Two polarizers (stacked) as variable ND?

[David R Munson]

Anyone here ever tried stacking two polarizing filters to achieve the effect of having a variable-density ND filter? I'm curious to try it, but I'm wondering if there are pitfalls to this that I'm missing. I figure I could mark the filter rings and measure light transmission with a light meter to figure out how much light is being transmitted, and then keep a reference card to know where to set it in order to get the density I need.

Thoughts?

[Mark Woods]

You can stack polarizers and end up with something between a "regular" polarizer and something that is totally black. Some people use them to do fade outs in motion picture work by rotating them during the shot. Not many. But some do. I don't know if this helps or not.

MW

[Andrey Donchev]

It will be extremely hard if not impossible to measure the correct factor of this ND filter and determine the correct exposure if one is not using a focal plane measuring device like Sinar's.

[Eric Leppanen]

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

I vaguely recall someone saying that this product used polarization to achieve its effect, but I can't swear to it.

[Jason Greenberg Motamedi]

From the link Eric provided: "... The Vari-ND can be used with most other Singh-Ray filters except Polarizers...", so I think it pretty certain that this filter simply two stacked polarizing filters.

[David R Munson]

The Singh-Ray is cool, but I figure I can pick up a couple good polarizers in the filter size I need (which isn't 77mm) and get the same effect for a lot less money.

[Doremus Scudder]

David,

I imagine it would work adequately. I assume you are trying for really long exposures to remove moving objects or blur a lot. In this case, you can err on the side of overexposure and have a little wiggle room.

To get the best optical quality, buy coated filters.

I would use a spot meter to take a meter reading directly through the stacked polarizers in exactly the relationship and orientation (i.e. which point is "up") that they will be on the camera. This should get you well into the ballpark (unless there is so little light that you cannot get a reading at all).

Of course, reciprocity correction and contrast control will have to be dealt with, as in any long exposure, but you know that.

Good luck,

[Brian Ellis]

You wouldn't have a "variable" netutral density filter. The factor of a polarizer remains the same regardless of the position of the polarizer and the angle of the light. So what you would have are two neutral density filters and probably some major vignetting proglems.

[Jason Greenberg Motamedi]

Brian,

You are mistaken; put two polarizing filters together and spin one. You will see that it is in fact a "'variable' netutral density filter.' When the slits are parallel the filter factor would be about two stops (?), and when they are perpendicular the filters would cut out just about all light.

[Alan Davenport]

I think there is no question that the Singh-Ray is using two polarizers to achieve the variable effect. I suspect they also have a quarter wave plate -- same as in a circular polarizer -- but in this case the 1/4 wave plate is in front of the polarizer elements. That would remove the polarization from the incoming light, and make the polarizers act the same amount on all parts of the scene.

To duplicate this effect with discrete polarizers, you'd want to have a CP reversed, then another polarizer (preferably linear, circular OK) behind it. With separate filters you'll have more air/glass interfaces, which might result in a little bit less quality overall. (Whether the human eye could discern the difference, I won't guess...)

[Brian Ellis]

JG said: "You are mistaken; put two polarizing filters together and spin one. You will see that it is in fact a "'variable' netutral density filter.' When the slits are parallel the filter factor would be about two stops (?), and when they are perpendicular the filters would cut out just about all light"

"Because of the change in values as the polarizer is rotated, photographers frequently assume the exposure factor must be increased as the degree of polarization increases. This is not the case! When used at the non-polarizing angle the polarizer acts as a neutral density filter with a factor of about 2.5; at the maximum polarization the same factor applies. If we increased the factor as the degree of polarization increasd, the non polarized areas would be overexposed. The same factor applies regardless of the film or the light source." Ansel Adams, "The Negative," p. 114.

I don't offhand see why using two filters or five filters or ten filters would change this principle. But if you do perhaps you could explain.

[Oren Grad]

Brian, the filters interact. Take a look at this demo (http://www.colorado.edu/physics/2000/polarization/blocking_light.html).

[Jason Greenberg Motamedi]

Brian,

AA was writing about using one polarizing filter not two.

As Oren points out, they interact; Polarizing filters (the regular kind, not circular) are filters with very narrow slits running in one direction. When you stack two of them together, and arrange the slits so they are perpendicular, essentially all light is blocked.

[Brian Ellis]

I thought I already posted an admission of error here but I don't see it so I'll try again. I was wrong in saying they wouldn't be variable, I didn't appreciate the effect of stacking the polarizers.

[David Grenet]

To calculate the factor of this ND filter you can use this formula (from "Physics for Scientists and Engineers", Serway and Jewett):

I = I_max cos²(theta)

where I is the intensity of the light after passing though the filter

I_max is the intensity if there was only 1 polarizer (generally 0.5) and

theta is the angle between the axes of the polarizers

So theta = 0 (polarizers lined up) gives a factor of 1 stop, theta = 45 degrees gives 2 stops, etc