Please Explain My Polarized Light Experiments

[Hornbein]

I wanted to filter out reflections from glass. So I bought a camera with a "circular polarizing filter." It filters out polarized light, adjustable for orientation. The result confuses some cameras, so it also has a second stage which induces circular polarization.

The results were disappointing. Most reflections from glass were not filtered out at all. What WAS filterable was glare or shine, as in shiny objects. The flatter and smoother the object, the better is the glare filtered out. I wanted to figure out what the difference was.

What about reflections from a mirror? That's about as flat as you can get. No filtering occurred. The result was identical no matter the orientation of the filter. It seemed to me that any image, even of the Sun, will not be filtered out. It has to be the sort of nebulous glare in which no image can be seen. I would like to understand this.

By the way, it is possible to use the filter backward. The light goes through the circular polarizer first then the linear filter second. A backward filter has no effect at all on simple glare.

Next I tried messing around with the light from a laptop. The filter is able to shut down the light completely. Then I tried the filter backward. It has no effect on the intensity of the light. Instead I got a very clear image of Haidinger's brush. It rotated as the filter rotated. As the filter rotates though the background changes from reddish to blueish. As usual the camera cannot capture and record the brush.

Next I tried reflecting the image from the laptop from a mirror. Here is a video I made of the results. Amazing, yes? The reflected image has lines of colors of the rainbow superimposed upon it, but the original does not. Note that the effect isn't a rainbow -- the colors are in a different order. As the filter rotates the colors reverse. I also tried it with the filter backward. It looked the same.

All this is far beyond my basic understanding of such things. Any help?

 

[FactChecker]

It will only filter out the polarized glare, not unpolarized reflected light. It sounds like that is what you are describing.

 

[tech99]

Summary:: Some very simple home experiments with "polarizing filters." I don't understand the results.

I wanted to filter out reflections from glass. So I bought a camera with a "circular polarizing filter." It filters out polarized light, adjustable for orientation. The result confuses some cameras, so it also has a second stage which induces circular polarization.

The results were disappointing. Most reflections from glass were not filtered out at all. What WAS filterable was glare or shine, as in shiny objects. The flatter and smoother the object, the better is the glare filtered out. I wanted to figure out what the difference was.

What about reflections from a mirror? That's about as flat as you can get. No filtering occurred. The result was identical no matter the orientation of the filter. It seemed to me that any image, even of the Sun, will not be filtered out. It has to be the sort of nebulous glare in which no image can be seen. I would like to understand this.

Well to make a start on the many questions...
My understanding of the camera is that it has a linear polariser first, which produces ther artistic efects, and then a circular polariser which then destroys the polarisation so as not to upset the camera electronics.
With glass the polarisation depends on the angle of reflection, being most at the Brewster Angle, which is approx 56 degrees to the normal. It is possible you were looking at a small angle to the normal, which does not create strong polarisation. On the other hand, when you say glare, I suspect you are looking at a greater angle to the normal, where polarisation is greater.
When you look at reflections from a mirror, it is probably reflection from a metalised surface, which does not produce polarisation.

 

[tech99]

Re the laptop observations, not absolutely sure but it sounds as if your laptop has linear polarisation. You were able to nul this out with the camera analyser used the right way round when the linear filter is in front.
When you reverse the camera filter, the linear polarised light from the laptop passes through the quarter wave plate is turned into CP. You can then rotate the linear polariser and obtain any plane of polatisation, so rotating the brush.
When you view the laptop using a mirror and the camera filter reversed, it sounds as if the mirror is birefringent, and is generating the opposite hand CP to the camera filter. As the cancellation is critically phase dependent, we find that the result is wavelength dependent, and so birefringent materils give colour bands due to small variations in phase caused by stresses in the material. But I am not sure of the exact sequence occurring in your camera filter.

 

[Hornbein]

Re the laptop observations, not absolutely sure but it sounds as if your laptop has linear polarisation. You were able to nul this out with the camera analyser used the right way round when the linear filter is in front.
When you reverse the camera filter, the linear polarised light from the laptop passes through the quarter wave plate is turned into CP. You can then rotate the linear polariser and obtain any plane of polatisation, so rotating the brush.
When you view the laptop using a mirror and the camera filter reversed, it sounds as if the mirror is birefringent, and is generating the opposite hand CP to the camera filter. As the cancellation is critically phase dependent, we find that the result is wavelength dependent, and so birefringent materils give colour bands due to small variations in phase caused by stresses in the material. But I am not sure of the exact sequence occurring in your camera filter.

Seems to me you have explained everything except one case. That is viewing the laptop using a mirror and the camera filter NOT reversed. Indeed that's how I made the video, as this is much more convenient than using the reversed filter. (Just screw the filter onto the camera.) Why do I get the color bands then?

 

[tech99]

It seems as if for this case the laptop is producing linearly polarised light and the camera is removing one component of it. I presume this rejection process is phase sensitive when the light passes through the mirror, which seems to be acting as a birefringent medium.This phase sensitivity makes the arrangement wavelength dependent and gives the various colours. But there is an element of mystery to your observations.

 

[Andy Resnick]

Summary:: Some very simple home experiments with "polarizing filters." I don't understand the results.
<snip>

All this is far beyond my basic understanding of such things. Any help?

Lots of interesting optics here.

First- that's a MASSIVE polarizing filter. For what lens is this used?

A circular polarizer is a 2-layer sandwich of a linear polarizer and a 1/4-wave plate oriented with the fast axis 45 degrees to the pass axis of the polarizer- this converts the linearly polarized light into circularly polarized light, so as you rotate the filter to achieve the desired result, the light entering the camera (and light sensor) doesn't change. This is the same principle behind current 3D movie glasses- you want to ensure left-right splitting occurs even if you tilt your head.

Ok- so 'reflections' are not singled out for cancellation by this device (or any simple linear polarizer)- as was pointed out, only one of the polarized components is canceled out- hopefully the one associated with glare.

Now the computer monitor: the light output from the screen is linearly polarized by construction- that's how flat screens work. Your result when holding the polarizer 'normally' is explained by this.

However, when reversing your polarizer, the light output from the screen is first converted into elliptically polarized light (since circular polarization only occurs for two specific relative angles between the polarization direction and the fast axis of the 1/4-wave plate). Where does the pattern come from? From back reflection of the light from the linear polarizer once more through the 1/4 wave plate, and then to your camera/eye. The pattern is difficult to describe analytically, but arises from interference between the two paths, similar to thin film interference. Complicating this simple model is the non-plane wave aspect of screen output.

If you look through the reversed polarizer at the screen and rotate your polarizer, you may notice gradual color shifting from green to purple. I think that results from chromatic aberration (meaning the 1/4-wave plate is not achromatic but has slightly variable retardance with wavelength) but I'm not sure.

Does this help?

 

[Orthoceras]

Tech99 has stated twice that the mirror is apparently birefringent. That is odd for a household mirror. OP, could you test and confirm that your mirror is birefringent? A very simple test is shown in the video below. If you are looking through the circular polarizer, held backwards, at the mirror image of the filter, the image will appear black if the mirror is plain. It will not be black if the mirror is birefringent.

 

[Hornbein]

Tech99 has stated twice that the mirror is apparently birefringent. That is odd for a household mirror. OP, could you test and confirm that your mirror is birefringent? A very simple test is shown in the video below. If you are looking through the circular polarizer, held backwards, at the mirror image of the filter, the image will appear black if the mirror is plain. It will not be black if the mirror is birefringent.

It looks the same, forward or backward. So the mirror is birefringent. It's a simple mirror of the cheapest kind. I'm in Japan, if that makes any difference.

 

[Hornbein]

Lots of interesting optics here.

First- that's a MASSIVE polarizing filter. For what lens is this used?

? Have you confused the circular mirror with the polarizing filter? The polarizing filter is attached to the video camera so it isn't directly visible in the vid.

If you look through the reversed polarizer at the screen and rotate your polarizer, you may notice gradual color shifting from green to purple. I think that results from chromatic aberration (meaning the 1/4-wave plate is not achromatic but has slightly variable retardance with wavelength) but I'm not sure.

Actually I got orange and blue with no intermediate colors. Just a tint though. Mostly white light.

 

[Andy Resnick]

? Have you confused the circular mirror with the polarizing filter? The polarizing filter is attached to the video camera so it isn't directly visible in the vid.

Ah. That makes more sense.

Actually I got orange and blue with no intermediate colors. Just a tint though. Mostly white light.

That sounds about right.

 

[Orthoceras]

Next I tried reflecting the image from the laptop from a mirror. .. Amazing, yes? The reflected image has lines of colors of the rainbow superimposed upon it, but the original does not. Note that the effect isn't a rainbow -- the colors are in a different order. As the filter rotates the colors reverse. I also tried it with the filter backward. It looked the same.

All this is far beyond my basic understanding of such things. Any help?

What you have been doing is conoscopy, a visual method to determine the birefringent properties (the optic axes and θ) of the biaxial material of your mirror. In the drawing, the plate is your mirror. If you, as observer, would have been closer to the mirror, you would have seen the other melatope too. The melatopes are the points where the optic axes intersect the plate. You would need an additional tool (a retardation wedge) to determine which axis is the slow one, and which is the fast one. Using the polarizing filter backwards should make some difference. As the filter rotates the colors should not reverse, and no isogyres should be present, if the filter is backwards. I hope I am not mistaking here. Conoscopy is a specialism for mineralogists.

 

[Hornbein]

I gave this another try. To get the bullseye pattern is a bit tricky. You have to orient the mirror in the somewhat unusual angle seen in the video.

I saw no sign of the double bullseye pattern. Single bullseye only.

I was mistaken in that the colors ARE in the order of the rainbow. The center is violet, outward to red, then violet again, etc. As the forward polarizing filter rotates, the colors rotate around the bullseye.

The view through the backward polarizing filter looks quite the same. The colors however do not rotate along with the filter. There is a blue tint then an orange tint, but not much change other than that.