The great EU lightbulb ban, 10 years on

[uppi]

Yeah, I did eventually do all that too. There are 850 nm LED flashlights for night vision devices that give off enough light that you can see a very short distance in front of them. I never did find any 780 nm ones - those would probably have been better given that they'd appear much brighter, but for that reason they wouldn't be popular for the sorts of purposes people would want NIR LEDs for. I've also messed around with lasers at various NIR wavelengths (780, 808, 830, 980) and had some fun that way.

Can you still see something at 850nm or are you seeing some shorter wavelength light that gets emitted by the diode? I did some experiments on my own, but they were limited by the 780nm laser diode which could only be tuned to slightly above 800nm. The visibilty seemed to be dropping off rather quickly beyond 800nm, but that might have been the laser emitting less power at the edge of its wavelength range.

780nm diodes are quite common. Because good laser diodes were needed for CD drives, a lot of optimization went into this wavelength, so that the cheapest and best NIR diodes are usually at 780nm. Maybe nobody bothered to sell them in a flashlight?

 

[rah]

Has anyone ever bought an incandescent light bulb for the purpose of heating?

For the easy bake oven, of course.

 

[Manfred Belheim]

Just bought some LED bulbs recently. At first glance they seem fine, but they're obviously flickering on and off at some high frequency which you can tell if you're moving anything around in the room at any sort of speed. Quite annoying. Not sure if this is a standard feature of LED bulb technology, or just the ones I got.

Also they generate a load of heat at the very base of the bulb, so might actually heat up sockets more than incandescent bulbs. Not sure if that's an issue either.

 

[civvver]

I hate cfls, they have to warm up to reach full brightness. But I also hate incandescent flood lamps for my recessed lights as they burn out super fast. I got warm leds, like 2700k, so they look soft white just like the old incandescent, and they haven't burned out in three years. Which is great since recessed in a 9 ft ceiling requires me to get the step ladder out of the garage to change. So overall thumbs up to leds. And they aren't expensive anymore, about $4-5 a bulb, which is probably the same price as incandescent of those type now that they're not being made. I think an 8 pack off amazon was over $20, and considering they last like 1/4th as long at best they cost more. I haven't seen any noticeable differences in my electric bills but I don't track usage and I'm sure rates have gone up anyway.

 

[Farm Boy]

LEDs are great in a lot of applications. People will/would have gravitated towards them without getting those awful cfls shoved at them or making it harder to get warm light bulbs in situations where warm light bulbs are the right tool for the job.

 

[rah]

Yes, once the price came down, but at first it was a hard sell for poor people. Long term savings were out of reach.

 

[Broken_Erika]

Just bought some LED bulbs recently. At first glance they seem fine, but they're obviously flickering on and off at some high frequency which you can tell if you're moving anything around in the room at any sort of speed. Quite annoying. Not sure if this is a standard feature of LED bulb technology, or just the ones I got.

Also they generate a load of heat at the very base of the bulb, so might actually heat up sockets more than incandescent bulbs. Not sure if that's an issue either.

Sounds suspiciously like cheap/defective crap to me, might want to try a different brand next time...

 

[civvver]

Yes, once the price came down, but at first it was a hard sell for poor people. Long term savings were out of reach.

Oh yeah even four years ago an led bulb was like $12. Now a couple bucks a piece for a regular sized one.

 

[Manfred Belheim]

Sounds suspiciously like cheap/defective crap to me, might want to try a different brand next time...

I doubt there will be a next time. I'll probably die before they need replacing.

 

[Farm Boy]

Sounds suspiciously like cheap/defective crap to me, might want to try a different brand next time...

They get legit hot on the heatsink. Railroad crossings have adaptor kits to spin the things around, mirror the light back out, and hook the now closer heatsink to the surface of the light to help melt off snow and ice. Well, at least I heard there were. Not sure if it became feasible/widespread. My attention span for railroad crossing lights isn't bottomless.

 

[FriendlyFire]

We went from 60W Incandescent to 15W florescent to 5W Led
With LED prices now being so cheap and more efficient. I've already made the switch to energy and money savings Leds there is no downside really. Iam glade that the EU lead the way pushing to ban Florescents and making Led affordable.

 

[rah]

I've replaced almost all to LEDs except for the one in my garage door opener. The vibrations from the door going up and down wears away the bulb contact on the base so I use the cheapest bulbs I can find in there since I have to change it frequently.

 

[Bootstoots]

Can you still see something at 850nm or are you seeing some shorter wavelength light that gets emitted by the diode? I did some experiments on my own, but they were limited by the 780nm laser diode which could only be tuned to slightly above 800nm. The visibilty seemed to be dropping off rather quickly beyond 800nm, but that might have been the laser emitting less power at the edge of its wavelength range.

780nm diodes are quite common. Because good laser diodes were needed for CD drives, a lot of optimization went into this wavelength, so that the cheapest and best NIR diodes are usually at 780nm. Maybe nobody bothered to sell them in a flashlight?

Now you've done it - I'm going to have to blabber on about all the lights I've shined at myself.

The LED wasn't emitting anything significant below the ~750nm of the real IR-pass filter I had (not the initial crude one). LEDs emit significant light for several tens of nm on either side of their peak, so I believe what would have happened is that the LED's spectrum and the eye's sensitivity would have multiplied to create a peak sensitivity of around 800 nm or thereabouts. Laser spectra of course are far narrower. At 830 nm a faint spot was still visible, although I can't remember what the output power was. An 808 nm spot is easily visible from a 200 mW laser.

At even more extreme wavelengths, light sources can still be perceived if you stare directly at them. A 940 nm remote control LED is just barely visible under perfect conditions; peak eye response multiplied by its spectrum is probably something like 890 nm. I first got interested in this after noticing I could see the flashing remote control LED in a dark environment.

I also did something very dangerous - I got a 980 nm, <5 mW laser diode and looked into it directly. There was a central red point of light surrounded by a cyan glow. The red light remained but the cyan light was filtered out when I put my 750 nm IR-pass filter between the laser and my eye, so it appears that I really was seeing the 980 nm light, and some nonlinear optic effect in the laser resulted in faint frequency-doubled 490 nm glow. I got to see two different octaves of light at once! And my eye was fine, although that was by no means guaranteed.

The photopic intensity curve is defined out to 825 nm, and is described by an approximate halving of sensitivity every 10 nm past 680 (at which point sensitivity is about 1.7% of its 555 nm peak). By 800 we're at about 3.7 ppm. Some papers I've seen indicate that this relationship continues all the way to 1064 nm, where a more careful "shine lasers into volunteers' eyes" experiment was able to get subjects to see it without exceeding exposure guidelines. The 1064 nm CW laser was red, but a pulsed 1060 nm laser was perceived as green due to two-photon absorption.

On the UV side, I was able to look directly into a 310 nm UV LED and see the UV light, too, as confirmed by a diffraction grating spectroscope. Here I took advantage of the fact that the human eye is potentially sensitive to UV, but it's filtered out by the lens. The filtering isn't perfect, though, and younger people have a window around 310. It was a pretty bluish violet, shifted a bit towards the white compared to 405 nm true violet. All three cones have secondary peaks in the UV, with the red cone kicking in first in the low 400s (that's why violet looks purple), and the other cones kicking in and shifting it more towards white as you get into real UV. It's still definitely blue-violet, just not as violet as true violet.

 

[Manfred Belheim]

Modern day Newton

 

[Lexicus]

Hey quick question do humans vary at all in the wavelengths they can see or are we all the same that way??

 

[Lexicus]

Seems likely as people have other ranges in senses such as hearing, smell etc.

Yeah, it does seem to make sense but I don't know that I've ever read anything about, like, one person being able to see further into the IR than another for example.

 

[Manfred Belheim]

Seems likely as people have other ranges in senses such as hearing, smell etc.

The case will be somewhat different to hearing in that it depends on three specific pigments with fixed physical properties, but I'd guess there are going to be other variables as well (general condition of the eye/cornea/various humours, optic nerves, plus whatever's going on on the brain side)

 

[Broken_Erika]

I have slightly different colour perception in each eye, my right eye sees colour more vivid than the left one......

 

[Bootstoots]

Hey quick question do humans vary at all in the wavelengths they can see or are we all the same that way??

Yeah, a little bit. For instance old geezers usually have some yellowing of the lens that reduces their sensitivity to blue/violet/near-UV, and people can have slightly different peak cone sensitivities, sometimes due to an unusual variant of the light-sensitive pigments in one or more of their types of cone cells. Of course they can also be missing a cone entirely as in many cases of red-green colorblindness, or very rarely they can have a duplicated cone so that they see in four primary colors (tetrachromacy). It's not as cool as in birds, most of which are tetrachromats with a UV cone - human tetrachromats, all women due to the X-linkage of the red and green cone cells, usually have their extra cone in between the red and green ones, so they gain a lot of sensitivity to different colors in the 540-610 nm region but not any extra spectral range.

Probably the coolest variant happens in aphakia, where some people without a lens can see ultraviolet light. The lens adds visual acuity but the cornea already does the majority of the work in focusing light, so for the expense of some blurry vision, people without lenses see deep into the UV down to the corneal absorption cutoff around 305 nm. It does appear bluish-white or violetish-white for them as it does for me when I just stare directly into a UV LED.

The earliest cataract surgery involved simple removal of the lens, which causes acquired aphakia. The most famous case of this was Monet, whose art suddenly shifted bluer after getting this surgery. Nowadays, though, they replace them with UV-blocking artificial lenses, so the effect isn't present. Lots of whiners would complain about macular degeneration and other such things if it weren't for the UV blocking, even though they'd get to push the bounds of perception for quite a while before that happened.

I've never heard of anyone being able to see unusually far into the IR, though. Animal vision in general doesn't seem to make much use of the NIR for some reason - many animals see into the UV but the only NIR vision I've ever heard of involved some fish in muddy rivers. Water absorbs significantly in part of the NIR but not all of it; it could just be that it's difficult to get NIR-sensitive variants or that there isn't much survival advantage to it for most animals.

 

[Lexicus]

Yeah, a little bit. For instance old geezers usually have some yellowing of the lens that reduces their sensitivity to blue/violet/near-UV, and people can have slightly different peak cone sensitivities, sometimes due to an unusual variant of the light-sensitive pigments in one or more of their types of cone cells. Of course they can also be missing a cone entirely as in many cases of red-green colorblindness, or very rarely they can have a duplicated cone so that they see in four primary colors (tetrachromacy). It's not as cool as in birds, most of which are tetrachromats with a UV cone - human tetrachromats, all women due to the X-linkage of the red and green cone cells, usually have their extra cone in between the red and green ones, so they gain a lot of sensitivity to different colors in the 540-610 nm region but not any extra spectral range.

Probably the coolest variant happens in aphakia, where some people without a lens can see ultraviolet light. The lens adds visual acuity but the cornea already does the majority of the work in focusing light, so for the expense of some blurry vision, people without lenses see deep into the UV down to the corneal absorption cutoff around 305 nm. It does appear bluish-white or violetish-white for them as it does for me when I just stare directly into a UV LED.

The earliest cataract surgery involved simple removal of the lens, which causes acquired aphakia. The most famous case of this was Monet, whose art suddenly shifted bluer after getting this surgery. Nowadays, though, they replace them with UV-blocking artificial lenses, so the effect isn't present. Lots of whiners would complain about macular degeneration and other such things if it weren't for the UV blocking, even though they'd get to push the bounds of perception for quite a while before that happened.

I've never heard of anyone being able to see unusually far into the IR, though. Animal vision in general doesn't seem to make much use of the NIR for some reason - many animals see into the UV but the only NIR vision I've ever heard of involved some fish in muddy rivers. Water absorbs significantly in part of the NIR but not all of it; it could just be that it's difficult to get NIR-sensitive variants or that there isn't much survival advantage to it for most animals.

Thanks, this is interesting and confirms what I thought about the four-cones thing (they can see more different colors in the spectrum but not an expanded part of the spectrum). IIRC snakes have those face-pit things that can detect IR emissions from nearby animals, but I guess that doesn't really count as "sight" since it isn't their eyes.