I mean, LED lights made it practical to shine 10.000+ lumens from a single socket. It would have been to hot/expensive with filament bulbs before that.
Result : we end up with 100 watts bulbs, same as before, even though the initial intention was to reduce energy consumption. The bulbs are even pricier and need more ressources to manufacture than before.
To me, this looks like a rather different use, actually, which wasn't practical before: daytime lights for dark dwellings.
I don't think that many people blast 100 W LED lights at night, which is the classic use-case of incandescent bulbs. And if they do, they probably replaced much more powerful bulbs.
This is exactly the rebound effect. The Wikipedia article unfortunately makes it out to be a Bad Thing and frames it as a wasteful increase in energy usage. It clearly isn't if you look at the given examples or think of some yourself. The example of LED lightbulbs, as you point out, shows that there is an existing need that was not being met due to the previously high cost of generating a lot of light indoors. An increase in efficiency allows that need to be met, energy consumption rebounds to the previous levels because the available money in the system hasn't changed. It's perverse that people suggest we should tax that extra efficiency to prevent such things happening, that would effectively hobble technological innovation because 1) why would you choose the more efficient option if it costs you exactly the same and 2) you are no longer freeing up that cash to be spent on novel uses of those more efficient systems. Again the LED example is illustrative, if you taxed them to the point that it is just as expensive as before to make the lighting in your room a closer analog of daylight then a lot of people, myself included, would simply stick with incandescent lighting since it has a nicer quality and you wouldn't be seeing any increase in activity around creating better indoor lighting by using more LED lights.
The same could be said for any example where the rebound effect applies, which is pretty much every technological advance in human history.
There is a blurry line between "new use case" and "more of the same old", I believe.
In transportation for example : cheaper fuel makes it easy to import exotic fruits from far away (avocado?). It is arguably a new use case : we did not eat those fruits before. On the other hand it does somewhat replace some local food (assuming we ingest the same amount of calories).
I did sometimes turn normal incandescent lights by day when I feel the light is too low. Now the threshold for "too low" seems to move slightly.
I also recommend SowiloDS led strips [1]. My family has used these in our various rooms and offices after the start of the pandemic, atop existing shelves or mounted on the wall, and I love them. The amount of light, quality, and light control they provide (with additional Hue controller) is great. I've come to really prefer the diffuse lighting of LED strips over more powerful bulbs, to light a space.
Their Bifrost-168 Pro strip produces up to 2800 lm/m, so 5600 lm for their 2m variant at $60 ($10.71/klm for the strip only), or 14000lm for 5m at $145 ($10.35/klm for the strip only). You could split the 5m into smaller strips, but that's only for experienced solderers.
You can plug them into a Hue controller (sold by Sowilo, likely cannibalized from Hue's own dimmer light strips), or whichever controller you want to use. For me, that control and integration was well worth the increased cost.
I recommend them to anyone who'll listen, and plan on using more of them in any future space.
I mean, I'm talking what is probably the highest-end LED lighting, and I haven't found anything equivalent, feature-wise, on AliExpress. (that said, maybe I missed it! It's a wide, disorganized place)
I have cheaper strips that I use for kitchen counter lighting, under the cabinets, and those work fine for that application. But for my main room lighting, I've come to really appreciate the ability to shift the whitepoint (to such a wide range), and the brightness the Bifrost provide.
Harbor Freight has 4 "bulb" type LED shop lights rated at 10k lumen for under $50 and 5.5k lumen ones for under $25. They're around 6k temperature and can be daisy chained. They aren't configurable, but are a great, cheap way to light garages or work areas.
It'll be interesting to see where the research around this stuff shakes out. There's research about nearsightedness[1] being linked to the amount of time spent with bright light, and some stronger research being done to dig into that further[2]. However there's also research showing that the blue light from LEDs could be damaging[3] (the fact that the human stand in is flies definitely means more research is needed). I wonder what a good balance will end up being.
Definitely not quicker and easier, but artificial sunlight (like this https://www.youtube.com/watch?v=6bqBsHSwPgw) has impressed me and wondering if there are ready made solution for that
Interesting. You could do this parallel rays effect with a lot of lower power leds and microlenses too? Could work with air cooling then (more surface area).
I think the second link with the collimating lenses is more what's needed. I was just looking at another page and saw that you can also buy collimating reflectors which might be cheaper than lenses. I wonder if that's something you could 3D print and then coat with something to make it reflective. It'd probably be better to build a tool to punch them out of stainless steel sheet metal and then polish them to a mirror finish.
I've been a proponent of natural, bright daylight for a long time. It's why I sit right in front of a south-facing window in my home office, and I try to take a 30+ minute walk/run every day in the middle of the day, no matter the season, to maximize my sunlight exposure. I live north of the 43rd parallel and our winters have short daylight hours (although nothing like Scandinavia, Alaska, etc). I have never been diagnosed for SAD (Seasonal Affective Disorder), but I am fairly sure I, and practically everyone, have some mild form of it. Taking vitamin D in the winter helps a little, but nothing beats actually getting out in the sun. Humans just aren't meant to sit inside in the dark all day. It's amazing what sitting outside in the sun, even in negative temperatures, do for my mood and productivity.
I have enjoyed reading all of the posts on HN lately about creating brighter rooms. However instead of creating a solution with LEDs, I am determined to further increase my exposure to natural sunlight. As my wife and I get ready to buy or build a new house, I am all-in on installing skylights and as many windows as possible. Even a fully overcast day has 1000-2000 lux of brightness at midday.
> For SAD you can buy these cute little light therapy boxes that are supposed to help, but they don't, because they are not bright enough to make a difference. Waste of money.
This is completely false. There are multiple studies that show the effectiveness of light therapy. I myself have been using it for years and it makes a huge difference in my mood. If the light is less bright, you'll need to use it for a longer time each day to compensate.
From some research I did a little while back light therapy does work according to the research papers but the problem comes when the end product does not deliver enough lux (amount of light over a given surface area at a given distance). IIRC the research papers either define the dosing over time in lux or in lumens at a given distance (pretty sure it is lux though). So the issue is that these light therapy boxes list dosing in lumens or lux at a very short distance and when sitting at a usable distance to reasonably perform another activity for 15-30 min you achieve far under the dosing amount listed in the papers. To top it off these devices were usually over or well over $100.
I went with a video recording LED panel set close to me on a tripod which ended up being the more economical option to achieve the desired lux. But to each their own.
If you’d like some more links send an email to the address in my profile and I’ll see what I can dig up again. I’m on my phone right now so my ability to do that is limited.
Seriously, they're like 10k+ lumens. They make a difference. Granted effectiveness could probably be improved by hanging a whole bunch of lights in the environment.
Corn lights at 50w and above tend to have loud fans to help keep them cool. That's probably fine if they're outdoors and you don't have to listen to them, but indoors that can be tiring.
At the very least, I'd suggest looking closely at the bulbs you're buying to see where they switch over to using fans.
That was what I was wondering about; I've got some 5-8W LED spots here and there and they get really hot, to the point where I'm confident the heat is why they die after a few years at most (much shorter than the advertised 10 year lifespan).
Would you recommend a warm or cold white light, after having one for over two years?
At the moment I only have 2 rgb led lightbulbs, which have less than 1k lumens each, so I'm thinking of getting more "white-only" bulbs for more lightning but I'm curious about the warmth of the light: I literally only always go below 2700k.
For a garage, a cold white light (higher color temperature) will give you more lumens per watt, so you'll see brighter light. If the CRI is high enough, then things won't look too weird and your eyes should adjust to what just looks to be a brighter and whiter light.
A higher color temperature light is usually also okay or even recommended for kitchen and bathroom use, if the CRI is high enough.
A warmer light (lower color temperature) is usually recommended for living rooms, bedrooms, and other places where you're going to be participating in more leisurely activities, or sleeping, etc....
Ideally, you'd have lights that can cover the gamut and switch from high color temperatures (whiter light) during the day, and lower color temperatures (warmer light) at night, and then you would be able to adjust in each location for the type of light that you like best and that seems to be best suited to the activity.
There are some "white" lights that include LED modules with two or more color temperature set points, and allow you to adjust where you want things to be between them. But they are more expensive, and more difficult to find.
It depends on the use case. As the sibling says, you want both, ideally a gamut, depending on the use case.
When I bought the light in my comment, I already had some Hue white lights, but they didn't make much of a difference during the day. Some pencil Osram lights [0] did help somewhat in the morning, so that's when I decided I should try a "serious" LED light. As this was for working during the day, I wanted a cooler light, but my experience with other "cool white" lights was that they looked too much like hospitals, so I settled somewhere in the middle.
While I never tried to measure the temperature, it's definitely cooler than the Hues and Osrams, but I still find it pleasant.
[0] https://www.amazon.fr/gp/product/B071JSB5D1 - R7s socket, 118 mm, 2700K. I had grabbed those because there were two lamps with this socket in the apartment I was renting at the time.
"For SAD you can buy these cute little light therapy boxes that are supposed to help, but they don't, because they are not bright enough to make a difference. Waste of money"
Hm. I have no idea how good the full spectrum white light boxes are for SAD. Instead I have a little Philips goLITE Blu -- a very small blue-LED-array therapy box that you're supposed to put just in the corner of your vision.
It is very effective for me, even on its lower powers, to the extent that I can overdose on it and become jittery (like I sometimes do in the spring). It's alarmingly bright on full power, and you really would need to put it much further away. I use it sparingly to help correct drifting sleep patterns in winter (by brightening the middle of the day, usually).
Fifteen minutes or half an hour, a few days a week in December and January, helps enormously.
I keep dreaming about rigging up some robot arms terminating in large mirrors, which would use inverse kinematics to follow the sun and reflect more light into my office window.
They are called heliostats. They actually built a massive one into a building near my office in order compensate for the loss of natural light in the village square. https://youtu.be/6nYkQR-kPyM
That paper pretty much just proved that blue light is more damaging than other colours. That would be expected, it is higher energy. Only the amount of light energy per area at the bottom surface of the cage was measured. It was unclear exactly how diffused the light source was. That would be important because the lens of the eye would focus all the light on a small area of the retina. To relate this to a human lighting situation you would have to specify the light energy per area at the light source. People normally will not tolerate bare point source LEDs for lighting. They are diffused in practice.
It's not just about the Lumen numbers. During the day/working a higher white-point 5000°k - 7000°k (i.e. "cooler") lights brighten your brain... Those are the daylight colors. In the evening you want lower white points, 2700°k - 3500°k (the so-called "warmer" colors). However, your mileage will vary. I know of one retail outlet that shifted from 5000°k lights to 3500° because the "warmer" light color made the inventory more pleasing and the whole store feel more homey.
You want a quality light as well. Color Rendering Index (CRI) points you to how well the light fits in with the spectrum output of the sun with 100 being like outdoor light. Higher is better. You probably want to avoid anything under 90... That said CRI can be hacked. I have seen lights with a CRI of 96 that add a greenish cast to the world. Your brain adjusts your perception so you might not notice, but take a picture and you will see the difference... particularly rake a picture of the same thing under two lights that you eye sees as the same and the difference can be surprising.
Having caught the light geek bug in the late 80s when I had a print shop and had to mix ink to fairly critical color matching. I have learned to pay attention. particularly with the development of LED technologies. Those early LED lights were horrible. Nowadays the technology is superb. My absolute favorite LED manufacturer is Yuji LEDs (https://store.yujiintl.com). They are my preferred LEDs for human and photographic lighting needs. I can walk from a sunlit day into my shop and not really feel much of a transition of the light quality. They sell bulbs & fixtures as well as raw cobs, rolls & strip modules. They are a small company so their inventory fluctuates and changes. Right now they even have a 1500w COB that outputs 120,000 lumen... That will brighten your day... You just need to figure out how to power it and cool it.
I considered corn bulbs, but my concern was UV radiation. No one has tested the long term effects of putting these things close to your face - even a small percent of off target emissions could end up doing damage at 30,000 lumens for an hour a day every day. Would love to see evidence that this is not something I need to worry about.
LEDs don't really emit in the UV, unless specifically designed to do it. This is a bandgap property, a fundamental part of the material the LED is made of, so it's not changing any time soon.
The phosphors in white LEDs shift the line spectrum of the (usually) royal blue primary dice down, not up, so they're not going to cause trouble either. LEDs are not blackbody sources, which is actually useful here instead of annoying.
You can get many "chip-on-board" (COB) LEDs with more than 200 Lumen / Watt in the 2700 - 6500 K range. They are blindingly bright, emitting up to 2.000 Lumen from a 5 mm² surface area, enough to permanently damage your retina if you get too close to them. A while ago I built a small light strip with several such LEDs, less than 25 € LED cost for more than 5.000 Lumen (a suitable control unit is around 10-30 € depending on whether you want it programmable and dimmable or not). So you could easily build a setup that delivers 100.000 Lux (maximum peak brightness you find outside during the summer at noon), though I'm not sure you'll enjoy it. Great for aquariums though.
The dangers of having a very strong point light source are, I think, a good argument for distributing the lights across a large area. Since distance matters with light intensity it would also help to never be very far from a lightsource wherever you go in a room. Maybe the best option would be to cover your ceiling in LED light strips or a bunch of those LED lights that look like fluorescent light fixtures.
Yeah but LED strips are less efficient in general. The appeal of the COB systems is that they have high Lumen / Watt ratios, drawbacks are the intensity and of course the cooling. I had some smoldered plexiglass from these COBs as the heat darkened it and the darker it got the more heat it absorbed. Pretty dangerous, so I also tend to use more smaller lights nowadays.
The certified small SAD lights are probably ok at short distances for the 20 mins or so they are recommended for.
I have a couple of soft boxes for photography with these corn lights that I bounce from walls if I want a bright room. If you are bouncing the light from a wall you don’t even need a soft box, just some type of reflector to aim the direct light at a wall. I don’t think I’d want to be staring at a naked bulb for long periods
Feit Electric PAR38 55W 5000-lumen Daylight LED light bulb has some advantages over the corn bulb:
- Mount it in a floor lamp and it throws most of its light upward to the ceiling.
- No fan.
The downside is that it's fewer lumens. I use two of these, in torchiere-style lamps on both sides of my desk. The lamps make my desk area as bright as outdoors on a cloudy day.
My productivity went up a lot when I started using these lamps.
While not quick and easy like the article says, having a decent house (or office!) with plenty of natural sunlight - windows, skylights, "light tunnels" - is the gold standard.
Our office is... kinda shit, too dark even during daytime, and the fluorescent tubes weren't bright enough at all. Glad I first moved desks next to a window, then WFH for two years now.
I can live with the fact that they built houses 50 years ago with poor natural light. But this article is talking about artificial light? I would lose my mind in that case.
Leaving aside general complaints regarding the housing crisis in the Bay Area, I would just note that it doesn't seem like a house has been built at all in my neighborhood in over 50 years.
Only endless renovations that make apartment living in Bay Area even more miserable.
While the buildings basically seem low quality to me - Hear your neighbors all the time. Definitely hear the people upstairs walking. No natural light or ventilation in bathrooms and kitchens. Old/artificial/synthetic/toxic materials including carpets.
But hey, techies make decent money, always scheming to get out of here.
It's not quite 20k lumens, and you can't adjust the color temperature, but it is very bright and fits in a standard light socket. For other bright lighting options look for "high-bay" or "shop" lights. High-bay lights are intended to be installed in warehouses or places with very tall ceilings so they tend to be very bright.
Not inherently, though.
I doubt it's much harder to find good high bay lamps than it is to find good E27 lamps, and expect the opposite once you cross the 20-30 W limit that's realistic for a passively cooled normal-sized E27.
Sure, you could build 95CRI lamps at 2700K, or 3000K, or whatever lower color temperature you want. Build whatever you want with the right LED type.
But can you actually easily find them on the commercial market, outside of tape reels?
Every garage light or high bay LED light I've found is always at or above 5000K, and their CRI is almost never listed. The CRI might never even be officially tested.
If you do have sources for garage or high bay lights that do have a low color temperature but a high CRI, please let me know. Because right now I'm stuck with a 50W corn bulb, and even though it's supposedly low color temperature, I can tell you that the CRI is absolutely horrible and this is the upper limit of how far I can go with this manufacturer.
Yes, you can buy them from Yuji.
Lower-than-2700K would likely involve the LED tape I use as a living room lamp (5m of 20W/m high-density at 1900K CRI 95+).
If you're willing to hook up the power supply and cooling yourself, then I suspect this would be excellent https://www.amazon.co.uk/dp/B08H5BWJ86 Also available off AliExpress (but it's not any cheaper). IME, if the CRI is high enough then the cooler colour temperatures are much more pleasant and 4-5k is ideal.
That would get you 20,000 lumens out of a 200w panel, which is a lot closer to what I think the original post was all about. You'd still have to wire the whole thing up, however.
Unfortunately, that doesn't help me with my problem of trying to get a brighter and higher CRI "bulb" for my garage.
I have resisted replacing the T-8 fluorescent light fixture, because I'd like to have a mix of technologies for the lights in that space. You know, in case the temperature in the garage gets too low and takes out one or the other technology -- I've seen major flickering and strobing with LED lights when it's really cold outside.
But I am starting to wonder if maybe I should replace the T-8 with LED-based models.
for high brightness high CRI, I build lamps of out of Bridgelux modules. where 'building a lamp' generally means 'screwing the module to a heatsink'. they are pretty cheap ($14 at digikey for a 30W module) last quite a long time if you cool them properly. you do have to hook them up to an appropriate constant current power supply which does involve some familiarity with what current and voltage are.
I readily admit that's too DYI for most consumers, but the market is really only doing a good job of pumping out LED bulbs that fit into standard sockets and have dubious and variable light quality. you kind of just have to buy of bunch of them and throw out 80% for being too harsh, or too green, or simply failing.
we could have better standards (mechanical, electrical) that accommodate the technology shift. we could have more useful and flexible product lines that rethink the way we distribute power for lighting. the dimmer situation is nightmarish for consumers - dimming LEDs is trivial, but the consumer definition of dimmer is 'triac AC chopper' by default - not because they chose that
why isn't consumer capitalism fixing this? this is supposed to be what its best at.
it is pretty easy to find strings of light sockets, similar to christmas lights, with sockets for regular bulbs, meant for stringing up around cookout areas. The bulbs with dozens of LEDs lined up on "filaments" don't need power conversion, so are very efficient and cool-running. Hanging them involves just strategically placed hooks, or even nails. They are perfectly silent.
I wonder if one can get some sort of a diffuser/shade for this kind of bulb, so as to be able to point it at the ceiling / top part of the wall, and get a more ambient light via reflection rather than a blinding super-bright spot.
Naturally, this will reduce the lux level, but I don't see how I could manage such bulbs otherwise - I would constantly have to avert my eyes.
That’s what I’ve been looking for. They make “deformable garbage” lights that rotate, but I can’t find any that actually confirm you can point it at the ceiling.
I have one of these things and at first it was like the sun brought to earth. But now I’m kinda used to it when on, and don’t even really notice the difference compared to the other regular bulbs.
And then the sun comes through the window just right and blows the cob out of the water.
I have one of these. I use it exclusively when I'm decorating/refurbishing a room, as it makes the room so bright I can clearly see where I'm painting etc.
Downside is, it totally nukes any WiFi signal in a 5 metre radius. :(
I wonder if it can substitute sunlight for the benefits of kids' eyes. Research shows nearsightedness is caused by insufficient outdoor time (not by looking at near things like screens), mechanisms unclear.
How does this research control for the fact that “being outside” generally involves a lot of focusing on stuff further away than a book/screen/etc in your lap? Did they have a cohort of kids who went outside and looked at a lot of bugs or sat around reading or something?
You'd also be changing the focus of your eyes nearly constantly as they saccade from one object to another. I agree that it's ridiculous using a study like that as a way to dismiss the conventional wisdom that staring at things close to you face for long periods of time is what causes nearsightedness since this study is proving the point that not spending a lot of time doing indoor activities is better for your eyesight.
There's been further research in this area that shows it's light exposure in general[1], not just outdoor time. There's also an interesting ongoing study to help determine the effectiveness of using bright artificial light as a treatment[2]
Nice idea! I grabbed a couple Dragonlights from Amazon - there are a bunch of knockoff models, but quite a few quality control issues with ones that aren't from the official Dragonlight store, just FYI.
I've been doing this, fake specs are a problem on the cheap ones, watch for this.
Also checkout "Four Leaf Lighting"
I'd replace any fluros, including tubes.
Don't look at money/watts saved vs replacement costs.
Increase the lumens to better use your space, your space is worth a fortune.
The instant on will surprise you how it matters compared to fluros, it's like when a constant hum is turned off.
You'll probably save on watts as well. Fluros chew a lot on top of their rated wattage, old ballasts and startups.
All that said, society also needs to move to better lighting spectrum's. But the above is a good start. I think Cool White is better than Warm White when it's a while to bed, but do your own research.
And if any a@#holes try to talk to you about saving on wattage, drop them from your life. Your health matters more than trying to save an insignificant amount of energy, they don't actually care about the world they just like talking down to you, they are toxic people.
I recently did fluorescent replacements to LED’s in my house. It’s surprisingly pretty straightforward. You can buy dual end power LED bulbs pretty cheap on Amazon - I paid around $30 USD for 4 (they will last around 10-12 years with regular use). They are significantly cheaper than that already pretty low price when you are buying 10+. Replacement is pretty simple: flip off breaker, cut the ballast wires about 1 inch before they go into the ballast, and strip the bulb wires where you cut and send them directly into the house wiring instead of into the ballast. I didn’t even bother removing the ballast and used the same wire nuts that were in there before so there wasn’t any hardware removal.
The only annoying part is getting rid of the old fluorescent tubes - they are hazardous waste due to the mercury content and there was only one place in my city that would take them and it was all the way on the far side of the rather large metro.
In this particular moment in human history we should aim to reduce energy consumption. Using 80 watt led bulbs to produce more light completely offsets the gain in efficiency that they provide. And It makes me feel more discouraged about our future.
A 80 watt incandescent light bulb will produce nowhere near close to 20000 lumens, more something along the lines of 1200-1300 lumens thus you would need 15+ times more to produce the same luminous flux.
Thank you for taking the time to say this. I've been thinking about this myself, having seen a few articles posted recently along similar lines.
I know that one architectural solution to getting more light indoors is often called daylighting[0]. I'm fortunate to live and work in old buildings with high ceilings and large, plentiful windows. What can we do about existing spaces with inadequate light, however?
High ceiling and large plentiful windows also increase the heating need. Should you want to reduce you energy consumption, you have to reach a balance. Within that balance light is pretty low in any energy budget. In France it's 10% of the electricity consumed by people.
Energy consumption for heat can also be solved by thicker walls and better insulation, but how are those walls and insulation produced? Are we talking about energy or carbon reduction? It's tough to analyze the whole chain. Burning wood is carbon-neutral, but can be carcinogenic when not fully efficient, and transporting wood can generate carbon...
What I'm getting to is that it's a complex system, and among that system, pollution due to light generation is really, really, really low at the bottom of the priority list.
I mean, LED lights made it practical to shine 10.000+ lumens from a single socket. It would have been to hot/expensive with filament bulbs before that.
Result : we end up with 100 watts bulbs, same as before, even though the initial intention was to reduce energy consumption. The bulbs are even pricier and need more ressources to manufacture than before.