If I send an artificial light source’s light through a linear polarizer, and that light reflects off a dielectric surface at a narrow angle, is the polarization angle maintained (in this case its linearity) or is the angle of all the light randomized?

I ask because I am working with cross polarization photography and I am wanting to understand conceptually what I am seeing happen before me.

Under cross polarization lighting emitting from a ring flash mounted around the camera, the object being photographed has mostly diffuse reflections which tells me that the linearly polarized light I am shining on it had to have had its polarization angle randomized or changed in some way otherwise all light would be filtered out prior to hitting my camera sensor.

Since there are no specular highlights visible, but plenty of diffuse reflections, I am left wondering the relationship of specular reflections and polarization.

Via Wikipedia I understand specular reflections as light reflecting along the same angle to the viewer, giving the appearance of increased intensity, as opposed to scattering randomly. However I also read about brewster’s angle which causes reflected light to be horizontally polarized, which we see as glare. Is the glare just a specular reflection and it’s polarization state is irrelevant to our viewing of it sans sunglasses, or is there some inherent relationship between specular reflections and linear polarization?

It’s hard to separate the two concepts while reading existing articles as they’re talking about it from the perspective of filtering linear light, so I appreciate your reading, and my apologies in advance for asking two questions in one.

P.S I read all the similar posts prior to submitting


2 Answers 2


Not sure if I get your question exactly, but I think you are asking two questions.

  1. Will the rule of the angle of reflection equal the angle of incidence change with polarization? Note that this is the definition of specular reflection.

That answer is no, assuming a specular reflection. The intensity of the light that is reflected off the surface will change with the polarization, the angle of the reflection will not.

  1. Will the polarization change after it bounces off a surface?

That is a little more complicated but often yes.

If it is a metal surface the answer is and it has to do with the boundary condition that the metal is a conductor. Since the electric field in the metal is zero (because the charges can rearrange to cancel the field out) the electric field of the reflected light wave has to change.

If it is a dielectric like glass, then it depends if you are going from glass to air, or from air to glass and depends on the angle. Again it depends on the boundary conditions.

What if it is not a smooth surface and you have a diffuse reflection, or if is something like air with dust in it, or water with particles in it. The it is a lot more complicated and the technical term is scattering and it depends on the size of the particles and the wavelength of the light.

An example of scattering changing polarization is the sky. If you are careful you can even tell there is some polarization by eye - Haidinger's brush

All surfaces, even pretty rough ones can have a high reflectivity at large angles. When at Brewsters angle one polarization is not reflected and the other polarization has some reflection, so the reflected light is more strongly polarized than the incident light before the reflection, so the polarized sunglasses with the polarizer in the correct orientation cut down the glare.

You might like the hyper physics section on light and vision

There are lots of very cheap, used versions of the second edition of Eugene Hect's book "optics". It has a lot of good diagrams and explains a lot. It also talks about polarization from a pretty practical point of view.

If you are doing cross polarization photography then the optical path length (the refractive index times the distance) also matters so keep that in mind. Stress birefringence occurs when the refractive index changes a little bit due to the stress.

Polarization can be one of the trickier part of optics, especially with some crystals, or specialty devices. In addition to linear and horizontal polarization you have right left circular or elliptical polarization etc.


If I send an artificial light source’s light through a linear polarizer, and that light reflects off a dielectric surface at a narrow angle, is the polarization angle maintained (in this case its linearity) or is the angle of all the light randomized?

No, generally the absolute polarization angle isn't maintained.

Consider for example, light incident at a shallow angle from the air onto the dielectric surface where the light is polarized at 45$^\circ$ relative the surface. This can then be considered a combination of light of equal intensity both perpendicular and parallel to the surface. Since each of these will be reflected with different intensity, the polarization angle will change.

Only in the case where the light is polarized exactly parallel or perpendicular to the surface will the polarization not change, and in random reflections within a room, that's not going to happen very often.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.