellow polar and it didn't seem to effect exposure. (Though I accpet that my orginal guess was wrong based on what others have said.)

[jean_jacques_lemaire]

Thanks for the correction, Steve :)