I’ve inadvertently (and advertently) been doing some experimentation with polarized light lately. Liquid Crystal Displays (LCD) typically emit linearly polarized light, and most decent sunglasses act as polarizing filters. This can cause some problems, if you happen to have some gadget whose display is inconveniently set to emit light with horizontal polarization — since reflected light tends to become polarized parallel to the surface from which it reflects, sunglasses are made to filter that light. But it also makes it tough to read any LCD that is oriented to emit that polarization.

There are ways around this, though. I’ve noticed that my iPod screen (unlike my GPS receiver and watch) doesn’t go black at any orientation of my sunglasses, though I do get some shifting rainbows on the screen. Here are two orthogonal orientations of a linear polarizer:

Here’s what happens when you rotate a polarizer in front of a normal LCD screen

What’s going on here is the cover to the iPod screen is birefringent — there are two optic axes, for each polarization, and the incident light is broken up into those two components. They have a slightly different index of refraction, and the larger index lags in phase. Since index depends on wavelength, the amount of phase delay will likewise change with wavelength, and will also depend on the thickness of the material. Some wavelengths will rotate such that will be completely out of phase and will destructively interfere, which gives you the colors.

If you choose your thickness properly, you can have one of the rays lag the other by 90 degrees, and this will result in circular polarization of the light, in what is called a quarter-wave retarder, or quarter wave plate. Since I observe that the intensity of the light from the iPod screen doesn’t noticeably change when I rotate it relative to my sunglasses, the light must be pretty close to being circularly polarized.

There a number of other applications for this; one of them is in the camera I used. An autofocus SLR camera sends its light into a prism, some of which goes to a light meter so the camera can adjust the exposure. But the amount of light that is transmitted or reflected is dependent on the incoming polarization — if this is adjusted by using a filter, the meter will be reading the wrong signal and under- or overexpose the picture. So filters for SLRs start with a linear polarizer and add a (nominally) quarter wave plate (but of course this will vary somewhat with wavelength) to give circularly polarized light, so that the transmission and reflection from the prism is the same as if the light were randomly polarized. This is why you see the warnings (you do see the warnings, don’t you?) about needing a circularly polarizing filter for these cameras.

If you have a polarizing filter, you can use an LCD as a polarized light source and view birefringent materials using it as a backdrop. Make the screen as white as possible, and rotate the polarizer until it blocks the light. Then place a birefringent material in front of the screen and look through the polarizer. Cheap clear plastic often will have stress-induced birefringence. I’ll post some examples soon.