A Little Weird Science -Looking For A Soft LED Lighting Color That Dims Sufficiently

IMG 0897

We are always looking for a better, simpler way of doing things, and this frequently includes in-house testing. Right now we are working on LED lighting, looking for the best combination of light, color temperature, and dimming ability.

At our official Pacific Northwest lab, Mark Fritzer put together the board in the first photo to test a variety of spot lights, rope lights, and dimming systems.

IMG 0891

This allowed us to compare various lights and and test with colored plastic for extra measure.

IMG 0894

One of the tests involved reading lights. This photo was of a Quick stalk light illuminated label held at 24”/60cm from the light fixture.

IMG 0888

The photo does not capture the lighting subtleties, but after much subjective analysis – we found our favored, in both the overhead, stalk, and strip lighting.

Stay tuned. There’s more to follow.

For more information on the FPB series, contact Sue Grant: Sue.Grant@Berthon.Co.UK.

Posted by admin  (May 1, 2014)

16 Responses to “A Little Weird Science -Looking For A Soft LED Lighting Color That Dims Sufficiently”

  1. Matt L Says:

    Setsail Team. I have spent a couple of decades in a business heavily influenced by the vagaries of “color science”. Spectrophotometers were at one time extraordinarily expensive. Now hand help spectro’s can be purchased for less than a $1,000 USD and in some cases much less than $400 USD. It is intriguing at times to use a hand held spectro configured to measure ambient light and take a look at the spectral qualities of various light sources. What you can do is find a light you think is pleasant, measure its spectral data and then compare that to other light sources that are not as pleasant. You can then check to see how the spectral response of the light changes as you dim it. The quality of the light might be fine at a medium intensity but be to cold (blue) at higher intensity.

    A spectro can be useful in evaluating whether a light source has a significant UV component which may cause fabrics to fluoresce in the presence of optical brighteners.

    Another nice trick would be to evaluate a light sources Color Rendering Index to avoid the ugly effects of metamerism. Nothing is worse than having your wife dress in 4300K light fluorescent light with everything matching only to have her walk outside in D65 light where nothing in the outfit matches. Of course it’s your fault at that point…but I digress. A quick rule of thumb is any light source with a CRI index of 92 or higher will render color very well. The ability to measure CRI is in most of the low end ambient light reading spectros.

    The CIE Metamerism Index is a newer (and theoretically superior) means of measuring metamerism, but it would not necessarily be required for this application.

    My apologies if Team SetSail is well versed in spectrophotometry, color temperature, UV excitation (actually it’s a bummer), CRI, and metamerism.

    Metamerism, which is when an object appears to be a different color under different light sources, should not be confused with mammarism, which is when the right one is a different size or shape from the left one. I had the pleasure of attending a color management conference where a colleague spoke in great detail on the subject of “mammarism.” This led us all to believe he had a much different experience the night before in his hotel room jacuzzi than the rest of us. So amused and enthralled we were with this most pleasant of malapropisms, we allowed him to continue for some time before someone mercifully pointed out that he should correct his pronunciation of “metamersim”. It was beers all ’round at the conference social that evening.

    Best Regards,

    Matt L

  2. Todd Rickard Says:

    Thanks for the input Matt. I am told Metamerism has been a significant component of LED selection by the design team at Westport Yachts. I am trying to see if they are willing to share any data from their testing. Our initial tests revolved around parameters that have yet to take this fully into consideration – but we understand its importance.

  3. Pedro Cuso Says:

    Hi Steve, as you know one of my Companies is a LED lighting company.
    Regarding color temperature is a matter of taste. 2600-3000ºK will give you a warm atmosphere, but for reading or work I prefer 4000-4500ºK.
    Regarding diming I recomend you to work with DC 12V, much easier to dim.
    AC 110 or 220/230V LED lights ussually have not the internal power suply for diming, is ussually switched P.S. and prepared to work on a wide range of input voltage with constant current or voltage output.
    For strips I will recomend you rigid strips due the back board is ussually aluminium and helps with the heather generated. Obviously DC.
    For strips that will make indirect lighting on the ceiling or in the floor as much leds as possible up to 120 led/m it will give you a much more homogeneous light without spots like the 30 or 60 led/m.
    Consider also the use of AR/QR 111 for sleeping rooms galey and lavatories instead of MR16.
    Are you thinking on some centralized control unit? It will allow you at the touch of a button to recall lighting scenarios.
    Any think I can help let me know.

  4. Steve Dashew Says:

    Thanks Pedro. We are looking at two different lighting regime – mood and task as youu have pointed out. We’ve only been working on the mood component so far as that seems the hardest to achieve. What we’d like is the equivelent of a substantially dimmed quartz halogen bulb, but in an LED. So far no luck. Suggestions aresolicited from one and all.

  5. Pedro Cuso Says:

    Hi Steve,
    Unfortunately what you are looking for is very difficult and or complicate.
    On quartz halogen the light is coming from an incandescent filament. When you dim current drops and temperature on the filament as well. When you dim on a incandescent light not only you get less lux also the colour temperature decrease and light became warmer.
    When you dim a fluorescent, with apropiat balast, you only get a decrease on lux but not any change on colour temperature.
    The white led works diferent but like a fluorescent tube.
    There is a phosphor coat on a blue led. The blue beam excites the phosphor and this one is giving the white light.
    Depending on the blend on the phosphor you will get warm neutral or cold white. But a decrease on the current in the led just will give lower excitation on the phosphor and less lumens but not any change on the colour temperature.
    However there is another way to make white light through rgb led. It means you need 3 led red blue and green.
    Then you need the control unit as a dmx one to adjust it.
    But I thing it goes against your “keep as simple as posible” theory that I share.
    If you want to test on rgb let me know

  6. Pedro Cuso Says:

    I will like to know if you are diming from AC or DC. If diming from DC it can be a solution for the strips, or two.
    We can dim with a double switch and diferent resistors for a 4000°K strip and a 2400°K strip. You will have a steeping dimer, but will get the effect you look for. Inconvenient is the residual heat from resistors. Any way power involved is very low. You will have constant current drain independently on the dim position. Other solution can be two dimer working with a double counterpointed potentiometer and 2 variable resistors to trim it. For halogen retrofitting I have no notice of any one with referred colour temperatures on same device.
    If wide beam (about 120 to 140°) on the halogens is not a problem (typically are around 30° what is a spot light) will be easy to make a custom ones. But in case of burning you will not find spare anywhere.

  7. Pedro Cuso Says:

    Regarding CRI almost any white LED device will give you values over 80 what is very right. Any regular tube T8 or T5 are on this range and all them fully aproved for office environment with the restringent regulation about lighting on work place. Over CRI 90 is considered “technical lighting” and used only in very specific and special environment like jewelery press test rooms photography special shops etc. What I understand about what you talk is related to colour temperature and not about colour rendering index.

  8. Hugh C Says:

    As I was reading your article above, I was thinking to my self, self, you need to comment on Spectrophotometers and CRI…. Thank you Matt L for getting to it before me! I believe it’s vital that you continue your research with these things in mind. LED’s, although clearly the way of the future, tend to have a pretty low CRI value and in my view aren’t quite ready for prime time yet (it won’t be long, though!). As with so many things, this is a “you get what you pay for” situation: the more expensive LED’s often have a much better quality of light. I look forward to hearing the results of your research.

  9. Matt L Says:

    Steve, I rant across this article surfing this Fathers Day afternoon. http://www.cnet.com/pictures/comparing-crees-100-watt-replacement-led-with-the-competition-pictures/. It has a few images that give some insight into CRI and color temperature.

  10. Anders M Says:

    Have you thought about sideemitting and end-emitting fibre optic lighting? not only can the light emitters be whatever you want them to be (Halogen, LED etc.), but are easilier dimmed and can have rotary colour wheel. I was in a home with a complete setup, utilising both HID,ordinary halogen, as well as LED emitters and the lighting was excellent. They mostly used solid core (i.e. not strands) fibre optic cables for their stuff.
    A bonus of fibre optics is that the place where it lights up does not get warm as all the warmth generated is from the emitter itself. Further, you can run the optic cables (be it side emitting or end emitting) to places that see a lot of water. Hell, you can have it in the floor in the shower compartment if you wish. The house I was in, had it vertically in the shower compartment as well as in the ceiling of same. The emitter itself was placed in a cupboard in the adjoining hallway.
    Anyway, just an idea, perhaps a bit off the wall.

  11. Steve Dashew Says:

    Sounds fascinating Anders. Any suggestions for sources of fiber optic side emitting lighting?

  12. Anders M Says:

    I’m pretty sure I responded to you (Steve) with a proper post with links and all, but maybe it didn’t go through because of the links?

  13. Steve Dashew Says:

    Try again Anders – Thanks

  14. Anders M Says:

    Okay, I will try again.

    I found some Australian companies, thinking they are closest to NZ, but have a larger variety.
    The first company is the one which seems the best (when reading about it – never dealt with any of them), and has the greater selection.

    And here is their solid core cables:

    And their light emitters:
    They have a 36W LED light emitter, which should put out enormous amounts of lights.

    And then there are these two company:
    (if for nothing else than for a few pictures.

    Personally, I’m not into “star ceilings”, I find them ridiculous, but using fibre optics for “proper” lighting seems like something worth considering. It has limitations, but it also opens up a new way of thinking light.

    In general, what I call “light/LED/HID/Halogen emitters” are also called “illuminators” and “light drivers” when you do your own searches.

  15. Steve Dashew Says:

    Thanks Anders – we will check this out.

  16. Anders M Says:

    I forgot to mention this time around, that you can get “fittings” such as angles to make the fibers run around sharp corners, or go from a chasing to a fixture. There are all sorts of end fittings/fixtures as well, so you could have a nice light for a bed lamp, for a painting and so on. And with fibre optic lighting, there is no UV being transmitted to you painting, which isn’t all-important for most, but a great little bonus.
    Some light emitters (even LED-emitters) comes with their own fan, while others are fan-less by design.