Broad band LED thoughts

[marspeed]

uv led info http://ledmuseum.candlepower.us/led/ledvio.htm

 

[Foxy Brown]

Just for fun, I overlayed an open ocean absorption spectrum with your LED graph. I'm not sure it's lined up exactly, but it makes the point.

This set of LEDs might be good for corals that live around the 50-75 feet deep range, and would also cause a good level of fluorescence.

Thanks for the comparison, that's big help.

I had this second idea...

For the actual tank I was planning for this LED combo, it will have a drop-off section with the shallow end being 18" and the deep end is 36". Since the light is about right for 50-75 feet, what do you think of putting optics (say 40-60 degrees) on the cyan and blues? This might give a whitish spectrum at shallow depths and bluer spectrum at the deep end because the lights with optics will penetrate better.

The idea is to compress the spectral variation found naturally in the sea to the small depth differences found in aquaria?

Thoughts?

 

[Koyaanisqatsi]

Thanks for the comparison, that's big help.

I had this second idea...

For the actual tank I was planning for this LED combo, it will have a drop-off section with the shallow end being 18" and the deep end is 36". Since the light is about right for 50-75 feet, what do you think of putting optics (say 40-60 degrees) on the cyan and blues? This might give a whitish spectrum at shallow depths and bluer spectrum at the deep end because the lights with optics will penetrate better.

The idea is to compress the spectral variation found naturally in the sea to the small depth differences found in aquaria?

Thoughts?

That's an interesting idea. The optics would push the blues, etc deeper and the warmer colors would spread more. You only want to do it over the section of the tank where it's deeper. And you'll probably still need to adjust the output ratios of the colors differently for the shallow and deep sections.

I envision a setting or knob that shifts the spectrum for deeper or shallower once we get a good idea of how the color mixing works out. It wold be really cool if that adjustment was calibrated so you could literally set it to "35ft deep" and get the color of the sun at that depth as a baseline. Then adjust for preference, such as increasing the fluorescence level or a color you like.

 

[blasterman789]

I do not see halides, well at least the 20K's I used to run producing much of anything in the green, red or cyan ranges and they looked spectacular and my sps growth was excellent

Which is exactly the point I've been trying to make.

My goal has been to try and emulate the color of popular halides with LEDs, not create a different aethestic and then try to justify it.

If our goal was to build a light with as much physical PAR at the expense of everything else we'd just blast our corals with royal blue and standard blue (windex) and be done with it. Corals would be happy but we'd soon get annoyed at our tanks glowing in radioactive actinic blue in a big hurry. Ever seen a tank lit with 10watt 445nm LED's? Christ...you need sunglasses. The actinic reflections are unbelieveable.

Ok, so we know what makes good actinic, and we know what produces high amounts of PAR. Our next task would be to make our tank look like those cool halides, right? So, what color would we add next? It would be a smidge of amber and red to balance out, and some 420nm actinic if we could. However, unless you missed grade-school coloring session, the last color you'd want to add is green.

Look at the graphs Sanjay made of all the halide bulbs here. All the 14 and 20k have pretty much the same spectral slope. Massive amounts of 450nm light, and a smidge of everything else. So basically, we've got it right with royal blue LEDs, (if not a bit deeper which is why I pushed for a 445nm LED). The million dollar question is what to use next, and it's simply not cool white *if* you want to match the look of those halides.

I'll say it again, but green is simply not a color you want in high quantities in a reef tank. It doesn't contribute much to PAR, and it cancels out complimentary colors we find pleasing (purple/pink/red). If daylight balance was an ideal aethestic for reef tanks we could just run high CRI T5's on our tanks at 6500k -vs- cool-white LED based lights at low CRI. I still don't think I've ever run across somebody running high CRI tubes over their reef, nor would want to.

 

[jtma508]

I just want to clarify an important concept surrounding PAR. PAR simply measures the light intensity from 400nm up to 700nm which is where terrestrial plants make their various chlorophylls (as opposed to measuring the entire spectral band). So, what that means is I could build two LED arrays, one with all royal blue and one with all green and generate the same amount of PAR (although there may be different numbers of LED's in each). The latter, of course would not only look bad but would be less helpful to our corals. We should not be as concerned about PAR as we are PUR, which measures photosynthetically usable radiation. The difference between PAR and PUR is that the latter, rather than measuring radiation across the entire 400 to 700nm band, ONLY measures radiation across those wavelengths that corals (zooxanthellae) actually USE for photosynthesis. Unfortunately, to do this we'd need a $4000 scanning spectroradiometer rather than a simple $300 PAR meter.

With those needs buttoned-up it's just a matter of adding whatever helps our tanks look the way we'd like them to. And I also agree that using popular MHs and T5s as models is the best course.

 

[Koyaanisqatsi]

We are not trying to make LEDs look like halides. Halides produce the spectrum they do because of a limitation of HID light. They have a broad spectrum, but with intense peaks in very narrow bands of color. LEDs are improving this problem by providing a light source with a smoother spectrum. A smooth, broad spectrum will illuminate all of the colors properly.

The main challenge we're facing is sourcing and mixing the colors properly. Once we solve that, we can tweak the spectrum for each reefer's preference of coloration in the tank. For example, some like the glowing fluorescent tanks. Other want good color pop, but not glowing. And some want a naturally lit look.

In terms of reef lighting, TONS of R&D has gone into trying to get halide lights to produce a smoother, more complete output spectrum. Their output is a significant compromise between intensity and color quality. We should not be trying to emulate that, but rather go back to what the halide manufacturers are comparing their stuff to: the sun.

 

[blasterman789]

PAR simply measures the light intensity from 400nm up to 700nm which is where terrestrial plants make their various chlorophylls

Partially true. Most general purpose PAR meters read spectrum intensity the inverse as lux meters which are calibrated for the human eye and peak in green/cyan. PAR meters typically have a peak at 455nm, then at about 630nm. Looks like a valley on a spectral graph with green at the bottom, cuz green has little photosynthetic value. Basically, a PAR meter is a 'purple meter' and they aren't as sophisticated as people think. The problem with PAR meters, as I've stated in many other threads, is most read red light just as much as blue because they are designed for terrestrial plant growing. In reefing, 455nm blue is far more important than 670nm red, and I've seen very few PAR meters with a true reef mode, although they do exist. I just use my dSLR and only read the intensity of the blue channel, which is about 455nm, and use that as reference.

We are not trying to make LEDs look like halides.

Maybe you are, but it doesn't sound like you're speaking for everybody. Again, full spectrum T5s exist at 6500 CCT and very high CRI's. If you were to put one of these over a reef tank it would look like a sewer. So, why are we trying to solve a problem when it's not a problem? Just use high CRI daylight T5's.

LEDs are improving this problem by providing a light source with a smoother spectrum. A smooth, broad spectrum will illuminate all of the colors properly.

Again, put a full spectrum, 95CRI 6500k T5 over your tank and you'll stop dead in your tracks when all you're rare morph zoas look the same shade of brown and green-yellow. Your purple acropora will turn dull tan. Also, there is no such thing as a full spectrum LED because market forces want high lumen numbers for advertising reasons and don't care about spectral quality (or CRI).

Theoretically you could add lots of amber and red to a cool-white LED and get a 90ish CRI at about 5000k CCT but it would end up looking just like the broad spectrum T5 with even less efficiency. As soon as you start adding the longer wavelengths to the LED light efficiency will take a huge hit (if that matter here). This is the same reason cool-white LEDs have much better efficiency than warm white LEDs - they have narrower spectral ranges to get uber high Lux readings.

I've tried high powered RGB DMX pars will full range channels over my tank, and it wasn't that amazing. As expected, I just shoved down the green as far as possible. If I were to theoretically add amber to the mix, and let fifty reef addicts adjust the colors to their taste I would guarantee you that none would want a broad and even spectral range. Again, if this were desired, just use high CRI T5's on your tank. For the record, high CRI daylight balanced halides do exist as well, but are just expensive. A plasma light accomplishes the same thing with far greater efficiency.

 

[Koyaanisqatsi]

Again, full spectrum T5s exist at 6500 CCT and very high CRI's. If you were to put one of these over a reef tank it would look like a sewer. So, why are we trying to solve a problem when it's not a problem? Just use high CRI daylight T5's.
...

Again, put a full spectrum, 95CRI 6500k T5 over your tank and you'll stop dead in your tracks when all you're rare morph zoas look the same shade of brown and green-yellow. Your purple acropora will turn dull tan.

Go back and reread my last post. Second paragraph.

I probably started a bad angle by saying we are trying to improve on the output of halides. Actually we are trying to improve on the output of LEDs, in terms of personal preference. In order to do that, we must start by targeting a common baseline - the sun and how it's affected by seawater. Then we can adjust for personal preference. (Is there an echo in here?) And in this thread, we can discuss coral health as well.

 

[Foxy Brown]

Look at the graphs Sanjay made of all the halide bulbs here. All the 14 and 20k have pretty much the same spectral slope. Massive amounts of 450nm light, and a smidge of everything else.

You are completely wrong here. Go back again and look more closely. If you want an idea of how important a peak is you must look at its integrated area, not its peak intensity. The blue peaks are often very impressively tall but they are also EXTREMELY narrow (typical of gas emission lines - mercury "i" line in this case @ 465nm). Because of the very narrow bandwidth, that peak is not nearly as important as it looks. The "smidge" of background peaks that you see filling the entire rest of the visible spectrum actually contain more energy combined than that "massive" blue peak, and it is energy that figures into the color, not peak intensity.

Also, There is a HUGE difference between an RGB led and a broad band light source. Talk to people in theater who use RGB lighting all the time and they will tell you the same.

Again, put a full spectrum, 95CRI 6500k T5 over your tank and you'll stop dead in your tracks when all you're rare morph zoas look the same shade of brown and green-yellow. Your purple acropora will turn dull tan. Also, there is no such thing as a full spectrum LED because market forces want high lumen numbers for advertising reasons and don't care about spectral quality (or CRI).

I'm not advertizing the use of a 5000-6000k spectrum, as like you keep bringing up. Of course people wont want that. Please don't compare "broad band source" to "low color temp" or "full spectrum" those terms do not mean the same thing and that argument is specious. I would still prefer an apparent 14k color temp, but that DOES NOT mean eliminating all other colors..

So basically, we've got it right with royal blue LEDs, (if not a bit deeper which is why I pushed for a 445nm LED). The million dollar question is what to use next, and it's simply not cool white *if* you want to match the look of those halides.

First there is no real difference between 445 and 450nm, when the source has a 30nm bandwidth. A new led at 445nm is completely silly. You're eye can't see the difference either.

If all you want is blue and royal blue (480-440 nm say) then just stick a bunch of actinic bulbs or blue leds over your tank... but wait no one does that - which should tell you something. In fact, if you tried it your coral wouldn't grow well at all. True actinics actually provide very little PAR or PUR or whatever you want to call it (420nm more so). Also, there are many other active pigments than chlorophyll a that are important. The blue is needed for the fluorescence and high color temp, but is not the only thing you need.

It would be a smidge of amber and red to balance out, and some 420nm actinic if we could. However, unless you missed grade-school coloring session, the last color you'd want to add is green.

That's just silly. In grade school you learn to mix blue yellow and red as primary colors. When dealing with light sources its blue red and green, to match the cones in your eyes... go look really closely at your TV or computer screen and you'll see how it can make every color your eye can see. If you want a really interesting experiment you can bounce "white light" from you screen off of a pure yellow or cyan pigment (not some mixed color or several pigments) and it will all look very dark because those the wavelengths are not there to be reflected... hence bad color rendition. There are other things to see in an average tank besides coral fluorescence.

Lastly, a true 14-20k blackbody source (like a star) still puts out tons of green, yellow, and red. In fact, it puts out more, all else being equal. It's just that blue violet are boosted relatively more. If you use only blue and a bit of red as you suggest, you will have a purple tank. Perhaps you like how that looks but not everyone does. In order to have a white light you must have green (yellow is just another way of activating your green receptors). All that water does is selectively reduce the longer wavelengths (effectively increasing color temp), not eliminate them entirely.

 

[jtma508]

Partially true. Most general purpose PAR meters read spectrum intensity the inverse as lux meters which are calibrated for the human eye and peak in green/cyan. PAR meters typically have a peak at 455nm, then at about 630nm. Looks like a valley on a spectral graph with green at the bottom, cuz green has little photosynthetic value. Basically, a PAR meter is a 'purple meter' and they aren't as sophisticated as people think. The problem with PAR meters, as I've stated in many other threads, is most read red light just as much as blue because they are designed for terrestrial plant growing. In reefing, 455nm blue is far more important than 670nm red, and I've seen very few PAR meters with a true reef mode, although they do exist. I just use my dSLR and only read the intensity of the blue channel, which is about 455nm, and use that as reference.

I'm sorry blasterman but you are mistaken. Quantum meters do, in fact, measure linearly across the spectrum of 400 to 700nm. They measure each photon equally. To be sure I contacted Apogee and they confirmed that fact. The curve you describe is the photosensitivity of terrestrial plants (as depicted in the curve on this page: http://www.specmeters.com/pdf/Light_Brochure.pdf) but that is NOT how the quantum meters measure.