Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

America's Drivers Agree: LED Headlights Are Just Too Bright

wsj.com

Post reply on HN
Print view View original

51–60 of 98 posts

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#51

Post by alrs »

sherr wrote: post #10

In addition to cars, I'm a cyclist and have noticed that LED cycle lights, both front and back, can also be far too bright sometimes. Blinding in fact. A few years ago, I was behind a bike which had such a high powered rear (red) LED light that I had to drop back to get away from it. It reminded me of the sort of powerful red light an airport might have. For bikes, it sometimes felt like an "arms race" in brightness.

It's not that bike lights are too bright, it's that they use garbage optics in countries that allow it. US brands have a bias towards off-road mountain bike lights that spray light everywhere. Look instead for bicycle lights that meet the German StVZO standard. Busch and Mueller stuff is good, you can also find generic stuff on AliExpress all day long.

https://www.bikeradar.com/advice/buyers-guides/stvzo-bike-li...

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#52

Post by oarsinsync »

clort wrote: post #42
simonbarker87 wrote: post #14

I think UK drivers also agree. It’s like lots of people are driving around with their high beams on. They aren’t because if they were you’d wouldn’t be able to look at them at all. They are 20% too bright and all too cold.

One issue I noticed when I rented a newer car a couple years ago, was that it automatically switched to full-beam and dipped where it thought appropriate though I only switched on low beam in the first place. I was driving a winding country road with a fair amount of oncoming traffic so I didn't really need full beam. The problem is that it only dips when it has seen the oncoming headlight, whereas to avoid dazzling…

I had this experience with an Opel (Vauxhall) recently. The solution was to manually deactivate the high beams when they were automatically engaged. They stayed off after that. Don’t know about a permanent solution, this was a one week rental.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#53

Post by Borborygymus »

I through it would be a nice idea to have lights polarised vertically and screens polarized horizontally... scattered light gets de-polarized so object illuminated by the lights would still be visible through screen, but direct light from other people's headlights would get blocked.

Don't know how that would look practically (how bright can one really make a polaralized light source that is cost-effective / how expensive is it to create whole windscreens that are polarized / getting it rolled out would be a challenge).

Still it's a very anoying problem and it would be nice to find a decent solution.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#55

Post by throwaway11460 »

Lammy wrote: post #41
hiddencost wrote: post #32

Earlier quoted context omitted.

House lights are fine, you either have old ones or bought the wrong ones.

I'm still using my stash of incandescents, and halogens for the track lights, but I don't control my friends' houses :)

Check out the IKEA light bulbs, they are cheap and great.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#56

Post by userbinator »

usrusr wrote: post #39
userbinator wrote: post #15

IMHO it's not the brightness, it's the contrast --- a diffuse field of light at the same brightness wouldn't be quite as uncomfortable as the sharp contrast between light and dark they create, and this goes for the one whose car has the headlights and the one being shone at. In my experience, newer brighter lights also means a sharper cutoff so there's a very bright area immediately in front of the vehicle followed b…

But it's also brightness: legal limits for brightness above the horizon were set in an age when light designers weren't able to do very detailed control of light distribution and the dimmed area was always a mix of some bright spots and on a much darker average. The legal brightness limits were selected under the assumption that the highest allowed was an exceptional peak that would only occasionally hit an observer…

because for the driver behind the wheel it's certainly awesome

I disagree --- I have been in a newer car with bright LED headlights, and while visibility in the bright area is certainly great, it makes it much harder to see anything beyond that, because the eyes are adapted to that brightness. The effect of having things "pop" into view as they cross that threshold was actually quite disconcerting.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#57

Post by lmpdev »

I sell these all day at work.

Aftermarket, the blame largely lies on the end user incorrectly selecting and fitted LED modules, but I still see this problem far too frequently with cars coming off the production line.

I cannot tell if it's the business incentives around the engineers designing these boards, an actual lack of understanding of optics from them, or a failure in due-diligence from companies they supply to.

As a person formerly heading down a physics career: LUX AND LUMENS ARE COMPLETELY DIFFERENT MEASUREMENTS!

I am worried far too many companies are taking Lumen readings from incandesent bulbs they are trying to replace, looking on DigiKey and selecting whatever COB from Cree with the same amount. This is not how it works physically and especially not how it works for how humans percieve light!

I wish more work would be put into the proper testing of the diffusion of light. Yes diffusers exist from proper companies like 3M and yes they are used, but it to me seems beyond something you can just dump on an otherwise busy engineering team in isolation and expect good results. You need proper user testing and more importantly need the cost of diffusing methods to come down many order of magnitude per mm^2.

And I'm not even covering other considerations like colour rendering index optically or standardisation of techniques compliance wise.

It is an emerging area, but I can't help but be aggravated by the seeming lack of process. Good LED specialists understand how to make them look correct, the rest of the world just can't seem to yet.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#58

Post by pi-e-sigma »

stareatgoats wrote: post #49

Please, just invent eyeglasses that dynamically subdue oncoming headlights (of any type) already. Incredibly, this golden goose has not yet been plucked AFAICS (the tinted co-called "night-driving-glasses" don't work - or did I just not find the right brand yet?).

[dead]

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#59

Post by gadders »

cesnja wrote: post #11
sherr wrote: post #10

In addition to cars, I'm a cyclist and have noticed that LED cycle lights, both front and back, can also be far too bright sometimes. Blinding in fact. A few years ago, I was behind a bike which had such a high powered rear (red) LED light that I had to drop back to get away from it. It reminded me of the sort of powerful red light an airport might have. For bikes, it sometimes felt like an "arms race" in brightness.

This is sometimes the only way to make sure the drivers notice you.

It's a good way to get them to shine their main beam headlights in your eyes in retaliation as well.

Re: America's Drivers Agree: LED Headlights Are Just Too Bright

#60

Post by crustycoder »

throwaway11460 wrote: post #16
tmnvix wrote: post #12

Earlier quoted context omitted.

It's not only oncoming drivers that can be blinded.

On my car it works for bikes and people too.

On my car it works from behind other cars as well. There's a tree lined road near me, drive behind someone at night and there's a dark box around them, with the the trees being fully lit up.

And it also works with over-reflective road signs, they too get a dark box.

Post reply on HN