Polarized vehicle headlights.

Oh dear. Does this mean you wouldn't see cars headlights coming straight at you?

Cross polerised filters render white light as purple (violet?) then there is road camber and vehicle roll to make the effect imperfect.

No. Polarizers are not 100% efficient, so you would still see a bright spot - but *most* of it would be blocked, so you wouldn't be blinded either.

Two problems though:

- Polarizers cut out about half the light. - That means the headlamps are only half as bright. - It also means that the windscreen cuts out about half the reflected light. - Net result: Your beam headlight needs to be 4x as bright! (Actually, LEDs might be able to emit polarized light directly, rather than blocking light of the wrong polarization. Still have x2.)

- The killer problem though is legacy. If I am driving towards you with my new car with polarized headlamps, how do I know I don't need to dip my headlamps?

How much ND filtering is involved in a typical pimpmobile windscreen?

I imagine one of the problems is that the polariser on the windscreen will cut out quite a lot of light, so it will be harder to see reflections off distant objects - so you need even brighter headlights to compensate. All other (unpolarised) lights would need to be brighter as well to compensate for losses in the polarising filter on the windscreen.

Actually this would not necessarily be a good idea. Sometimes it is better to have constant illumination of the road ahead from lots of lights dotted along the road, rather than rely on lights projecting forwards from you car, which illuminate by inverse square law: objects close too will be very bright and objects further away will be less bright. Providing that glare from overhead lights as you pass each one isn't a problem, I prefer constant illumination to light which is less good on distant objects.

No, there is always some leakage. It was being worked on in 1958 by some colleagues.

Not to mention side lights that are unpolarised.

Yes, but this was thought about soooo long ago, when there were far fewer vehicles on the road. We shouldn't put off really good ideas indefinitely.

Does inverse square apply to a directed beam of light? You could have multiple light sources on the car, each highly directional.

Hmmm. You *may* have a point there. My knowledge of optics isn't good enough to know.

There are other road users than cars, some don't have windscreens to polarize.

If its legal, not much. They have to transmit 96% IIRC.

Polaroid visors on crash helmets? :-) Polaroid sunglasses for other road users such as pedestrians and cyclists? :-)

Do circular and linear polarising filters reduce the light intensity by about the same amount, or is one better than the other?

I am still awaiting a prescription windscreen so that I don't need to wear glasses.

Reflections (from shiny cars, puddles etc) rotating the polarisation and causing dazzle?

Umm.. I have read bits of Haldane and concur on the politics.

If you removed 50% of the light energy from a filament source, aren't you also going to gather 50% as heat at the filter?

I thought it was nearer 70%, however it's some years since I looked into this. One state in the US won't allow tinted front windscreens at all if IRC. There is also the tilt factor of the windscreen relative to the source which also comes into this. I had a bad experience with a coated winscreen before Xmas where the in car wireless RFID toll tag would only work on the LHS of the mirror, even though the coating was supposedly not present all over. It was a pig to remove the tag, even with a hair dryer and even then only 70% of the glue transferred.

Good for him :-)

Although he did seem to have a very strange view about what went on in the USSR . . .

Yes, but the visible light power is a small fraction of the total input power in an incandescent bulk (about 1% for a torch bulb, should be a bit better for halogen).