Two Polarizing Filters = One Variable ND Filter?

[SueMe82]

I've read on the net (here) that if you use two polarizing filters together you can make a "infinitely variable ND Filter", i was just wanting to know have any of you tried this and if so how does it compare to using a higher strength ND Filter?

 

[CPayton]

Ya, I tried that. It does work but with the filters I have (not very expensive) it tended to leave a purplish cast at higher densities. I guess it would be possible to correct it using a whibal or other white balancing method. I'd probably use it in a pinch but would use ND filters if available.

 

[Sideburns]

It works. My photography teacher actually said that if we had two polarizers to try it...

You can put them 90 degress to each other and get a pretty much black viewfinder....and do exposures for quite some time in bright daylight.

I haven't tried it though...

 

[usayit]

I purchased a UV filter and two polarizing filters. Popped out the polarizing glass and the UV glass from their filters and inserted the polarizing glass into what was once the filter ring for the UV glass. I then screwed it too the rotating filter ring of the 2nd polarizing filter. Having two rotating filter rings made it difficult to adjust and make the whole setup a bit too thick (possible vignette).

It works like a champ....

A while back, one of my buddies didn't believe me so I made a quick video

http://www.silveredemulsions.com/temp/VarND.mp4

 

[usayit]

btw... no idea about their quality but they are also made for purchase too.

http://www.singh-ray.com/varind.html

 

[Garbz]

I also suggest you get one linear and one circular polariser. Depending on the design the linear one on the outside will allow you much greater consistency and the true ability to get 100% blackness.

 

[Josh66]

Cool, this sounds like something I have to try.

 

[jstuedle]

I also suggest you get one linear and one circular polariser. Depending on the design the linear one on the outside will allow you much greater consistency and the true ability to get 100% blackness.

The only issue with this solution is the linear poly will prevent the AF system on most if not all cameras from working. AF is the reason circular poly's were made.

 

[Helen B]

That's why Garbz suggested that the linear should be at the front. The light arrives at the linear polarizer, then passes through the circular polarizer - you can consider a circular polarizer (in simplistic terms) as being a linear polarizer followed by a depolarizer.

Two circular polarizers would not work unless the front one was the wrong way round, nor would a circular in front of a linear.

Best,
Helen

 

[usayit]

Two circular polarizers would not work unless the front one was the wrong way round, nor would a circular in front of a linear.

Really? Mine works and it is made from two circular polarizers... I have to admit, I don't have an scientific understanding of polarizers. All I know is that I figured it out via experimentation and set on a DIY project. Perhaps when I popped out one of the polarizers from its rotating filter ring, I flipped it around when I remounted on the UV's ring. I swear that I had them both facing forward when I was "playing" with them at the counter.

Garbz suggestion of using a linear polarizer has me very curious. I think I'll have to find one and give it a shot. I'm not concerned with how it affects AF as the camera that the "VariND" filter is used on doesn't have AF.

 

[Helen B]

Here's a test.

If your first circular polarizer is mounted with the delay plate in front of the polarizer it will have no effect as a polarizer - ie it won't affect the sky or reflections as it is rotated. This is, in fact, the best way of making a variable ND that has no polarizing effect. One circular polarizer backwards, then one forwards.

Best,
Helen

 

[RacePhoto]

Here's a test.

If your first circular polarizer is mounted with the delay plate in front of the polarizer it will have no effect as a polarizer - ie it won't affect the sky or reflections as it is rotated. This is, in fact, the best way of making a variable ND that has no polarizing effect. One circular polarizer backwards, then one forwards.

Best,
Helen

So what you are saying, is when I used two circular polarizing filters for a test last Summer, I was only using one? :lmao:

I understand what you are saying and the why, I just don't see how the filter is polarizing and unpolarizing and making any difference if there's only one? Wouldn't it remove the effects, and make the filter do nothing in that case as well.

But you do make a good point. Now I have to go test the theory and see if I was just wasting time and money by using the second filter. :er:

Good news is that old linear 58mm filter from the 60s is still in a box somewhere, and my new 100mm lens is a 58mm threading. Perfect for side by side testing against two 77mm circulars on a different lens.

 

[Helen B]

So what you are saying, is when I used two circular polarizing filters for a test last Summer, I was only using one? :lmao:

If you used two circular polarizing filters, both mounted the same way, then only the first filter would have an effect on polarized light, but both filters would act as neutral density. If the delay plate at the back of the first filter isn't doing its job properly then there will be some variable density effect as the two filters are rotated in relation to one another.

I understand what you are saying and the why, I just don't see how the filter is polarizing and unpolarizing and making any difference if there's only one? Wouldn't it remove the effects, and make the filter do nothing in that case as well.

Just to recap, a circular polarizer consists of two elements, a linear polarizer foil and a 'delay plate'. The polarizer foil removes light. Only light polarized in one plane gets through, the rest is lost. The remaining light is then 'depolarized' by the delay plate. The delay plate does not restore the light that has been lost, it just depolarizes the polarized light coming from the polarizer (in simple terms).

This is important because as we have discussed above, the coefficient of reflection at an air/glass surface is different for different planes of polarization. If TTL exposure or autofocus sensors are behind beamsplitters (ie sending some light to the viewfinder screen, some to the sensor) the proportion of light sent to the sensor will vary with the polarization of the incoming light. This isn't a good thing.

Best,
Helen

 

[Garbz]

Simple explaination:

Circular in | circular out -> circular in | circular out -> camera

If you take the 0degree polarisation in and spit a circular out, and then take the 90degree polarisation at the second interface you can not predict which way the light hits it at the interface.

You will get either a non-black image, and most likely an image which changes in tone across it.

Now suppose circular in | linear out -> linear in | circular out

You can guarantee the first filter will produce a 0degree polarised light so if the second filter is polarised at 90degrees you get zip out. If it's at 45degrees you get the 50% component out circular at the back.

This image very badly illustrates the example. This is what happens when you have 3 polarisers and shows how light takes the polarisation of the filter preceeding it. The screen is a linear polariser, the other two are circular. The interface between the two there's one behind the filter which is 90degrees out of phase with the screen. This is an interesting property of having 3 filters and has nothing to do with the post in question HOWEVER:

Have a look at the interface between the two filters? The left part is slightly darker than the right which is due to the circular nature of the furthest filter.

 

[Helen B]

I’ve mentioned in another thread (http://www.thephotoforum.com/forum/showthread.php?p=1088303) that the materials used to make photographic polarizers are never perfect. This means that polarizers will never behave exactly as theory would suggest. For example, circular polarizers may not produce perfectly circular polarized light – in fact it is almost impossible to do it for all wavelengths of light. There will be some linear component. This doesn’t matter much in normal use as a single polarizer, but it could affect their performance when used in crossed polarizers when they will behave like imperfect linear polarizers.

Here is a list of the ideal properties of various elements. ‘Unpolarized light’ refers to light with no dominant direction of polarization.

Linear polarizer

Unpolarized light in: linear polarized light out, with one stop reduction.

Linear polarized light in: linear polarized light out, with the degree of reduction varying from none to all depending on the angle between the planes of polarization of the incoming light and the linear polarizer.

Circular polarized light in: linear polarized light out, with one stop reduction.


Quarter-wave delay plate

Unpolarized light in: unpolarized light out, no reduction.

Linear polarized light in: circular polarized light out, no reduction.

Circular polarized light in: linear polarized light out, no reduction.


Circular polarizer, used forwards (ie linear polarizer followed by a quarter-wave delay plate at 45° to the linear polarizer)

Unpolarized light in: circular polarized light out, one stop reduction.

Linear polarized light in: circular polarized light out, with the degree of reduction varying from none to all depending on the angle between the planes of polarization of the incoming light and the linear element.

Circular polarized light in: circular polarized light out, one stop reduction.


Circular polarizer, used backwards (ie quarter-wave delay plate followed by a linear polarizer at 45° to the delay plate)

Unpolarized light in: linear polarized light out, one stop reduction.

Linear polarized light in: linear polarized light out, one stop reduction for all angles of the polarizer, the plane of polarization may be rotated.

Circular polarized light in: linear polarized light out, either no reduction or total reduction, depending on the sense of rotation (ie left-rotating light may pass, right-rotating light may be blocked, or vice-versa depending on how the filter was assembled).

Best,
Helen