Changing LED Tint With Filters

[Anders Hoveland]

I wonder if we will eventually see retrofit LED bulbs with colored filters for residential use, much like the filtered coatings already being used to modify incandescent light bulbs (such as Chromalux and Reveal). Maybe some people just do not really like the light from LED bulbs, and by filtering it the light illuminating the room could be subjectively made more pleasant.
http://www.meijer.com/assets/product_images/styles/xlarge/1001029_046677135591_A_400.jpg


Stating the obvious again, but filters may affect the color of LED light sources differently than other types of light sources. Each type of light source tends to have its own unique frequency profile, and these frequencies will uniquely match to the frequency being absorbed in the filter. It is not just about "color".

For example, neodymium glass appaears purplish under incandescent and LED light, but appears dull bluish grey under typical fluorescent lighting. The fluorescent lamp gives off discrete frequencies, and the main yellow frequency absorbed by neodymium glass is not one of those.

So some filters may result in different types of color changes depending on the type of light source. Since this thread is about LEDs, we should note that typical white LED light has deficiencies in deep red, blue-green, and violet frequencies. So filters that absorb these frequencies might not affect the LED as much. Also, because LED has a relatively narrow blue spike, there is potential for filter frequency mis-match. A filter designed to filter out blue light (and thus appear orange) might not work as well on LEDs. If the filter is filtering out multiple frequencies, the ratio of color filtering will be off.

Specifically, many red filters may significantly reduce the efficiency of LED flashlights, since LED light actually only has a limited ammount of red frequency light, and much of the light that is being perceived by red color receptors in the eye is actually orange frequency. It would be especially difficult to really get a true red color (that is not somewhat orangish) using a filter over a regular white LED flashlight.

Interesting... phosphorous filters...

I suspect that a phoshor coating may actually be used in a certain type of incandescent bulb, in Osram-Sylvania's double-life halogen with "modified spectrum technology". There's something about the light from these bulbs that I just do not like. It's definitely not the same effect as neodymium glass. The outer bulb is a barely noticeable slightly pink color, so I think it might involve some type of phosphor rather than the usual colored filter coating. Very interesting. Presumably some of the violet and blue light (even some of the near ultraviolet being emitted by inner halogen capsule) is being converted to red light. The intended effect apparently was to try to lower the correlated color temperature, in an attempt to mimmic the feel or normal incandescent bulbs, while still getting the higher efficiency of the hoter halogen filement. I am speculating a little here, I really cannot find any information about what exactly this "modified spectrum technology" advertized on the packaging really is.

http://www.lighting-gallery.net/gallery/albums/userpics/11216/normal_SANY3047.JPG

 

[uk_caver]

Maybe some people just do not really like the light from LED bulbs, and by filtering it the light illuminating the room could be subjectively made more pleasant.

Maybe some people like that should spend time checking out LEDs to find ones they do like the colour of.

In the long run, there seems likely to be a meaningful market for colour-tunable LED lights for mood lighting for normal people, though obviously that would be easier if there were some economic standard control system for such devices.

Stating the obvious again, but filters may affect the color of LED light sources differently than other types of light sources.
[...]
Specifically, many red filters may significantly reduce the efficiency of LED flashlights, since LED light actually only has a limited ammount of red frequency light, and much of the light that is being perceived by red color receptors in the eye is actually orange frequency. It would be especially difficult to really get a true red color (that is not somewhat orangish) using a filter over a regular white LED flashlight.

Not only is that obvious, but I think it's been adequately addressed earlier in the thread.

I suspect that a phoshor coating may actually be used in a certain type of incandescent bulb, in Osram-Sylvania's double-life halogen with "modified spectrum technology". There's something about the light from these bulbs that I just do not like. It's definitely not the same effect as neodymium glass. The outer bulb is a barely noticeable slightly pink color, so I think it might involve some type of phosphor rather than the usual colored filter coating. Very interesting. Presumably some of the violet and blue light (even some of the near ultraviolet being emitted by inner halogen capsule) is being converted to red light. The intended effect apparently was to try to lower the correlated color temperature, in an attempt to mimmic the feel or normal incandescent bulbs, while still getting the higher efficiency of the hoter halogen filement. I am speculating a little here, I really cannot find any information about what exactly this "modified spectrum technology" advertized on the packaging really is.

Seriously?

I just Googled 'sylvania modified spectrum' and the first hit was http://www.google.com/patents/US20130063016 which goes into great detail about various types of modified spectrum bulb, including what the official meaning of 'modified spectrum' is.
Research doesn't really get any easier than that.

 

[Anders Hoveland]

Maybe some people like that should spend time checking out LEDs to find ones they do like the colour of.
there seems likely to be a meaningful market for colour-tunable LED lights for mood lighting for normal people.

Filtering is not just about the overall color, it also effects the spectrum. Filters would be able to achieve spectrum changes that LED phosphors by themselves are unnable to accomplish. A combination of filters could even potentially be used to increase CRI of LED light. For example, if the spectrum needs a higher ratio of red and blue-green frequency light, more Nd:YAG phosphor could be used, and then some of the excess orange and yellow frequency could be filtered out.

 

[uk_caver]

Filtering is not just about the overall color, it also effects the spectrum.

Any changing of colour affects the spectrum.
It's just that with electronically tunable lights, the overall output spectrum is a linear combination of two or more LED spectra, whereas if using filters, the overall output spectrum is the overall LED output spectrum modified by multiplying it with the filter's transfer function.

Filters would be able to achieve spectrum changes that LED phosphors by themselves are unnable to accomplish. A combination of filters could even potentially be used to increase CRI of LED light.

Wouldn't any change to output colour by electronic colour tuning also be likely to have an effect on CRI, and wouldn't the effect of some changes be to increase CRI as well as having some changes reducing it?

For example, if the spectrum needs a higher ratio of red and blue-green frequency light, more Nd:YAG phosphor could be used, and then some of the excess orange and yellow frequency could be filtered out.

Assuming it was actually thought worth incorporating it as a design goal (ie it wasn't just something thought likely to appeal to an insignificant minority of real-life consumers, or some theoretical 'need' without actual supporting evidence), it would be possible to make electronically changeable LED sources with additional LEDs to boost various areas of the spectrum without necessarily compromising efficiency the way that significant filtering would.

Possibly some such devices could be useful, even if only as experimental units to see if quietly changing the extent to which gaps were filled was something which people really noticed or cared about.

Have you read the link regarding the secrets of modified spectrum technology?
I found it rather interesting.

 

[bshanahan14rulz]

1.4% CrSn pink, 0.2% red iron oxide


Sounds like a filter.

Also, I was under the impression that a phosphor layer on an LED was a combination of various fluorescent materials, not a single one. I think this is the way to go, rather than wasting more energy on downconverting photons just to filter them out. Although I think I am lost with regards to what we're debating here, but then again, I'm just a dancing monkey.

 

[Anders Hoveland]

Also, I was under the impression that a phosphor layer on an LED was a combination of various fluorescent materials, not a single one.

It can be, but standard white LEDs just use a single phosphor, Nd:YAG. Apparently the Nd:YAG can be tweaked to emit more into the green (EQ-white), but I am not really sure how this works.

I think this is the way to go, rather than wasting more energy on downconverting photons just to filter them out.

Ideally, yes. But in reality, there is a limited number of practical phosphors to choose from, ones that will not degrade over time under the intense light exposure.

 

[uk_caver]

Anders,
Have you read the patent description about the modified spectrum technology you were so interested in, and which you had been unable to find any information about?

Given the several seconds of work it took me to locate it, I'd hope you'd find it enlightening.

 

[bshanahan14rulz]

I think Nd doped YAG emits in the IR region. Assuming you mean Cerium. I think all the YAG does is provide a crystal with specific optical properties for the dopant ions to live in. Also, if all LEDs use one single phosphor, how do we get LEDs with a reddish tint or a greenish tint, when the phosphor only emits yellow? I would think that there would be many suitable practical phosphors to choose from based in sturdy glass or garnet. I understand that they are pumped differently, but fluorescent tubes have different phosphor "mixes," so I don't think it's much of a stretch to think that LEDs use a mixture of various fluorescent phosphors too.

Once we get LEDs efficient enough, who cares if we apply filters to "tone down" some of the spikes to make the rest of teh spectrum more noticeable? And if we can just add some new phosphors to boost the deficient parts of the spectrum, even better! I think filters are never the right solution when you are looking for efficiency of a system, but I think you are more interested in obtaining a fuller, more even spectrum spread. Sometimes you can't have cake and eat it too :-/

 

[Anders Hoveland]

Have you read the patent description about the modified spectrum technology you were so interested in, and which you had been unable to find any information about?

I had already found that, but I have doubts that it is the same "modified spectrum technology". "Modified spectrum" is a very generalized term. That patent did not specifically refer to one filter type.

This reason why I suspected the bulb may have a phosphor coating inside is because, holding the bulb up to a light source, transmitted light thrugh the bulb seems to only have only a very slight pink tint, almost not noticeable. But when the bulb is actually turned on, the pink tint is much more pronounced. In contrast, a regular incandescent bulb has a more orangish tint, which I generally do not mind. The pink tint just bothers me for some reason. Strangely, the pink tint is somewhat reminiscent of the eery pinkish light from "warm white" fluorescents.

I am not sure, perhaps you are right and I am just speculating too much. It would be interesting to analyze the light with a spectrometer to try to confirm if my suspicions are correct. If there is a little spike in the spectrum, it is a phosphor, if there is a depression, it is just an ordinary filter.

 

[uk_caver]

Sure.
You'd found it, you'd read it, and you'd understood that 'modified spectrum' was a term relating to incandescent bulbs being tinted rather than 'coloured' and being allowed to be less efficient than regular bulbs as a result of using filters.

Then you'd concluded that a particular Sylvania modified spectrum bulb might be using a phosphor because it didn't look like neodymium-tinted bulbs and it looked pink, whereas, on the other hand, the referenced patent which you'd already read was talking about bulbs which didn't use neodymium tinting and which instead used a coating which was likely to appear pink. Obviously quite different.

And, after all, if a phosphor was being used to boost halogen efficiency compared to using a filter, it's understandably not the kind of thing a manufacturer would try to make a big deal of (or even mention (or even patent)), given how no consumers or regulators really care about light-bulb efficiency these days.

 

[bshanahan14rulz]

or you could bust the bulb and pump it with various different monochromatic sources and watch for fluorescence.
Halogen bulb and a long life incan are both filament bulbs, but with very different light coming out of 'em. Perhaps there is a halogen capsule inside of the bulb, so less reds produced + blues being filtered out by the pink and red particles may create efficacy of halogen bulb with more reds of standard incan.

 

[Anders Hoveland]

After seeing the new Cree LED retrofit, which has a glass bulb, I got to thinking, how would it affect the light if they had used neodymium glass? Would the effect be the same as with a regular incandescent bulb? (For those of you who are familiar with Chromalux "Daylight" or GE's "Reveal" bulbs, these use neodymium glass)

Probably it would not be quite the same because LED has a different spectrum than incandescent. It could potentially be used to raise the correlated color temperature without making the blue spike bigger. Yes, it would make the LED spectrum even a little more discontinuous, but nevertheless I have a feeling it might slightly raise the CRI index. Basically, the LED phosphor already has too much yellowish frequency light, and to a much lesser extent the neodymium glass also filters out a portion of the yellowish-green, so by reducing these, it would effectively increase the relative proportions of red and blue-green frequency light in the spectrum. The bandwidth of the yellow frequency being filtered out is also relatively narrow, so it would not result in too much color shift.

Something else that should be mentioned is that the amount of filtering can make a difference. For example, with neodymium glass (at least using an incandescent source), a lesser amount of filtering results in an increase in CCT, while excessive filtering results in a decrease in CCT. This is because, while the neodymium glass readily filters out the yellow, there is also some absorption elsewhere. Once all the yellow has been completely filtered out, additional filtering just starts taking away from the parts of the blue and green. So, if the filter absorbs more than one part of the spectrum, using two layers of filtering can potentially result in a very different color than just one layer. Just something interesting to consider.

 

[Derek Dean]

I've got to say, I'm still amazed at how many threads I read from folks who are unhappy with the tint of their lights and think there is nothing they can do about it, aside from either returning it or trying to figure out (using charts) which LED to order so they can mod it, and then hope the tint of that LED is acceptable.

While I'll agree that a HiCRI LED will generally give the best results (I do love my Nichia 219 modded NovaTac 120P), I've found it possible to get 95% of that performance from my filtered lights, which is pretty darn good in my book.

I use most of my lights every day (er.....every night), both on my job and bicycling home afterward, and the difference the filtering makes is HUGE. Anyway, I'd be interested to hear from those of you who have filtered your lights, what you're impressions are after having used them for a while, and if any of you have HICRI lights, how your filtered lights compare to those.

 

[Cataract]

^^^ +100

I just had a GT last weekend and impressed Cavannus with my filters (I already impressed the other guys last year). I agree that you can get close to 95% of a HiCRI. I just put filters on my 2 work EDCs a few weeks ago, but I only went for reasonnable color rendition rather than trying to imitate a HiCRI (didn't want to loose too much output.) The difference in depth perception is still staggering. I was looking for new EDC's because I wanted neutral, but now it can wait till I find my perfect dream EDC.

 

[riccardo]

@Cataract:

I just got a TM26, it's a quad-xm-l in cool white tint, but I can't really understand which dominant it has. Often it seems pretty neutral but it has the extreme part of the spill with a purplish color, the corona is a little creamy and the hotspot .. Can't really say.

Indeed the light is cold, if I use it to light a room I feel like I'm in a creepy old hospital.. under this light the yellows are changing to a lighter and brighter tonality, reds are also getting lighter and closer to orange side. Anyway these changes are not always evident, the light is quite white, the tint isn't really bad.

Anyway I'm a neutral white guy and I would warm it a bit but I'm having some troubles in getting the Lee swatchbook here where I am. I'll probably have to order some sheets of filters but I don't know which. Can you suggest some from their catalog? Should I get the high temp resistant version?

 

[Derek Dean]

riccardo, the wonderful, and important, thing about getting a filter swatch book is that it contains a large number of filters, many of which are close in tint, but different enough to make a difference. This is what allows us to find the tint that works for us. What might be a lovely tint for Cataract, or myself, might not be quite right for you.

Also, there is enough variance in LED manufacturing tolerances that even LEDs on the same spool can have different tints, so it can vary greatly from light to light. I've even read reports of folks with mulit-LED lights having LEDs of different tint in the same light.

Where are you located? I believe Lee Filters has an office in the UK which serves Europe, but there are also other companies that manufacture filters who offer swatch books.

Even for a multiple LED light you'd want to get a filter swatch book first to find just the right tint, then you could order the somewhat expensive sheet to cover the entire front. The alternative would be to order several swatch books at the same time. Each swatch is large enough for two 1 1/2 inch circles, so, once you find the right tint you could cut out 4 circles and cover just the front of the reflectors.

One thing you could do would be to call a local theater company and see where they buy their filters from, then contact them and ask for a couple of swatch books. Or, call you local photography store and ask if they can help.

By the way, I found the same thing with the XM-L LED in my Zebralight SC600, it has a purplish ring around the outside with a cool, green, cyanish spot. Weird. It took some time playing with the filters, but I finally found one that works well. The tint is still a bit weird around the edges, but not bad, and the beam tint in general is quite nice now.

 

[riccardo]

I'm living in Moscow, mails and small packets from Europe can take up to 8 weeks...!

I'll try to contact the local dealer of lee products but I'm afraid I know their reaction.

 

[Derek Dean]

riccardo, it's worth waiting 8 weeks for these (order 2 swatch books). Once you have them, you'll be able to dial the tint in very closely to want you like, plus, they include many different diffusion swatches. Good luck.

 

[Cataract]

I'm with Derek on this one; for every flashlight I picked a filter for, I ended up using a completely different filter and most often a different color. The process of choosing a filter can take a couple of hours and I am usually very surprised at what works best as it is very hard, if not impossible, to guess which color is the real dominant spike. Trial and error is definitely the only way to go since the wrong filter can make color rendition much worse.

I got my filter packs from a photography store; It cost me the same as ordering online, but I can get a new pack anytime I want and they can order sheets for me, saving me transport fees. I'd suggest you try that first as they might have better delivery times if you still have to order a sample pack. If that doesn't work, the wait is definitely worth it.

 

[riccardo]

well well... it's much better than what I tough.
I called a local LEE dealer here in Moscow and they told me that the swatch books are available for free, moreover the dealer is just on the way to office of one of my colleague's so tomorrow morning he'll stop there and bring me a couple of swatch books.

Watching the LEE website I've seen several filters that according to the description can do the job with my TM26. Anyway it is not clear to me if to improve the "3D" vision and reduce the "alien abduction light effect" it is better to correct the CT or just to remove the dominant (that often are complex)... of course it is a purely academic discussion since the way to go is to try and try and try playing with the swatch book but still I'm curious to understand better.

I also noticed that for each filter they provide the value of light transmitted in percent, does it apply directly to lumens value? For instance, this filter: http://www.leefilters.com/lighting/colour-details.html#249&filter=tf has a transmission of 82,5%, does it mean that with such filter the 3500 lumens of a TM26 will get to about 2900 lumens? If so than a filtered CW LED is not loosing much compared to the same LED in NW...