Could someone please start a bright home lighting company?

[lincolnquirk]

Link post

Ele­va­tor pitch: Bring enough light to simu­late daylight into your home and office.

This idea has been shared in Less Wrong cir­cles for a cou­ple years. Yud­kowsky wrote Inad­e­quate Equil­ibria in 2017 where he and his wife in­vented the idea, and Rae­mon wrote a play­book in 2018 for how to do it your­self. Now I and at least two other friends are try­ing to build some­thing similar, and I sus­pect there’s a big­ger-than-it-looks mar­ket op­por­tu­nity here be­cause it’s one of those things that a lot of peo­ple would prob­a­bly want, if they knew it ex­isted and could ex­pe­rience it. And it’s only re­cently be­come cheap enough to ex­e­cute well.

Coelux makes a high-end ar­tifi­cial skylight which cer­tainly looks awe­some, but it costs up­wards of $30k and also takes a lot of head­room in the ceiling. Can we do bet­ter for cheaper?

Bright­ness from first principles

First let’s clear up some defi­ni­tions:

• Watts is a mea­sure of power con­sump­tion, not bright­ness.

  • “Watt equiv­a­lent” bright­ness is usu­ally listed for LED bulbs, at least for the stan­dard house­hold bulb form fac­tor. You should gen­er­ally ig­nore this (in­stead, just look at the lu­mens rat­ing), be­cause it is con­fus­ing. Nor­mally “watt equiv­a­lent” is com­puted by di­vid­ing lu­mens by 15 or so. (bulb man­u­fac­tur­ers like to make LED bulbs that are easy to com­pare, by hav­ing similar bright­ness to the in­can­des­cents they re­place, hence “watt equiv­a­lent”)

• Lu­mens out­put is a mea­sure­ment of an in­di­vi­d­ual bulb, but says noth­ing about the dis­tri­bu­tion of those rays of light. For that you want to be do­ing math to es­ti­mate lux.

• “Lux”, or “lu­mi­nous flux”, is the mea­sure­ment of how bright light is on a cer­tain sur­face (such as a wall or your face). Lux is mea­sured in lu­mens per square me­ter. Usu­ally, your end goal when de­sign­ing light­ing is to cre­ate a cer­tain amount of lux.

  • Direct sun­light shines 100k lux (source for these on Wikipe­dia)

  • Full daylight (in­di­rect) is more than 10k lux

  • An over­cast day or bright TV stu­dio light­ing is 1000 lux

  • In­door office light­ing is typ­i­cally 500

  • In­door liv­ing room at night might be only 50

Side note: This scale sur­prises me greatly! We use­fully make use of vi­sion with four or more or­ders of mag­ni­tude differ­ences in lux within a sin­gle day. Our hu­man vi­sion hard­ware is do­ing a lot of work to make the world look rea­son­able within these vast differ­ences of amount of light. Re­gard­less, this post is about get­ting a lot of lux. I hy­poth­e­size that lux is as­so­ci­ated with both hap­piness and pro­duc­tivity, and dur­ing the “dark sea­son” when we don’t get as much lux from the sun, I’m look­ing to get some from ar­tifi­cial lights.

If you put a sin­gle 1000-lu­men (66-watt-equiv­a­lent) om­ni­di­rec­tional bulb in the cen­ter of a spher­i­cal room of 2m ra­dius (which ap­prox­i­mates a 12′ square bed­room), the lux at the ra­dius of the sphere is 50. So now we can get a sense of the scope of the prob­lem. When doc­tors say you should be get­ting 10,000 lux for 30 min­utes a day, the de­faults for home light­ing are two or­ders of mag­ni­tude off.

• Rae­mon’s bulbs are “100W equiv­a­lent” which is ~1500 lu­mens per bulb. So he’s got 36k lu­mens. If we treat this as a point source and ex­pect that Rae­mon’s head is 2m away from the bulbs, then he’s get­ting 1800 lux, which is twice the “TV stu­dio” light­ing and seems pretty re­spectable. I haven’t ac­counted for re­flected light from the ceiling ei­ther, so re­al­ity might be bet­ter than this, but I doubt it changes the calcu­la­tion by more than a fac­tor of 2 -- but I don’t have a ro­bust way of es­ti­mat­ing am­bi­ent light, so ideas are wel­come.

• David Chap­man’s plan uses three 20k-lu­men LED light bars for offroad SUV driv­ing, for a to­tal of 60k lu­mens. But be­cause the light bars aim the light at a rel­a­tively fo­cused point on the floor, David es­ti­mates that most of that light is be­ing de­liv­ered to a roughly 6-square-me­ter workspace for a to­tal of 10k lux. The pho­tos he shared of his workspace seem to sup­port this es­ti­mate.

Other im­por­tant fac­tors be­sides brightness

Color tem­per­a­ture seems im­por­tant to well-be­ing. Color tem­per­a­ture is mea­sured in kelv­ins with refer­ence to black-body ra­di­a­tion, but you can think of it as, on the spec­trum from “warm white” to “cool white”, what do you pre­fer? Rae­mon’s plan uses an even split be­tween 2700K and 5000K bulbs. 2700K is quite yel­low-y, 5000 is nearly pure white. In my ex­per­i­men­ta­tion I dis­cov­ered that I liked closer to 5000 in the morn­ings and closer to 2700 in evenings.

And what about light dis­tri­bu­tion? Large “pan­els” of bright light would seem the clos­est to daylight in form-fac­tor. Real win­dows are brighter near the top, and it is con­sid­ered dra­matic and un­nat­u­ral to have bright light­ing com­ing from the ground. Also, sin­gle bright point sources are painful to look at and can seem harsh. I think there’s a lot of flex­i­bil­ity here, but I think my per­sonal ideal light would be a large, win­dow-sized panel of light mounted on the ceiling or high on the wall.

Also, color ac­cu­racy: LEDs are no­to­ri­ously nar­row spec­trum by de­fault; man­u­fac­tur­ers have to do work to make their LEDs look more like in­can­des­cent bulbs in how they light up ob­jects of differ­ent col­ors. Check for a mea­sure called Color Ren­der­ing In­dex, or CRI, in product de­scrip­tions. 100 is con­sid­ered perfect color ren­der­ing, and any­thing less than 80 looks in­creas­ingly awful as you go down. The differ­ence be­tween CRI 80 and 90 is definitely no­tice­able to some peo­ple. I haven’t blind tested my­self, and definitely might be imag­in­ing it, but I feel like there was some kind of no­tice­able up­grade of the “co­z­i­ness” or “warmth” in my room when up­grad­ing from CRI 80 to CRI 95 bulbs.

Dimma­bil­ity? (Are you kid­ding? We want bright­ness, not dim­ness!) Okay, fine, if you in­sist. Most high-end LED bulbs seem dimmable to­day, so I hope this is not an oner­ous re­quire­ment.

Last thing I can think of is flicker. I have only seen flicker as a ma­jor prob­lem with re­ally low-end bulbs, but I can eas­ily see and be an­noyed by 60hz flicker out of the cor­ner of my eye. Cheap Christ­mas LED light strings have su­per bad flicker, but it seems like man­u­fac­tur­ers of nicer LEDs to­day have caught on, be­cause I haven’t had any flicker prob­lems with LED bulbs in years.

Okay, so to sum­ma­rize: I want an all-in-one “light panel” that pro­duces at least 20000 lu­mens and can be mounted to a wall or ceiling, with no no­tice­able flicker, good CRI, and ad­justable (per­haps au­to­mat­i­cally ad­just­ing) color tem­per­a­ture through­out the day.

A red­di­tor made a fake win­dow for their base­ment which is quite im­pres­sive for un­der $200. This is definitely along the axis I am imag­in­ing.

I haven’t men­tioned op­er­at­ing cost. Full-spec­trum LEDs seem to out­put about 75 lu­mens per watt, so if our panel is 20k lu­mens then we should ex­pect our panel to draw 266 watts. This seems rea­son­able to me. If you leave it on 8 hours a day, you’re go­ing to use 25 cents per day in elec­tric­ity (at $.12 per kWh).

Mar­ket­ing and Costs

What do you think peo­ple will pay for the product? I have already put 6+ hours into re­search­ing this and don’t have a satis­fac­tory solu­tion yet. I would prob­a­bly pay at least $400 to get that time back, if the re­sult satis­fied all my re­quire­ments; I ex­pect to put in quite a bit more time, so I think I could prob­a­bly be con­vinced to pay north of $1000 for a re­ally good product. Hard to say what oth­ers would pay, but I wouldn’t be sur­prised if you could build a good product in the $400-1200 range that would be quite pop­u­lar.

What about costs? To­day, Home De­pot sells Cree 90-CRI, 815-lu­men bulbs on their web­site for $1.93 per bulb for a cost of $2.37 per 1000 lu­mens. This is the cheap­est I’ve seen high qual­ity bulbs. (The higher lu­men bulbs are an­noy­ingly quite a bit more ex­pen­sive). To get 36k lu­mens at this price costs un­der $100 re­tail. Pre­sum­ably there are cool­ing con­sid­er­a­tions when pack­ing LEDs close to­gether but those seem solv­able if you’re do­ing the “panel” form fac­tor. There are other costs I’m sure, but it seems like the LEDs and driver are likely to dom­i­nate most of the costs. Th­ese are dimmable but not color tem­per­a­ture ad­justable.

Yuji LEDs sells 2700K-6500K dimmable LED strips, also with 95+ CRI, at $100 for 6250 lu­mens (so a cost of $16 per 1000 lu­mens). This is 7x more ex­pen­sive per lu­men, but know­ing that it ex­ists is re­ally helpful.

Pro­mo­tion and Distribution

Kick­starter is the ob­vi­ous idea for get­ting this idea out there. I would also recom­mend start­ing a sub­red­dit (if it doesn’t ex­ist; I haven’t checked yet) for do-it-your­selfers who want to build or buy re­ally bright light­ing sys­tems for their homes, as I think there is prob­a­bly enough sus­tained in­ter­est in such a topic for it to ex­ist.

You can also try to get press. The idea of “in­door light as bright as daylight” is prob­a­bly some­what viral so I’d hope you can get peo­ple to write about you. Coelux got a bunch of press a few years ago do­ing this ex­act thing, but their product is so ex­pen­sive that they don’t even list their price on their web­site, but in ar­ti­cles about Coelux you can see peo­ple com­ment­ing that they wish they could af­ford one.

I do think the idea needs to be spread more. Most peo­ple don’t know this is pos­si­ble, so there’s a lot of work you’ll be do­ing to just ex­plain that such a thing is pos­si­ble and healthy.

Com­pe­ti­tion?

I don’t think there’s any rele­vant com­pe­ti­tion out there to­day. Coelux is su­per high end. The com­pe­ti­tion is do-it-your­selfers, but this mar­ket is far big­ger than the num­ber of peo­ple who are ex­cited to do-it-them­self.

Some have men­tioned “high bay” lights, which are de­signed to be mounted high in ware­houses and such, and throw a light cone a long dis­tance to the floor. I am ex­cited to try this and I will prob­a­bly try it next, but I am not su­per op­ti­mistic about it be­cause I ex­pect it to be quite harsh. This is the one that Yuji sells, but you can find cheaper and pre­sum­ably lower-qual­ity ones on Ama­zon.

Part of my mo­ti­va­tion for writ­ing this blog post is to source ideas for other things that ex­ist that could fill this niche. Com­ment here if you solved this prob­lem in a way I haven’t de­scribed! I’ll up­date this post with ideas. If you start this com­pany, also email me and I’ll buy one and try your product and prob­a­bly write about it :)

Build­ing a Sus­tain­able Business

If you put a bunch of re­search into de­sign­ing a re­ally great product and it suc­ceeds but gets effec­tively copied by low-cost clones, you’ll be sad. I am not sure how to defend this, and I think it is prob­a­bly the weak­est point of this busi­ness model; but it is a weak­ness that many hard­ware com­pa­nies share, and a lot of them still carve out a niche. One idea would be to build up your product’s brand­ing and rep­u­ta­tion, by ex­plain­ing why low-cost clones suck in var­i­ous ways. Another is just to give re­ally good ser­vice. Lastly, if you avoid man­u­fac­tur­ing things in China, maybe Chi­nese clone com­pa­nies won’t copy your tech­nol­ogy as quickly.

[benkuhn]

I’m one of the friends men­tioned. Here’s some more anec­data, most im­por­tantly in­clud­ing what I think is the cur­rent eas­iest way to try a lu­me­na­tor (re­quiring only one fix­ture in­stead of huge num­bers of bulbs):

I don’t have sea­sonal de­pres­sion, but af­ter spend­ing a win­ter in a trop­i­cal coun­try, it was ex­tremely no­tice­able that it’s harder for me to fo­cus and I have less willpower when it’s dark out (which now starts at 4:15). I bought an ex­tremely bright light and put it right next to my desk, in my periph­eral vi­sion while I work. It was an im­me­di­ate and very no­tice­able im­prove­ment; I es­ti­mate it buys me 30-120 min­utes of fo­cus per day, de­pend­ing on how over­cast it is.

You can see a be­fore-and-af­ter here, al­though my phone cam­era’s dy­namic range is not good enough to re­ally cap­ture the differ­ence.

Every­one who has vis­ited my house since I got the light­bulb has re­marked on how nice it feels, which I was ini­tially sur­prised by since the bulb is 5600k and not par­tic­u­larly high-CRI.

My cur­rent setup is hon­estly kinda crappy (but still amaz­ing). I’m work­ing on a much nicer DIY ver­sion, but in the mean time, here’s the stuff I bought:

• 250-watt corn bulb (~= 40 60w-equiv­a­lent bulbs; $100)

  • This bulb has a pretty loud fan (~50db at close range); if you don’t like noise, you can buy two of the 120-watt ver­sion.

• this E39 fix­ture ($15)

  • the clamp is too weak to hold the bulb, but you can jerry-rig a sup­port by em­bed­ding the socket into the sty­rofoam pack­ag­ing that the light comes in :P

  • Also if you use this you’ll need to turn it off and on by un­plug­ging as there is no switch on the fix­ture.

• these E39 to E26 adapters ($10 for 4)

  • buy if you want to put in an over­head light or tra­di­tional lamp

  • note that the bulb does not fit well in many fix­tures be­cause it is very large and heavy

(Ama­zon links are af­fili­ate so I can see whether they are use­ful to peo­ple)

[gwern]

More dis­cus­sion: https://​​www.benkuhn.net/​​lux https://​​news.ycombi­na­tor.com/​​item?id=21660718

[mako yass]

What if we just had brighter screens?

If it just needs to be bright­ness in the field of vi­sion rather than bright­ness in the room, well, most of the time there’s a (very large) screen dom­i­nat­ing my field of vi­sion.

I have now set my screen bright­ness in un­com­fortable ranges. Hav­ing difficulty ad­just­ing but feel­ing very awake. Will re­port back in a week, I guess.

I was con­sid­er­ing pro­ject­ing bright light onto the wall be­hind the screen (this would al­low the light to be de­fused a lot, and it would be very easy to de­ploy, wouldn’t even need to hang the thing, let alone make a power socket), but it oc­curred to me that hav­ing the back­drop be brighter than your screen tends to cause headaches.

[Bucky]

I think this is likely to be or­ders of mag­ni­tude away from the kinds of things which have been effec­tive for oth­ers (see e.g. this rough calcu­la­tion on red­dit)

[Raemon]

Do you cur­rently have SAD?

[mako yass]

No and it’s sum­mer in my hemi­sphere any­way (but I spend a lot of time in­doors)

[Raemon]

What are you test­ing, then?

[mako yass]

I am fre­quently af­flicted with the kinds of drowsy de­pres­sive states that I would as­so­ci­ate with a state of dor­mancy in a deep win­ter. I think I heard that brighter lights gen­er­ally in­crease alert­ness and pro­duc­tivity. My cur­rent model is.. the mechanisms for de­ter­min­ing whether the hu­man is in­doors and (there­fore?) about to sleep are just very very crude. The model is also try­ing to ac­count for the the CO2 con­cen­tra­tion thing, which, last I heard we didn’t have any other plau­si­ble evolu­tion­ary ac­counts for.

[Koen.Holtman]

I used to work in the light­ing in­dus­try, so here are some com­ments from an in­dus­try per­spec­tive.

There are sev­eral high-qual­ity stud­ies about how more light, and be­ing able to con­trol dim­ming and color tem­per­a­ture, can im­prove sub­jec­tive well-be­ing, alert­ness, and sleep pat­terns. It is gen­er­ally ac­cepted that you do not need to go to di­rect sun­light type lux lev­els in­doors to get most of the benefits. Also, you do no need to have the bright­est dim level on all the time. For some peo­ple, the thing that will re­ally help is a reg­u­lar sched­ule that dims down be­low typ­i­cal in­door light lev­els at se­lected times, with­out ever dim­ming above typ­i­cal lev­els. I am not an ex­pert on the lat­est stud­ies, but if you want to build an in­door ex­per­i­men­tal setup to get to the bot­tom of what you re­ally like, my feel­ing is that in­stal­ling more than 4000 lux, as a peak ca­pac­ity in se­lected ar­eas, would definitely be a waste of money and re­sources.

If I would want to in­stall a has­sle-free bright light setup in my home cheaply, I would buy lots of high-end wire­less dimmable and color tem­per­a­ture ad­justable LED light bulbs, and some low-cost spot lights to put them in, e.g. spot lights that can be at­tached to a ceiling mounted power rail. If you make sure the bulbs sup­port the ZigBee stan­dard, you will have plenty of op­tions for con­trol soft­ware.

If power rails with lots of ~60W equiv­a­lent bulbs lacks aes­thetic ap­peal for you, then you could go for a high-end spe­cial form fac­tor product like that from Coelux men­tioned above. The best way to think about the Coelux product, in busi­ness model de­vel­op­ment terms, is that it is not re­ally a light­ing product: it is a spe­cial­ised piece of high-end fur­ni­ture. So if you want to de­velop a busi­ness model for a bright home light­ing com­pany, the first ques­tion you have to ask your­self is whether or not you want to be in the high-end fur­ni­ture busi­ness.

By the way, the main rea­son why the light­ing in­dus­try is not mak­ing any 200W or 500W equiv­a­lent LED bulbs that you could put in your ex­ist­ing spot lights is be­cause of cool­ing is­sues. LEDs are pretty en­ergy effi­cient, but LED bulbs still pro­duce some in­ter­nal heat that has to be cooled away. For 60W equiv­a­lent this can hap­pen by nat­u­ral air flow around the bulb, but a 200W equiv­a­lent bulb would need some­thing like a built-in fan.

[benkuhn]

Thanks, this com­ment is re­ally use­ful!

It is gen­er­ally ac­cepted that you do not need to go to di­rect sun­light type lux lev­els in­doors to get most of the benefits.… I am not an ex­pert on the lat­est stud­ies, but if you want to build an in­door ex­per­i­men­tal setup to get to the bot­tom of what you re­ally like, my feel­ing is that in­stal­ling more than 4000 lux, as a peak ca­pac­ity in se­lected ar­eas, would definitely be a waste of money and re­sources.

Do you have any poin­t­ers to where I might go to read the lat­est stud­ies?

[Richard_Kennaway]

For 60W equiv­a­lent this can hap­pen by nat­u­ral air flow around the bulb, but a 200W equiv­a­lent bulb would need some­thing like a built-in fan

From the net I find that a typ­i­cal 60W in­can­des­cent is 10% effi­cient, there­fore gen­er­ates 54W of heat within the light fit­ting. A typ­i­cal 60W equiv­a­lent LED bulb draws 7.5W and is 90% effi­cient, there­fore gen­er­ates 0.75W of heat in the fit­ting. There­fore, for an LED bulb to gen­er­ate as much heat as a 60W in­can­des­cent, it would gen­er­ate ⁵⁴⁄₀.75 = 72 times as much light, and be equiv­a­lent to 72*60 = 4320W of in­can­des­cent light­ing.

Since 60W in­can­des­cents do not need fan cool­ing (or even 150W, which I have used at home in the past), why would a high-pow­ered LED bulb?

[FactorialCode]

LED bulbs aren’t made of tung­sten and so can­not heat up to 3000+ de­grees with­out tak­ing dam­age. LED’s are much more sen­si­tive to heat and will burn out very quickly if not prop­erly cooled.

[benkuhn]

A typ­i­cal 60W equiv­a­lent LED bulb draws 7.5W and is 90% efficient

Where are you get­ting this num­ber? As far as I know, the most effi­cient LEDs to­day are around 50% effi­cient.

[Richard_Kennaway]

Oh, some­where on Google.

[jonathanclark]

Check out our product, it’s a far-cry from the $400 you men­tioned but many times less ex­pen­sive and thin­ner than coelux.

Check out a video here:

https://​​www.youtube.com/​​watch?v=XVV_-aB1jpg&t=57s

Some com­par­i­sons with Coelux: https://​​bit.ly/​​2YszzsO

We have done 7 years of re­search on this and just re­cently brought “Vir­tual Sun” to the mar­ket.

If you want to repli­cate a win­dow, there is more than just lu­mens, CCT, and CRI. In or­der for your brain to re­ally be­lieve some­thing is a win­dow or a skylight, you also need:

1. Separate sun and sky light com­po­nents. The real sun is very di­rec­tional, by the time it reaches the earth the rays of sun­light di­verge by only 0.5 de­grees, have a much lower CCT than the sky, and a much brighter lux pro­file than the rays of light from the sky. In­ner­scene Model A7 out­puts a rec­t­an­gu­lar sun­beam the di­verges at less than 2 de­grees and has a CCT of 4500K and is over 2500 lux at head level when sit­ting un­der it. Com­pared that with Coelux which has a 10 de­gree di­ver­gence an­gle. A7′s sky changes from 3000K up to well over 10000K.

2. The illu­sion of in­finite space be­hind the fix­ture. The LEAD build­ing stan­dard re­quires all rooms to have eye level sight lines to the out­doors. Psy­cholog­i­cally a room with win­dows is much more com­fortable to work in than those lit only by ar­tifi­cial light, part of this is that most peo­ple have some level of claus­tro­pho­bia—rang­ing from very mild (but no­tice­able) to panic at­tack in­duc­ing. Be­ing able to see out­side ele­vates a lot of that. We in­vented a new 3d dis­play tech­nol­ogy that al­lows users to see a sun at in­finity, rather than some­thing paint­ing on the sur­face so it looks like a por­tal, not a TV or flat sur­face. This turns out to be an ex­tremely hard prob­lem, and I haven’t seen any­one else do it—even coelux’s sun only ap­pears to be about 20 feet away from you.

3. Vari­abil­ity in CCT and in­ten­sity. Ar­tifi­cial light is too static which cre­ates monotony, we are used to a va­ri­ety of light pro­files from nat­u­ral light be­cause of the earths ro­ta­tion and at­mo­spheric effects. When the sun is ob­scured by clouds tem­porar­ily, the light in the room changes slowly but very sig­nifi­cantly—there is more diffuse light and less di­rec­tional light, the CCT gets warmer, and bright­ness in the room may go up or down de­pend­ing on the cloud thick­ness. Similar the CCT of light the sky changes through­out the day be­cause of the earths or­bit and how much light goes through the at­mo­sphere at differ­ent an­gles. In­ner­scene is able to repli­cate this and it makes a room feel “al­ive” and con­nected to the out­doors, com­bined with the 3D effect it is com­fortable to work/​live in a space with no real win­dows for long pe­ri­ods of time.

You can’t com­fortably work in a room where there is strong glare shin­ning in your eyes, which you would get if you buy ran­dom high pow­ered LEDs and stuck them on your ceiling. From a perfectly diffu­sive light ob­ject like the sky there is no glare, but you will have glare from an un­shielded light fix­ture un­less it is well diffused or you have baf­fles that block the light which isn’t go­ing down. Get­ting the di­rec­tional pro­file of light is very im­por­tant in this re­spect. Be­cause our sun is always at a high an­gle (30 de­grees) it doesn’t shine into your eyes di­rectly and the sky por­tion of the fix­ture is perfectly diffused like the real sky.

[Raemon]

(quick mod note: we nor­mally delete first com­ments that are mar­ket­ing a product. In this case it was sur­pris­ingly on topic and in­for­ma­tive, but wanted to be clear that it wasn’t prece­dent set­ting)

[lincolnquirk]

Hey wow, neat com­pany and I’m glad you posted about it here!

Un­for­tu­nately, I get the sense that your product is up in the Coelux range of pric­ing, be­cause you don’t list the price. I think a lot of peo­ple here are go­ing to im­me­di­ately dis­miss it as an op­tion given that we can’t eas­ily figure out how much it costs.

Your mar­ket­ing is also aimed at busi­nesses in­stead of homes. Espe­cially given that my post was about home light­ing, do you have any­thing you can share about home ap­pli­ca­tions of your product?

(BTW, I re­mem­ber read­ing an ar­ti­cle in the 90s from a tech mag—like Wired or some­thing—about crack.com, al­though I had no idea what you did, but the no­table thing that stuck in my brain from that ar­ti­cle was “why on earth would any­one name a com­pany af­ter a hor­rifi­cally ad­dic­tive drug?” I’d be cu­ri­ous if you know what ar­ti­cle I’m talk­ing about and whether you have a link to it!)

[Jonathan Clark]

re­gard­ing crack.com, funny you con­nected the dots! I was a co-founder of that com­pany which made the video game “Abuse” back in the 90s. The story be­hind the name… origi­nally we called the com­pany Chameleon Soft­ware but found some­one else was us­ing that name and had to change it. At the time the in­ter­net was just start­ing out and we thought it would be cool to use a do­main name to name the com­pany. We went through the dic­tio­nary and de­cided to call it “Crack Dot Com”, in those days we spel­led out the whole thing. This is in part be­cause at the time, there was a pop­u­lar model for sel­l­ing games called “Crack ware” be­cause the first one is free, you get ad­dicted and want to buy more :) We as­pired to make ad­dic­tive games and it was a bit edgy so I think it served us well. We didn’t re­al­ize it then, but crack was one of the top 100 search terms on the in­ter­net be­cause peo­ple were look­ing for soft­ware cracks so we got some free traf­fic (though ended up get­ting a lot of piracy!!) haha. My sec­ond com­pany, which turned into Thin­stall (now Vmware ThinApp) started out look­ing at how to stop soft­ware piracy by mak­ing desk­top soft­ware have a small SAS com­po­nent that was re­quired.

Any­way, I wanted to get back to my com­puter graph­ics roots so I started In­ner­scene to re­search how to make 3D dis­plays that look like win­dows (VR with­out a head­set). Turned out to be a much harder prob­lem that I ex­pected but we have some­thing re­ally amaz­ing now.

[Raemon]

Was this an “Abuse” game that was dis­tributed some­how by Bungie? If so I played that and have some fond mem­o­ries. Hi!

[Jonathan Clark]

yes! We pub­lished Abuse on DOS, Mac, and Linux (Red­hat). When the com­pany shut­down we re­leased the code to the pub­lic do­main so you can find it on many other plat­forms to­day in­clud­ing Win­dows or di­rectly in­side a browser ( https://​​play­clas­sic.games/​​games/​​run-and-gun-dos-games-on­line/​​play-abuse-on­line/​​ ), and on iPhone/​​An­droid—though I find it hard to play on a mo­bile phone :)

[Lukas Finnveden]

Check out our product, it’s a far-cry from the $400 you men­tioned but many times less ex­pen­sive and thin­ner than coelux.

How ex­pen­sive is it? Maybe there’s some rea­son that you don’t want an ex­act price in the com­par­i­son, but can you give a rough range?

[Jonathan Clark]

As a rough guide, you can in­stall 5-6 In­ner­scene A7 units in a pro­ject for the same price as one Coelux LC45 when you con­sider ship­ping and in­stall cost.

Although un­der­stand­ably frus­trat­ing, we don’t have pub­lic pric­ing for a few rea­sons.

1. Our product is gen­er­ally in­te­grated into a con­struc­tion pro­ject where the pur­chas­ing pro­cess can be more com­plex. In such a pro­ject, you might have a Gen­eral Con­trac­tor who sub­con­tracts out the elec­tri­cal work to an elec­tri­cian who pur­chases from a elec­tri­cal dis­trib­u­tor—and each party has a mark-up they charge for the value they are pro­vid­ing to the price to an end-user can vary. In ad­di­tion, for larger pro­jects there can be light­ing de­signer or light­ing sales agency in­volved that help with de­sign, code com­pli­ance is­sues, etc. Often these guys will work for free but charge a mark-up on the prod­ucts they spec­ify. For these, rea­sons you find most com­pa­nies that sell light­ing into com­mer­cial or larger res­i­den­tial pro­jects don’t have pub­lic pric­ing. Prod­ucts you find at home de­pot or on ama­zon are usu­ally com­mod­ity prod­ucts where ev­ery­one has pretty much the same thing and com­pa­nies com­pete mainly on price.

2. We work with part­ners who are al­lowed to set their own prices above our whole­sale price to them. The amount of markup they charge will de­pend on how much of the above leg­work men­tioned above they take on, the size of the pro­ject, li­a­bil­ity/​war­ranty/​ser­vice they take for the prod­ucts they provide, etc. Depend­ing on which coun­try you are in, there could be im­port du­ties, ship­ping charges, etc. That be­ing said, we do sell di­rectly to cus­tomers for most of the world and happy to provide a quick price by email us­ing info@in­ner­scene.com

3. As we work on op­ti­miz­ing our prod­ucts for higher vol­ume and less ex­pen­sive parts, we pass on cost sav­ings to our cus­tomers—so our prices con­tinue to come down year by year—for ex­am­ple A7 is >30% more af­ford­able com­pared with A6 and it’s a bet­ter product. How­ever, many places on the in­ter­net are per­ma­nent and we risk of peo­ple see­ing a price from 3 years ago and not look­ing any fur­ther be­cause it’s out of their bud­get.

[johnswentworth]

How would very bright in­door lights han­dle the $\frac{1}{r^2}$ prob­lem?

With the sun, it’s already very far away (rel­a­tive to the size of a room, or even the size of the whole earth), so light in­ten­sity doesn’t vary much as we move around - $\frac{1}{r^2}$ stays ba­si­cally con­stant. But in a house, you’ll some­times be 10X fur­ther from the light source than other times—so the light will some­times be 100X brighter than other times. (I don’t know if other peo­ple have this prob­lem, but those kinds of sharp light gra­di­ents tend to give me headaches—so mostly I ei­ther use dim light­ing in­side or go out­side.)

[Blake]

Based on this video, the Fres­nel lens and diffu­sion lay­ers in mod­ern TVs do a great job of push­ing light out­wards and soft­en­ing the gradient

[johnswentworth]

That is SO COOL, thanks for the link.

[Jonathan Clark]

The guy in the video is us­ing in­ac­cu­rate vo­cab­u­lary, there are no fres­nel lenses in a flat panel LCD/​TV. This page shows the typ­i­cal lay­ers you’ll find in a mod­ern dis­play:

https://​​pid.sam­sungdis­play.com/​​en/​​learn­ing-cen­ter/​​blog/​​lcd-struc­ture

In the video when he shows a film, you can tell it is bend­ing light not fo­cus­ing it—so it is likely a fres­nel prism sheet that he is hold­ing (not a fres­nel lens as he states).

You can use a fres­nel lens to make a light source ap­pear to be fur­ther away as he sug­gest, but that re­quires more depth as the light source needs to be at the fo­cal point of the lens. The other prob­lem is that you can only have 1 LED at the fo­cal point for the whole sheet, so you won’t get that much light com­ing out even if you use a high pow­ered LED.

You are likely to get bet­ter re­sults buy­ing a $50 LED light panel on Ama­zon com­pared to try­ing to re­cy­cle parts from an old LCD (though per­haps not as much fun). LED sources for LCDs of­ten have poor CRI so the light qual­ity you’ll get from this ap­proach is likely to be bad in per­son. I sug­gest search­ing for 90+ CRI with CCT >6000K LED Light pan­els.

[Shmi]

I won­der if the claim of ex­treme bright light alle­vi­at­ing the symp­toms of SAD has been tested by more than one per­son. While effi­cacy is not cor­re­lated with pop­u­lar­ity, given that this is a ra­tio­nal-think­ing based fo­rum, it would be at least nice to know if the de­vice works.

[ESRogs]

Did you miss the part about how a bunch of differ­ent peo­ple were do­ing this and benefit­ted from it? Or are you just as­sum­ing that those peo­ple didn’t offi­cially have SAD?

[Shmi]

I don’t see that part any­where, only the one ex­am­ple that ev­ery­one refers to.

[Douglas_Knight]

What do you want?

Do you want to buy some­thing for your­self, or do you want a com­pany to change the world?

Yes, there is room for a bet­ter product, but I think that off-the-shelf prod­ucts are pretty good and you should just get them. If you want to change the world, maybe you should just pro­mote these ex­ist­ing prod­ucts. In par­tic­u­lar, for your short term needs, just do it.

I think that the right an­swer for most peo­ple and most pur­poses is Rae­mon’s in­struc­tions, $300 for 300 watts, same to­tal wat­tage as coelux. Why did you write this post already know­ing about Rae­mon’s in­struc­tions? What are they lack­ing? That they re­quire in­stal­la­tion? If you have 24 sep­a­rate bulbs spread around the room, in­stal­la­tion is un­avoid­able. Light strips may be a bet­ter solu­tion, but they re­quire even more in­stal­la­tion.

Some peo­ple want differ­ent things. David Chap­man seems to want to illu­mi­nate his desk, not his room, so he might not like Rae­mon’s setup. If you want to min­i­mize in­stal­la­tion, you might want a sin­gle light. This leads to Ben and Ashen’s sug­ges­tions. They prob­a­bly aren’t as nice as coelux, so, yes, it would be nice if some­one made nicer ver­sions (which should be pos­si­ble). Ashen’s out­door floodlights prob­a­bly have lousy CRI. Ben’s corn­cob isn’t the stan­dard res­i­den­tial fix­ture, and thus re­quired some as­sem­bly. Both prod­ucts prob­a­bly shine out­wards to illu­mi­nate an area, rather than the coelux which is in­tended to mimic the sun through a win­dow push­ing light in a sharp line. This illu­sion is prob­a­bly lux­u­ri­ous, but I’m skep­ti­cal that it is ac­tu­ally good for the goal.

I was go­ing to fol­low up by say­ing that if you like the form fac­tor of coelux, there are similar prod­ucts on the mar­ket for maybe $2/​watt, only twice as ex­pen­sive as Rae­mon’s setup. They aren’t as bright as coelux, but you could get 5 or 10. There is the sec­ond product Ashen linked or light ther­apy boxes (ap­par­ently 72W) are prob­a­bly a good op­tion with full spec­trum and good lenses. But then I read more I heard a lot of ac­cu­sa­tions of poor qual­ity and fraud around light boxes, so I dunno.

[ashen]

The best solu­tion I could find was LED se­cu­rity flood lights, e.g.

https://​​www.ama­zon.co.uk/​​Cam­pHik­ing®-FloodLight-Land­scape-Spotlight-Water­proof/​​dp/​​B07JMQXXMW/​​ref=sr_1_9?key­words=500w+led+flood+lights+se­cu­rity&qid=1574927272&s=light­ing&sr=1-9

So this would be less than 100 dol­lars for ap­prox 40000 lu­men, with all in one en­clo­sure in­clud­ing head sink, can be wired to mains or wall socket. Also with this de­sign they can be an­gled.

Not ad­justable in bright­ness or colour tem­per­a­ture, but my prefer­ence would be over­head light­ing for day, and then in­can­des­cent side lamps af­ter dark (a sun above vs. fire).

[ashen]

In terms of con­sumer product, I think some­thing like this might be ideal:

https://​​www.ama­zon.co.uk/​​Ceiling-Dim­ming-Bath­room-Cor­ri­dor-6M5252TYQ/​​dp/​​B07MCTVH5V/​​ref=sr_1_fkmr0_2?key­words=lED+ceiling+lamp+office+lamp+with+re­mote+con­trol%2C+28W+2800lm+3000k-6000K+dimmable+LED+cir­cuit+board+ceiling+light%2C+splash­proof%2C+round+bed­room+light%2C+for&qid=1574930392&sr=8-2-fkmr0

But more like 300W in­stead of 28W. So tak­ing the en­clo­sure and re­mote con­trol of the bath­room type light, with the LED setup of a floodlight. As for CRI, I guess it would be bad. How im­por­tant is CRI though? Does this re­late to sub­jec­tive sense of “harsh­ness”?

I also found this: https://​​nofilm­school.com/​​diy-light-panel-busted-lcd-tv

Stresses im­por­tance of size, diffu­sion pads and fres­nel lens for cre­at­ing a soft, diffuse light.

Was think­ing a good DIY pro­ject would be to take a LED floodlight and wire in be­hind a busted 50″ dis­play as a side panel (ar­tifi­cial win­dow).

So the ideal con­sumer product might need to have a pretty wide sur­face area and fres­nel lens which would drive up costs.

[ashen]

Also saw this on hacker news to­day: https://​​news.ycombi­na­tor.com/​​item?id=21660718

one com­ment “Light­ing is a re­ally hard busi­ness (es­pe­cially res­i­den­tial)”

[Raemon]

Note for peo­ple build­ing a nu­anced model of the mod­er­a­tion policy:

I’ve front­paged this. This was a bit of an edge case for the front­page guidelines since it’s a call to ac­tion, and we don’t nor­mally front­page “startup an­nounce­ments,” which is a similar genre to this, at least su­perfi­cially. But, the differ­ence seemed im­por­tant: this post’s pri­mary goal seems more like ag­gre­gat­ing in­for­ma­tion rather than di­rectly bid­ding for at­ten­tion.

[romeostevensit]

De­toolbox has a bunch of mar­ket val­i­da­tion check­lists.

[daniel zev]

I have been look­ing into do­ing this. Pm me if you want to do it to­gether. I’m from Toronto but am will­ing to work with peo­ple from el­se­where.

(Espe­cially if you have any product de­sign ex­pe­rience, or think you can figure it out)

[philh]

I don’t have a ro­bust way of es­ti­mat­ing am­bi­ent light

Did you con­sider buy­ing a lux me­ter?

I’ve or­dered one of these, so I hope I’m not miss­ing some­thing.

[Matt Goldenberg]

If some­one does this—I think for this par­tic­u­lar product, the ob­vi­ous way to get val­i­da­tion is kick­starter or Indiegogo. It’s what could be a good look­ing tech product, with a group of peo­ple with very high de­mand, and a great story, but also high fixed costs.

Pretty perfect for the Kick­starter model.

[FactorialCode]

I’m cu­ri­ous to what ex­tents the benefits of this de­pend on race. I would ex­pect the op­ti­mal light­ing for an Afri­can who’s an­ces­tors were ex­posed to far more so­lar ra­di­a­tion to be higher than for a Euro­pean who’s an­ces­tors were ex­posed to far less.

[Brendan Long]

Huh that com­pany you link to at the end also has this sec­tion: https://​​store.yu­jiintl.com/​​col­lec­tions/​​high-cri-led-emitters

Maybe this is just a mat­ter of buy­ing this emit­ter https://​​store.yu­jiintl.com/​​col­lec­tions/​​high-cri-led-emit­ters/​​prod­ucts/​​bc-se­ries-high-cri-high-power-cob-led-5600k-bc270h-unit-1pcs?var­i­ant=25100635143 and hook­ing up to a suffi­ciently quiet high­sink + pack­ag­ing it in a more us­able form fac­tor.

I won­der if there would be is­sues about heat dis­si­pa­tion /​ po­ten­tial fires though?

[benkuhn]

I’ve done a bit of re­search on this. I think some­thing along these lines is prac­ti­cal. My biggest un­cer­tainty is what a “us­able form fac­tor” is (in par­tic­u­lar I don’t know how much diffu­sion you’d need, or in what shape, with a very small emit­ter like this).

FWIW, the Yuji chips are in­sanely ex­pen­sive per lu­men and seem to be on the low end of effi­ciency (ac­tu­ally they seem like such a bad deal that I’m wor­ried I’m miss­ing some­thing). The chip that came out on top in my spread­sheet was this Bridgelux chip which is about ¹⁄₁₀ as ex­pen­sive per lu­men and 2x as effi­cient (but has a larger light-emit­ting sur­face and a CRI of 90 in­stead of 95).

With a low-pro­file, silent fan and effi­cient (120+ lm/​W) emit­ters, you shouldn’t need that much of a heat sink to keep the fix­tures be­low, say, 60° C.

Dis­claimer: I haven’t ac­tu­ally finished build­ing my DIY lamp at­tempt, so this could all turn out to be wrong :)