Linear polarizer, auto focus and metering with SLR.

[Thonord]

<Quote, CinC tutorial on "Camera lens filters":
Linear polarizers are much less expensive, but cannot be used with cameras that have through-the-lens (TTL) metering and auto focus — meaning nearly all digital SLR cameras.
<Unquote.

I struggle to understand this. I don't doubt it being correct, but I don't understand why.

Why won't auto focus, and metering, (single spot), work with a linear polarizer?

[davidedric]

As I understand it, the beam splitter, or semi-silvered mirror, which diverts light for metering and autofocus is sensitive to linear polarised light.

Incidentally, a circular polariser is actually a linear polariser backed by a "quarter wave plate" which transforms the linear polarisation into circular.

As to WHY the mirror is sensitive to linear polarisation, I'm afraid my high school physics are too far behind me

(edit: light dawns) Aha! A mirror polarises light. That's why the filter works to reduce reflections. Hence, the amount of light reflected by the mirror depends on the angle between the light's polarisation and the polarising plane of the mirror. Well, it makes sense to me!

Dave

[Kodiak]

Why won't auto focus, and metering, (single spot), work with a linear polarizer?

A very very simple answer:
Because the CP is too dark! …this will affect the auto focus. Not a problem
since you most likely using a tripod and do a manual focus prior to mounting
the filter. When it comes to the spot metering, the CP is not (my experience)
affect light metering.

[DanK]

My understanding of optics is limited enough that I have to rely on what I have found. One very trustworthy source is Bob Atkins, who explained why linear polarizers can mess up metering in this article.

Older SLRs used fully reflecting mirrors, which reflect all polarizations with equal intensity. Newer SLRs, and particularly autofocus SLRs, often use partially reflecting mirrors. The reflected light goes to the viewfinder and metering systems, while the transmitted light goes on to the auto focus sensors. If a linear polarizer was used on such an SLR, the intensity of light sent to the metering system would depend not only on the intensity of the light, but also it's polarization angle. However, when a picture is taken the mirror is out of the light path and the film is only sensitive to the intensity of the light, not its polarization. Thus metering errors can (and do) occur when a linear polarizer is used with an SLR which has a partially reflecting main mirror.

I don't have a known source for the AF issue, but I did find this:

Modern autofocus systems use mirrors that act as beam-splitters: most of the light is reflected to the viewfinder for metering and viewing, while the rest (typically 25%) is transmitted and then reflected by a secondary mirror to the autofocus sensor, which is in the camera body. With a polarizing filter attached, the ratio of reflected/transmitted light (fixed for non-polarized light) varies with the polarization plane orientation of polarized light, relative to the beam-splitter, causing the amount of light reaching the autofocus sensor to vary. This effect can be particularly troublesome if your front element moves with the lens, as it would cause the amount of light reaching the sensor to change while the lens focuses.

While you can see (rather easily) when your autofocus system has been confused by a linear polarizer, there isn't such an easy way of knowing whether or not the internal light metering system has been stumped.

[JR1]

In fact

There are two types of polarizing filters readily available, linear and "circular", which have exactly the same effect photographically. But the metering and auto-focus sensors in certain cameras, including virtually all auto-focus SLRs, will not work properly with linear polarizers because the beam splitters used to split off the light for focusing and metering are polarization-dependent. Linearly-polarized light may also defeat the action of the Anti-aliasing filter (Low-pass filter) on the imaging sensor.

[rpcrowe]

I was first introduced to a Circular Polarizer in 1980 when a Tiffen Representative gave a filter presentation at my Navy Unit at the time: Pacific Fleet Audio Visual Command.

At that time I had a Pentax ME and was not getting good exposure while using a linear polarizer (which was the only polarizer I knew about). The Tiffen Rep sent me a circular polarizer which worked fine. BTW... of course the Pentax ME was a manual focus camera so the linear polarizer did not interfere with focus...

BTW: at that time, Tiffen filters were considered top-line gear. Unfortunately, that is not so today!

[Manfred M]

The answers around the beam splitter are correct. A semi-silvered mirror is used in all DSLRs that I know of, and the semi-silvering is created by making the mirror with many very fine lines when the main mirror is constructed, allowing some of the light to pass through it. This means your mirror effectively is effectively a linear polarizer, hence a linear polarizer added on top via an external filter means that the amount of light being transmitted to the metering cells is reduced, causing the light meter to produce an incorrect reading.

[Glenn NK]

I used a linear polarizer way back when and don't recall the problem with the part of the sky that was the most off-axis to the sun being darker (might have forgotten).

Question: would a linear polarizer have the same effect - one part of sky darkened more than the other?

If not, then a linear polarizer would still be very useful for landscape work - how often does one use AF for landscapes? I never do.

Glenn

[revi]

I used a linear polarizer way back when and don't recall the problem with the part of the sky that was the most off-axis to the sun being darker (might have forgotten).

Question: would a linear polarizer have the same effect - one part of sky darkened more than the other?

If not, then a linear polarizer would still be very useful for landscape work - how often does one use AF for landscapes? I never do.

Glenn

Short answer: yes.

This effect was known already when linear polarisers were the norm. And, as already stated above, a circular polariser is a linear polariser followed by a 'quarter-wave plate', so you still get the effects on intensity of a linear polariser.

[Thonord]

What a wonderful source of information this is. I need to educate myself om beam splitters, and the construction of circular polarizers.

Thanks to all

[pnodrog]

My understanding of optics is limited enough that I have to rely on what I have found. One very trustworthy source is Bob Atkins, who explained why linear polarizers can mess up metering in this article.



I don't have a known source for the AF issue, but I did find this:

Interesting - now take away the mirror and use live view - what happens then?

Actually I must try it. I have a good linear polarizer kicking around somewhere and I might just check what happens.

[Tringa]

I don't want to hijack this interesting thread but, are linear polarisers rare beasts?

When I bought my first DSLR a few years ago I read somewhere that the older linear polarisers could not be used with DSLRs and a newer circular polariser would be needed.

I thought a polarising filter I had bought around 1969 would be no use but decided to give it a try. As it works perfectly I assume it must be a circular one which surprised me. Is it just that linear polarisers are not that common and were not, even 40 plus years ago?

Dave

[DanK]

I don't want to hijack this interesting thread but, are linear polarisers rare beasts?

When I bought my first DSLR a few years ago I read somewhere that the older linear polarisers could not be used with DSLRs and a newer circular polariser would be needed.

I thought a polarising filter I had bought around 1969 would be no use but decided to give it a try. As it works perfectly I assume it must be a circular one which surprised me. Is it just that linear polarisers are not that common and were not, even 40 plus years ago?

Dave

I don't know if this is generally true, but if I compare the one linear polarizer I still have to my circular polarizer, the effect of rotating the outer element of the filter is far greater with the linear polarizer. Might be worth comparing.

[Mike Buckley]

Is it just that linear polarisers are not that common and were not, even 40 plus years ago?

I'm not a gear head, but I never heard of a circular polarizer until I began using a digital point-and-shoot camera (not a DSLR) that required using one instead of a linear polarizer. Linear polarizers have always been readily available at least from the 1980s to the present. Conduct a search of linear polarizers and you will find them offered by a variety of companies that are widely known for their circular polarizers.

[Inkanyezi]

Most polarisers back in fifties and sixties were linear, but with the event of zoom lenses for movie cameras, containing beam splitters, the circular polarisers were introduced. A linear polariser will not work as designed with such a lens, as the beam splitter redirects light polarised in one direction and lets the rest through. It is possible to blackout the viewfinder with a linear polariser and such a lens.

Then in the sixties, Canon made a camera with a semi-transparent (beam-splitting) mirror, and it would need circular as well. At that time, the circular polarisers were uncommon. However the TTL metering cameras from Topcon from early sixties and further would not meter correctly with a linear polariser, although it wasn't mentioned in the manual.

Now, when all DSLR cameras have beam splitters for AF as well as for spot metering, the circular polariser is more common than the linear.

Point and shoot cameras or mirror-free systems however do not need a circular polariser, but work well with linear polarisers. The linear polariser transmits somewhat more light than the circular.

[Tringa]

Thanks for the info on polarisers, all.

Dave

[azkitch1]

To bring new life to this thread...Linear polarizers weren't called linear until the proliferation of circular polarizers made it necessary.
It almost sounds like the mirrorless systems, like Sony Alpha, could get away without CPLs. But then I saw the post saying the anti-aliasing filter on the imaging sensor. Hmm.
I have an adaptor for Minolta lenses onto my Sony A6000. It's MF only, so if I stumble across a 49MM LPOL, I'll give it a whirl.

I hope someone reads this great thread again!

[Manfred M]

To bring new life to this thread...Linear polarizers weren't called linear until the proliferation of circular polarizers made it necessary.

That is not correct. My Leica R3 (from 1976) required the use of a Circular Polarizer based on the way the metering system was implemented. and that is exactly what the users manual refers to it as. So far as I know, both the Leicaflex SL and SL2 used a similar dichroic (beam splitting) mirror arrangement, which suggests that circular polarizers were in use as early as 1968, when the SL camera was launched.

[xpatUSA]

To bring new life to this thread...Linear polarizers weren't called linear until the proliferation of circular polarizers made it necessary.
It almost sounds like the mirrorless systems, like Sony Alpha, could get away without CPLs. But then I saw the post saying the anti-aliasing filter on the imaging sensor. Hmm.
I have an adaptor for Minolta lenses onto my Sony A6000. It's MF only, so if I stumble across a 49MM LPOL, I'll give it a whirl.

Dave, I have a 49mm linear polarizer and a mirror-less camera with that filter thread. How would I give it a whirl?

[Manfred M]

Dave, I have a 49mm linear polarizer and a mirror-less camera with that filter thread. How would I give it a whirl?

I suggest he wants to know if autofocus works and the exposure is correct, regardless of the orientation of the linear polarizing filter.