LED Bulbs

[Brian-M]

Here's something to try while we save up to buy LEDs.
Bring your own Bottle.
http://www.bbc.co.uk/news/magazine-23536914

For those who can't be bothered to click the link, it's about someone who came up with the idea of using making skylights from discarded plastic drink bottles. You fill the bottle with water to refract the light (and a dash of bleach so that algae doesn't grow), drill a hole in your roof, shove the bottle halfway through the hole and glue it in place.

 

[Olowkow]

For those who can't be bothered to click the link, it's about someone who came up with the idea of using making skylights from discarded plastic drink bottles. You fill the bottle with water to refract the light (and a dash of bleach so that algae doesn't grow), drill a hole in your roof, shove the bottle halfway through the hole and glue it in place.

Did anyone understand the significance of "works best with a black cap" or using a black film can? I don't see why that would make any difference.

 

[Soapy Sam]

For those who can't be bothered to click the link, it's about someone who came up with the idea of using making skylights from discarded plastic drink bottles. You fill the bottle with water to refract the light (and a dash of bleach so that algae doesn't grow), drill a hole in your roof, shove the bottle halfway through the hole and glue it in place.

A couple of years ago I actually investigated fitting one of these-
http://www.esi.info/detail.cfm/Monodraught/SunPipe-domestic-light-pipe/_/R-26140_HP123SE

at home, as we have little southerly light and parts of the house are quite dark in daytime, especially winter. I concluded the effectiveness was too low to justify the cost and associated installation upheaval- easier to install a window. Moser's version is so much cheaper and more elegant you have to admire it. Pretty useless in Scotland , admittedly.

Back to the LEDs , I guess.


Olokow- No, that puzzled me , too. You would expect maximising the clear area is the way to go. Most drinking water bottles have blue caps in my experience. Maybe a translucent cap just gives an objectionable colour cast?

 

[marting]

Our eyes evolved to see black body radiation style spectra (actually, just one spectrum, that of the sun). For millennia, the only artificial light we had was black body radiation of various color temps.

Analog photography was developed with black body radiation spectra in mind, because it was the only game in town.

Spectra of FLs and LEDs are different, and new. But, the way I see it, they're here to stay, LEDs in particular. As a photographer, you have to evolve to deal with that. Digital photography can deal with the color changes quite nicely; even cheap cameras have a setting for FL lighting, and you can adjust the colors in post. Of course, when you present your work, you have to make the print with the lighting at the showroom in mind. You always had to that, BTW. It's just that in the past all spectra were black body, and the color temp and its effect on the colors of the photo could be calculated. FLs and LEDs are more complicated, as they may have vastly different spectra that are not easily calculable with a simple number like color temp.

That brings up another complication.

When a camera takes a pic of a scene illuminated by CFLs or other fluorescents is the desire to accurately render (apart from color temperature) colors as illuminated by CFLs or adjust them to render colors that would look approximately like the scene would be rendered in natural light? This adds to the question of how a rendered print will be viewed since it might be desired to also render a print that most accurately represents a scene viewed in natural light even though it is illuminated by fluorescents.

There are techniques to deal with all of these but they are complex enough that most photographers avoid them, let alone realize how these interact. www.color.org has a lot of tech info related to these issues.

 

[Brian-M]

Did anyone understand the significance of "works best with a black cap" or using a black film can? I don't see why that would make any difference.

I assumed that it was so that you don't see distorted refractions of the regular cap from inside. You'd just see a dark blur instead, which I guess might be more aesthetically pleasing.

What has me wondering is how long they'll last. UV light breaks down the polymers in plastic, making it brittle. Eventually cracks will form.

Moser's version is so much cheaper and more elegant you have to admire it. Pretty useless in Scotland , admittedly.

Pretty useless in any building with a ceiling as well, it'd just illuminate the empty roof-space.

 

[BenBurch]

It is worse than that. Here is something all good photographers know.
1. light an object using only CFLs.
2. Take a photo without flash.
The colours in the photo will appear wrong. Even though our eyes tell us the colours appear normal. May work with other types of bulbs such as LEDs.

But, this can be fixed; http://edspstuts.blogspot.com/2011/05/quick-and-easy-color-correction-in.html

 

[Delvo]

...

 

[marting]

But, this can be fixed; http://edspstuts.blogspot.com/2011/05/quick-and-easy-color-correction-in.html

The optimum way to do this is with an icc profile generated from the same camera image using a known color chart such as the x-rite ColorChecker SG. The chart's patch colors, which are measured and known to the profiler software, are then mapped and a 3D interpolation from this is used to map other pictures taken with the same illumination to a color corrected image.

It isn't perfect for many reasons mostly because a camera normally has only three color filtered sensors.

There are high end setups museums use to very accurately capture colors of expensive artwork using more than 3 filters to capture and estimate the spectral reflectivity of the artwork.

 

[elgarak]

That brings up another complication.

When a camera takes a pic of a scene illuminated by CFLs or other fluorescents is the desire to accurately render (apart from color temperature) colors as illuminated by CFLs or adjust them to render colors that would look approximately like the scene would be rendered in natural light? This adds to the question of how a rendered print will be viewed since it might be desired to also render a print that most accurately represents a scene viewed in natural light even though it is illuminated by fluorescents.

There are techniques to deal with all of these but they are complex enough that most photographers avoid them, let alone realize how these interact. www.color.org has a lot of tech info related to these issues.

Absolutely.

However, that is not a new thing at all. There were test charts and color charts around way before FLs or LEDs or digital photography were developed.

In fact, I would go so far to say that things gotten easier, since olden days analog films had additional problems: You wouldn't know beforehand how a batch of film would react. I recall a lot of empirical knowledge, things like films from one company would react so-and-so, and films from the other company such-and-such. Then company X changed the formula, and photographers and filmographers had their hands up in protest.

 

[BenBurch]

Absolutely.

However, that is not a new thing at all. There were test charts and color charts around way before FLs or LEDs or digital photography were developed.

In fact, I would go so far to say that things gotten easier, since olden days analog films had additional problems: You wouldn't know beforehand how a batch of film would react. I recall a lot of empirical knowledge, things like films from one company would react so-and-so, and films from the other company such-and-such. Then company X changed the formula, and photographers and filmographers had their hands up in protest.

Reciprocity failure and the attendant color shift was always what got me with film. I did a lot of low light work.

 

[marting]

Reciprocity failure and the attendant color shift was always what got me with film. I did a lot of low light work.

Always a problem with low light film.

Digital photography has had to adapt to the peculiarities of film like the "s curve" which results in more contrast and color saturation in the midrange at the cost of the same in highlights and shadows. It's now what people expect and is fairly easy to do digitally and is built in in most camera jpeg image processing. There are even folks offering various alteration techniques, some working better than others, to achieve the particular color characteristics of different film/ print processing. For those of us doing repro work the intrinsic linearity of digital sensors is a godsend though we have to work in RAW mode with non-linear (outside of gamma which just helps with dynamic range) s-curve effects removed.

For regular photo work "pleasing" color prints is paramount and that translates into the kinds of non-linearities people are used to in prints with their higher color saturation amongst other things. Luckily it's far easier than it used to be to achieve either of these goals.

 

[Soapy Sam]

Pretty useless in any building with a ceiling as well, it'd just illuminate the empty roof-space.

I doubt ceilings are part of his target market.

 

[marting]

Where I live LEDs with their long life, even at their high initial cost are pretty big cost savings. They even save compared to the more closely efficient and much cheaper CFLs. Juice is pretty expensive here, just over $0.30/kwh. Even makes solar panels financially attractive.

And I hated dragging out a ladder to replace the overhead bulbs.

 

[Soapy Sam]

Where I live LEDs with their long life, even at their high initial cost are pretty big cost savings. They even save compared to the more closely efficient and much cheaper CFLs. Juice is pretty expensive here, just over $0.30/kwh. Even makes solar panels financially attractive.

And I hated dragging out a ladder to replace the overhead bulbs.

I believe it's about 16 pence here.
(0.16 * 1.55 = 24.8 cents). Good god, we have something cheaper than you?

 

[wobs]

LEDs arrived.
They are brighter than I thought they would be.

Its been difficult to gauge what strength I needed, as the fittings are recessed in the ceiling. The 4 reflector bulbs have little spread, which gives a lot of light under them.

However, the LEDs are longer, and GLS bulbs, and so have a wider spread of light.

As I previously tried a 600ln LED from Ikea, and found it a bit lacking, I tried 1000ln bulbs. These are actually brighter than the old halagons.

I think 800ln would have been better in hindsight, but it has been difficult to judge until you actually get them.

They are Samsungs, and have not buzzed (I've read reviews of some that have). The light is fine for me, but I'm not too fussy about that, as long as the level is right and uses the minimum power, I'm happy.

 

[Dr. Keith]

As a side note:

It also is worthwhile to reconsider how you install the lighting in your home, when using LED's. After all, LED light bulbs that fir in regular sockets are more of an compromise, to make the switch-over easy. But they are far from optimal when it comes to really useing the benefits of LED's.

Those bulbs (and by extension, high powered LED modules like the one i mentioned) are meant to produce a lot of a light in a single place, in the hope it spreads out into whatever you are going to illuminate. But there are rather easy to implemet options, if you are willing to depart from that philosophy.

Just as a quick example, if you have rudimentray skills to build your own stuff. Use L or U shaped profiles, thin ones. Mount them on your walls, a few cm below your ceiling. Then place flat LED strips in them, shining upwards to your ceiling. The profile is there so you never look directly into the bright LED's. That way you get a nice, indirect lighting in your room.

There is virtually no way to do that nicely with incadescent lamos, due to the heat. Even fluorescent tubes are limited here, because they are rather thick and need a bulky ballast to operate. With LED's you can make that virtually invisible. If you are into electronics (or get the proper constant-current supply) you can make it dimmable.

And heck, you don't even need to use white LED strips for that. Use RGB LED strips. They cost a bit more, but that way you can add some nice colour to your ambient lighting.

Really, i think that traditional lighting concepts are not really suitable for LED lighting. Not because they wouldn't work, they do, but because those traditional concepts are limited by the physics of regular bulbs (= heat, big size and high current consumption in low voltage applications). LED's allow _much_ more freedom in the way you arrange them.

My hope is that some day organic LED's will have a better lifetime expectancy and become cheaper. With that technology you can literally have your wallpaper _be_ the light source instead of just reflecting light from some other source. But i guess that will be quite a long way to come...

Greetings,

Chris

I was wondering if you could point to some sources for elaborating on the bolded above.

I'm in the very early stage of designing a space that will need very efficient lighting, so I know I will be using primarily LED. I would like to design the lighting layout to get the most from this type of light instead of using a typical incandescent layout and then dropping in LED replacements.

It is a project that will not likely be built for a few years, but I don't think I can count on LED wallpaper.

Thanks!

 

[Dr. Keith]

LEDs arrived.
They are brighter than I thought they would be.

Its been difficult to gauge what strength I needed, as the fittings are recessed in the ceiling. The 4 reflector bulbs have little spread, which gives a lot of light under them.

However, the LEDs are longer, and GLS bulbs, and so have a wider spread of light.

As I previously tried a 600ln LED from Ikea, and found it a bit lacking, I tried 1000ln bulbs. These are actually brighter than the old halagons.

I think 800ln would have been better in hindsight, but it has been difficult to judge until you actually get them.

They are Samsungs, and have not buzzed (I've read reviews of some that have). The light is fine for me, but I'm not too fussy about that, as long as the level is right and uses the minimum power, I'm happy.

Ours seemed "too bright" at first, too, but either they dimmed a shade over time or we just got used to them.

 

[marting]

I believe it's about 16 pence here.
(0.16 * 1.55 = 24.8 cents). Good god, we have something cheaper than you?

Tell me about it. One of the fallouts of Enron's shenanigans.

I have a property in Texas where electricity only costs $.04/kwh. I wish I could import it.

 

[Soapy Sam]

Opposite side of this argument is that if your power bill is mainly for winter heat, running tungsten bulbs as heaters gives you both light and heat, so LEDs may not be such a boon as you expect.

They are expensive and at 58 years of age I'm less than 100% confident I need bulbs with a 20 year lifespan...

I tried to fit LEDs in our bathroom, but despite claims of compatibility from the manufacturer of the light fittings, they did not , in fact, fit. The bulbs have external ribs, presumably for cooling, though it's hard to see why that's necessary, - and the ribs would not fit the recessed downlighter fittings. So although advertised as GU10 bulbs, they did not fit all GU10 holders. I shifted them to my computer room, where the lights are on far longer anyway. They certainly give a bright light- not like the disastrous CFL crap the industry (and EU government) has been foisting on us.

The "savings" advertised do not always make sense though, unless you leave lights on most of the day- as in shops or offices. I reckon it will take nearly five years of use to recover the costs on LED bulbs (as I use them)- and if I drop just one and break it, that time goes way, way up.

If price comes down, they will take off, but we're not there yet.


ETA- One variable I'd like to know more about is decreased efficiency over time. I have no idea if this affects LEDs significantly or at all.
Some CFLs (even from good brands) dropped by 50% in output in as little as six months. I have the photos to prove it.

#

Postscript.
I installed those bulbs sometime around the end of April last year.
Between May and September they would see very little use. Given that I worked 28 / 28 on drilling rigs and was therefore away from home one month in two, I think it likely that they had seen at most 1000 hours of use.

One failed on February 28th and a second died today.
In each case , one of the three LEDs in the unit failed and the two survivors went into a flashing mode which meant the bulb had to be removed immediately.
I assume this implies more than a failure of a single LED, but failure of some circuitry within the bulb.
At almost £15 a time, I consider this totally unacceptable.
The bulbs concerned are Sylvania ES 50 350 lumen models.
I have emailed Sylvania UK's customer support, asking for comment.
I will post any reply here.

In the meantime, I strongly urge anyone not to install these bulbs.

 

[a_unique_person]

#

Postscript.
I installed those bulbs sometime around the end of April last year.
Between May and September they would see very little use. Given that I worked 28 / 28 on drilling rigs and was therefore away from home one month in two, I think it likely that they had seen at most 1000 hours of use.

One failed on February 28th and a second died today.
In each case , one of the three LEDs in the unit failed and the two survivors went into a flashing mode which meant the bulb had to be removed immediately.
I assume this implies more than a failure of a single LED, but failure of some circuitry within the bulb.
At almost £15 a time, I consider this totally unacceptable.
The bulbs concerned are Sylvania ES 50 350 lumen models.
I have emailed Sylvania UK's customer support, asking for comment.
I will post any reply here.

In the meantime, I strongly urge anyone not to install these bulbs.

I have plenty of led bulbs, no failures yet. They cost a lot less than that. I would demand a refund.