Polarized Light -- Idea for headlight safety

[Johnnyallen]

I've been a semi-pro photographer for a long, long time. I know the value of having a polarizing filter in my camera bag. I also wear polarized sunglasses for driving during the day. You can't beat 'em.
I know some people who are complaining about these LED head lights from oncoming cars blinding them at night.
How about polarizing the light from the headlights and having an opposing or canceling filter on the windshield. Or windscreen for our British friends.
Two questions: would this work? would this actually cancel out the light from the drivers own headlights and not illuminate the highway?

 

[anorlunda]

Yes it could work. You must be careful however to only block a portion of those headlights. Otherwise, you might not see the car at all.

I think the difficulty is convincing the manufacturers. Any polarizing filter reduces the amount of light somewhat. When manufacturers brag, "maximum possible lumens on the road" as a sales feature, they don't want to give it up.

For old folks like me, the best solution to glare when driving is cataract surgery. I use a yellow colored filter on the visor at night. It helps.

 

[sophiecentaur]

For linear polarisation , the cancellation would not be complete unless the cars were on a level straight road and a polariser on the front of headlamps would reduce the light flux by half, which could be a problem. The light thrown back by random angled plane objects could have its polarisation altered and that would further reduce the light admitted by the windscreen, making objects less visible. I was trying to work out whether circular polarisation might do the job better.
I guess it's time that headlamps were all engineered so that they could identify oncoming cars and dip / deflect appropriately. That would be best achieved if headlamps were all modulated so as to be recognised as headlamps and not just bright roadside lights (not hard to do`). I believe there already systems on up market cars that do some automatic dipping but I couldn't quote a source for that.

 

[Tom.G]

https://www.popularmechanics.com/ca...bile-and-the-first-headlight-dimmer-16301766/

above found with: https://www.google.com/search?&q=automatic+headlight+dimmer
It even found how to build your own! And there are some on ebay for those that want to experiment.

Cheers,
Tom

 

[Johnnyallen]

Yes it could work. You must be careful however to only block a portion of those headlights. Otherwise, you might not see the car at all.

I think the difficulty is convincing the manufacturers. Any polarizing filter reduces the amount of light somewhat. When manufacturers brag, "maximum possible lumens on the road" as a sales feature, they don't want to give it up.

For old folks like me, the best solution to glare when driving is cataract surgery. I use a yellow colored filter on the visor at night. It helps.

I had to have cataract surgery over a year ago. Big difference.
I also didn't sort of take into account the fact that a polarizing filter is also a neutral density filter for you photography bugs out there. This would actually act as a tinted window. That could be counterproductive.

 

[Bystander]

these LED head lights from oncoming cars blinding

LED? Or, quartz halogen? Blue light IS blinding, scatters more, something.

 

[Johnnyallen]

For linear polarisation , the cancellation would not be complete unless the cars were on a level straight road and a polariser on the front of headlamps would reduce the light flux by half, which could be a problem. The light thrown back by random angled plane objects could have its polarisation altered and that would further reduce the light admitted by the windscreen, making objects less visible. I was trying to work out whether circular polarisation might do the job better.
I guess it's time that headlamps were all engineered so that they could identify oncoming cars and dip / deflect appropriately. That would be best achieved if headlamps were all modulated so as to be recognised as headlamps and not just bright roadside lights (not hard to do`). I believe there already systems on up market cars that do some automatic dipping but I couldn't quote a source for that.

I'm going to do a little research on the difference between linear and circular polarizing.
When I made the plunge to a Nikon DSLR about three years ago, I was told to buy a circular polarizing filter because my old linear filter (which I had no idea that it was) would render the autofocus non-functioning.
I mentioned to anorlunda that, I also didn't sort of take into account the fact that a polarizing filter is also a neutral density filter for you photography bugs out there. This would actually act as a tinted window. That could be counterproductive.
And...
I remember as a kid that Cadillac had a high beam dimming feature, and Googling it, I see they they now call it Intellibeam. However, this is only for going from high beam to low beam (and vice versa) and is not addressing my original issue.
If you're interested: http://gmauthority.com/blog/2014/11/cadillacs-intellibeam-system-automatically-enables-high-beams-feature-spotlight/

 

[sophiecentaur]

None of this is going to help my problem. I live in Essex UK and a vast number of cars are actually missing one headlamp and the other one is wonky. I assume the drivers are all busy using their mobile phones hand-held at the same time as they throw empty beer cans out of the window. Oh - that makes no hand free to steer the damned thing!

 

[anorlunda]

None of this is going to help my problem. I live in Essex UK and a vast number of cars are actually missing one headlamp and the other one is wonky. I assume the drivers are all busy using their mobile phones hand-held at the same time as they throw empty beer cans out of the window. Oh - that makes no hand free to steer the damned thing!

I wonder if you play the same kids road game in the UK. In the USA and Canada, we tell kids to watch for a car with one headlight. When they see it, they yell "peh-diddle". A one taillight car is a "peh-dunk".

It's even in Wikipedia https://en.wikipedia.org/wiki/Padiddle

 

[Klystron]

Circular polarization works as a light filter; my experience was with old protective lenses. Cylinders with two grids. Rotate one lense to inhibit, rotate back to increase light. Circular could solve some of listed problems but could introduce its own such as chromatic distortions and moire patterns depending on the car windshield design while also dimming incoming light. Designing a circularly polarized lense for the headlight seems doable, but the windshield design could be tricky.

 

[FactChecker]

When I made the plunge to a Nikon DSLR about three years ago, I was told to buy a circular polarizing filter because my old linear filter (which I had no idea that it was) would render the autofocus non-functioning.

My understanding of the photography circular polarizing filter is that it has two parts. The first part is a regular linearly polarizing filter to selectively cut glare in the orientation that it is turned. The second part then turns the linearly polarized light from the first part into circularly polarized light that the autofocus can work with. They definitely behave like the old filters that changed which light they let through as they were turned. I do not think that is the same as the circular polarization that has been mentioned here.

PS. This Wikipedia article "confirms" the above, but it might be where I originally got the impression long ago. So it may not be an independent confirmation. https://en.wikipedia.org/wiki/Polarizing_filter_(photography)

 

[Tom.G]

IIRC, circularly polarized light changes handedness upon reflection, therefore the approach fails.

• Headlamp Right-circularly polarized
• Reflected light is Left-circularly polarized
• ∴ WIndshield must be Left-circularly polarized to see what you illuminated
• Opposing traffic has same configuration, Right-circularly polarized headlamp
• However, you are looking at the opposing lightbeam from the other end, making it appear to you as Left-circularly polarized

Does that mean your windshield has to be a "leaky" (non-ideal) polarizer?

Cheers,
Tom

 

[sophiecentaur]

On the face of it, a system based on Polarisation seems elegant and desirable but, however you use polarisation discrimination, you lose half power or worse and, unless the driver wears goggles, you need a good quality polarising sheet in the windscreen. It would have to be pretty uniform if annoying patterns are not to be formed.
The whole idea would have been (and was) considered in the past but things are different nowadays. Cars are pretty intelligent and they all could have steerable lamps and they could easily spot an oncoming car and dip a well profiled beam to avoid the driver's face. Or how about some optics to announce the actual position of the driver's face itself.
For a really good system, it would be important to discriminate between an oncoming car and a bright roadside light so car lamps would need a marker / modulation system to announce themselves. Pretty trivial in principle but there would need to be good standardisation; that could be a serious hurdle as usual. And then there's the old COMPATIBILITY issue which would demand a certain level of protection for all vehicles (in both directions. of course).

I also had an idea of an autonomous system which could blank out bright lights from a driver's visual field but leave the rest of the road fully visible. Goggles would probably be the cheapest way to achieve this.

 

[anorlunda]

How did circular polarization enter this thread? I thought that polarized sunglasses try to block horizontal polarization, because glare reflected from water or highways is horizontally polarized.

And chromatic distortion should be the goal, because the blue content of glare is the most blinding. I use a device like the one in the picture. I think it is not polarized. Although the yellow filter necessarily reduces total light transmission, it reduces the blue more so that perceived glare is reduced so that the net effect is not counter-productive.

 

[sophiecentaur]

How did circular polarization enter this thread?

I introduced it because systems that use linear polarisation are very angle-critical. This is why (despite how it's described in the noddy books) 3D cinema uses crossed Circular Polarisers for R and L images. They allow people to 'loll' in their seats and still get isolation between L and R images. The same problem can arise with the headlamp system. (Natch, the polarisations in a linear system would need to be in diagonal planes.)
Of course, there are mechanisms all over the scene that can generate a cross polar image but tilting the head or the car would not be one of them if circular polarisation were to be used.
Sunglasses are a separate issue but they would probably not be used for night driving (except for cool dudes).

the blue content of glare

This is true and makes me wonder why so many vehicles use such lighting. My Landrover has halogen filaments in the main beam and Xenon in the dip. I don't understand why.

 

[Bystander]

I don't understand why.

People trying to be "cool."

 

[Klystron]

How did circular polarization enter this thread? I thought that polarized sunglasses try to block horizontal polarization, because glare reflected from water or highways is horizontally polarized.

And chromatic distortion should be the goal, because the blue content of glare is the most blinding.

Different polarization techniques seem inherent to the original question. Quality photography lenses and filters are aligned and carefully designed to eliminate unwanted optical aberrations as are eyeglasses. The OP specified headlight to windshield filtering methods at night.

Glare from sunlight and glare from intermittent headlights at night probably require different optimum filtering methods.

My ancient (2003) truck windshield (yes, a black Toyota pickup with full-cab assembled in California to that state's specifications) contains embedded grids of dark pigments (I'm guessing) designed to mitigate glare. When I bought the truck in Nevada (at night) the seller added polymer sheets to the interior side windows. Owner chooses filter levels from 1 (mildest) to 10 (almost black). After driving from the lot (dark night) and attempting a left turn, I returned the truck to the lot to replace the dark tinted panels with a lighter tint. State law, often flaunted, prohibits adding anything to the windshield except for rear-view mirrors.

 

[Janus]

The whole idea would have been (and was) considered in the past but things are different nowadays..

I have an old physics text, copyright 1936, which brings up the suggested idea of using polarization to reduce headlight glare. So, as you said, it is not a new idea. The fact that it never became common usage in the 80+ yrs since seems to speak to its practicality.

 

[Tom.G]

My Landrover has halogen filaments in the main beam and Xenon in the dip.

Are those Xenon lamps the Xenon filled incandescent type, or the HID arc discharge thru Xenon gas? I understand that Xenon incandescents are used to prolong filament lifetime without the high temperature needed for the halogen fill to be effective.

 

[sophiecentaur]

Are those Xenon lamps the Xenon filled incandescent type, or the HID arc discharge thru Xenon gas? I understand that Xenon incandescents are used to prolong filament lifetime without the high temperature needed for the halogen fill to be effective.

Afaik, they are only fancy versions of Halogen lamps. The only arc lamps for cars that I can find are very expensive and probably not appropriate for a bottom of range Landrover type. They are very blue, though - not suggesting a low filament temperature.

 

[Jon Richfield]

Probably the best is to wear polarising glasses, or fasten a polarising sheet to your windscreen, strategically placed. That would work dramatically for glare and for polarised headlamps. It also would help somewhat for ordinary car headlights (just move your head to dim troublesome oncoming light, polarised or not).

 

[Eric Bretschneider]

This has been tried over and over again.
The fact that it was suggested in the 1930s and it has yet to be implemented should tell you something. What do you think happens to the polarization when the headlights get dirty? scratches/scuff/haze on the outer optic?

All optical (scattering) elements that change the polarization of the emitted light. Once that is lost the whole system falls apart.

Consider it a beautiful idea slain by an ugly fact - while theoretically useful it is extraordinarily difficult to implement in practice.

 

[Klystron]

Consider consolidating new tech. Advanced aircraft have used HUD (heads up displays) and integrated visual/voice information systems for decades. Driver inputs come from multiple sources including gauges, cameras and EM sensors, audio/visual cues and alarms, radio broadcasts, and mirrors. Modern 'cockpit' designs filter, sort, and integrate driver information in an unobtrusive format where the driver literally "keeps their head up" while still receiving external and operational data.

Head worn apparatus -- glasses, goggles, half-lenses, helmet visors, 3-D headgear -- assist the driver but with inevitable trade offs. Human factor studies indicate an adaptation period required to understand the inputs. Many people are affected by vision enhancements including disorientation, loss of balance, even nausea [citation needed].

Placing and projecting external inputs into driver line-of-sight -- including glare mitigation whether specific polarization methods apply -- makes sense given a trend toward automating driving within integrated traffic control systems. Pardon, if I exceed thread scope.

 

[sophiecentaur]

Probably the best is to wear polarising glasses, or fasten a polarising sheet to your windscreen, strategically placed. That would work dramatically for glare and for polarised headlamps. It also would help somewhat for ordinary car headlights (just move your head to dim troublesome oncoming light, polarised or not).

Those comments do not address the details of the polarization required and how it could work. Sun specs could not be part of a working system.

 

[Tom.G]

Jon Richfield said:

Probably the best is to wear polarising glasses, or fasten a polarising sheet to your windscreen, strategically placed. That would work dramatically for glare and for polarised headlamps. It also would help somewhat for ordinary car headlights (just move your head to dim troublesome oncoming light, polarised or not).

Those comments do not address the details of the polarization required and how it could work. Sun specs could not be part of a working system.

You are right, that does not address headlight polarizarion. It does however suggest a solution to the underlying problem, that of headlight glare.

 

[FactChecker]

You are right, that does not address headlight polarizarion. It does however suggest a solution to the underlying problem, that of headlight glare.

Polarized glasses can reduce reflected glare from regular headlights off of wet road surfaces and car metal. The reflected light is partially polarized. It would not cut the unpolarized direct light from the headlight except as it reduces all light. But wearing polarized glasses at night makes everything too dark in general.

 

[Jon Richfield]

This has been tried over and over again.
The fact that it was suggested in the 1930s and it has yet to be implemented should tell you something. What do you think happens to the polarization when the headlights get dirty? scratches/scuff/haze on the outer optic?.

I have been aware of that since the 1960s (wasn't around in the 1930s, sorry, I bow to your seniority )

That ugly fact is slain by the beautiful fact that polarisation is not all-or nothing. If I have half of my incoming light cut out by polarisation, that is pretty good going and probably more than I need. Notionally polarisers should cut out 100%, but in practice won't happen for many reasons, including the fact that the cars aren't bolted to a stationary lab bench, and that is good, because we don't WANT to render oncoming traffic blackly invisible. So a bit of mist or dust between the polarisers will do no harm; in fact they will block more than they will let through the polarised light by rotating it.

All optical (scattering) elements that change the polarization of the emitted light. Once that is lost the whole system falls apart.

Nothing of the kind; how much have you experimented? Not much, to guess from your assertion...? Try it! Set up two polarisers at right angles and blow some dust or mist or smoke between them and see how much light gets through; THAT is what you call the whole system falling apart? Some people are very hard to please...

And even if the oncoming traffic has no polarisers, polarising screens remove a proportion of incoming light, polarised or not. (Cos is not only lettuce hm?)

Consider it a beautiful idea slain by an ugly fact - while theoretically useful it is extraordinarily difficult to implement in practice

The really ugly fact is that at night one does not want to lose much light except oncoming headlights, so a fully polarising windscreen could be a menace. That is why I proposed putting only a patch of polariser on the windscreen, for the driver to duck behind to dull bright headlights or sun on the horizon.

 

[FactChecker]

That is why I proposed putting only a patch of polariser on the windscreen, for the driver to duck behind to dull bright headlights or sun on the horizon.

I think that I would rather put a dot of filter on glasses that can cover the headlights. It would be easier to move my head a little and wouldn't have to be polarized. But there are often many cars and headlights.

 

[Eric Bretschneider]

Why should I repeat work (testing degradation of polarization due to dirt/haze) when it’s already been done. Just consider the situation.

There are much more elegant solutions. Work has been done on DMD projection systems for headlights. Coupled with a camera system, the pixels corresponding to the head location of a driver in an oncoming vehicle can be dimmed or even shut off completely. This can also be done for pedestrians. No visible light = no glare.

More advanced implementations can project around falling snowflakes to improve visibility.

 

[cjl]

Afaik, they are only fancy versions of Halogen lamps. The only arc lamps for cars that I can find are very expensive and probably not appropriate for a bottom of range Landrover type. They are very blue, though - not suggesting a low filament temperature.

Xenon HID lamps for cars aren't a true xenon short-arc bulb, but they also aren't incandescent lamps or halogens. They're metal halide gas discharge lamps, with a bit of xenon to boost startup intensity. They're used in cars because they have substantially better luminous efficiency than halogens, provide a spectrum that better approximates daylight, and are significantly closer to a point source than filaments, allowing better beam control, a more uniform illumination pattern, and a sharp cutoff. Properly adjusted OEM-spec xenon projector headlights shouldn't cause glare issues, since they have optics to provide a very hard cutoff above which almost zero light is projected. Unfortunately, since these did come out on high-end cars first, there are a number of cheap ebay-spec headlight "upgrade" kits that are much worse, and extremely prone to glare. In most cases, when you see cars that are really bad, it's either because they're running a kit like this or because their headlights are way out of adjustment.

 

[sophiecentaur]

Xenon HID lamps for cars aren't a true xenon short-arc bulb, but they also aren't incandescent lamps or halogens. They're metal halide gas discharge lamps, with a bit of xenon to boost startup intensity.

I'm a bit confused by that. The Wiki article says that they take a warm up time of a few minutes. Mine are on instantly (or at least very bright at the start). Wiki says that automotive bulbs are metal halide gas discharge. I gave up looking because nearly all the hits are sites to sell you replacement bulbs.
Also, when I enter the Reg Number of my car, everyone claims that dip and main are the same type. They are not - one's blue and the other's hello halogen colour

 

[Sherwood Botsford]

You can put the same number of lumens on the road by jacking up the lamp power before it gets filtered. This takes a rule change in max wattage for headlamps. Given the difference between tungsten and LED efficiencies, it should probably be defined in terms of illumination anyway.

I think reflection from a non-metalic surfaces randomizes polarization.

So if random, it will be in effect rotated 45% -- half way -- So you would get back 1/sqrt(2) of the illumination -- a drop of about 30%

But the opposing headlights would be blocked close to completely. So the signal to noise ratio would be markedly improved. There would be some glow from forward scatter reflection of the opposing car's headlights, which would reveal the car's location even if headlight dimming were perfect, and would help you see the road.

 

[FactChecker]

I think reflection from a non-metalic surfaces randomizes polarization.

My experience is with the camera polarizing filter. There is a noticeable reduction of glare reflecting off of leaves, glass, water, etc. when the filter is turned correctly.

 

[sophiecentaur]

My experience is with the camera polarizing filter. There is a noticeable reduction of glare reflecting off of leaves, glass, water, etc. when the filter is turned correctly.

I would be very surprised if linear polarisation (as in camera filters) were used for this sort of project. Cancellation is very much affected by tilt when linear polarisation is used. That's the reason for using circular polarisation for stereo cinema and I think all the same advantages would apply for the headlight trick. However, it is all beginning to appear less and less attractive to me.
The problem with a system like the one being discussed requires everyone to be taking part and could not increase illumination above the present legal cut off angles. I think we would have to assume that the majority of vehicles would initially be using the standard dip arrangement. A more intelligent system, not involving polarisation at all, would allow the driver of a modified car to have better aimed headlamps, protecting any oncoming driver yet giving much better illumination of the parts of the road at higher angles and which are so hard to see at present. Since the early proposals for polarised systems, things have changed so much and the dynamically pointed and profiled beams are well within existing capabilities. The snag would be with regulations and specifications rather than the technogoly, I think.

 

[Klystron]

Why should I repeat work (testing degradation of polarization due to dirt/haze) when it’s already been done. Just consider the situation.

There are much more elegant solutions. Work has been done on DMD projection systems for headlights. Coupled with a camera system, the pixels corresponding to the head location of a driver in an oncoming vehicle can be dimmed or even shut off completely. This can also be done for pedestrians. No visible light = no glare.

More advanced implementations can project around falling snowflakes to improve visibility.

Google found this pdf. DMD (digital micro-mirror device) begins on page 2.
http://www.utc.ices.cmu.edu/utc/TNCCRKN-ECCV14.pdf