Open letter to the LED industry

[Michigan Mike]

Thank you for making so much sense.

I see you didn't comment on light diffusion and filter color tints which is fair if it's part of your current research. However, a scientific question you probably could answer, as I'm not even a novice at best when it comes to quantum mechanics and radiation.

How does light, particularly 400nm-510nm + 560nm-660nm behave on surfaces? Reason I'm asking is simply because in LED tanks you will see amazing colors (those tanks that run little white intensity) on top where they are hit with direct light, but so strong contrast to where there is no light hitting directly. LED, as the sun, is a hard light source. MH is also, and T5 is soft.

Again, I'm suspecting that different energy waves will reflect differently on surfaces. Is this correct?

Interesting question, all I know is my halides reflect water lines everywhere on my ceiling. My t5 does not.

 

[Vannpytt]

From an article on advanced aquarist, found here:
http://www.advancedaquarist.com/2007/10/aafeature2
"As we have seen, most chromoproteins found within corals absorb light energy at ~570-580 nanometers (since they do not preferentially absorb blue and red wavelengths, they reflect it), often giving corals an apparent blue-purple coloration."

Please, can anyone elaborate on the quantity of this the coral needs? I find it strange to say something like "Your reef needs calcium". Sure it does, but how much? Too much is really bad, too little is also really bad. That argument holds little value if not quantified and put in context.

 

[Reef_Breeders]

Please, can anyone elaborate on the quantity of this the coral needs? I find it strange to say something like "Your reef needs calcium". Sure it does, but how much? Too much is really bad, too little is also really bad. That argument holds little value if not quantified and put in context.

Not really, it holds plenty of value, as your are saying that it is not needed at all. You don't need a ton of deep red or yellow, just a small amount. If you need me to be more specific, I will break out the spectroradiometer and measure a light for you. Look at any graph that shows the absorption spectra of Chlorophyll a, and the peak at 660nm is almost as big as the one at 430nm. So chances are, too much would be so red you wouldn't want to put the light over your tank. And, the size of the peak also shows that it should not be ignored- it will benefit your corals growth.

 

[Kattes]

Wonderful thread by the way. If this sort of thing went on here regularly, I'd definately spend more time over here.

A big thank you to everyone contributing to this discussion.

 

[penfold2]

Please, can anyone elaborate on the quantity of this the coral needs? I find it strange to say something like "Your reef needs calcium". Sure it does, but how much? Too much is really bad, too little is also really bad. That argument holds little value if not quantified and put in context.

If you look at the spectral output of most fixtures using white leds in a typical manner, you'll see that the 550-600nm peak that you're worried about is actually one of the weakest parts of the spectrum when combined with other colors. You need to look at the total output of the diode combination rather than just the white diode output. White leds simply prevent the curve from falling around the yellow portion of the spectrum. There is no spike there when used in typical ratios.

 

[Vannpytt]

Not really, it holds plenty of value, as your are saying that it is not needed at all. You don't need a ton of deep red or yellow, just a small amount. If you need me to be more specific, I will break out the spectroradiometer and measure a light for you. Look at any graph that shows the absorption spectra of Chlorophyll a, and the peak at 660nm is almost as big as the one at 430nm. So chances are, too much would be so red you wouldn't want to put the light over your tank. And, the size of the peak also shows that it should not be ignored- it will benefit your corals growth.

Fair enough. I will stand corrected again, and gladly admit that there is a need for red/yellow. While we are not dependent on "White" LEDs to supply this, it might be easier. That said, I believe the ratio is totally off. If you don't mind running those tests, that would be really awesome. (Not sarcasm, I really would love if you could as I don't have access to that kind of equipment myself.)

If you look at the spectral output of most fixtures using white leds in a typical manner, you'll see that the 550-600nm peak that you're worried about is actually one of the weakest parts of the spectrum when combined with other colors. You need to look at the total output of the diode combination rather than just the white diode output. White leds simply prevent the curve from falling around the yellow portion of the spectrum. There is no spike there when used in typical ratios.

No. When run at correct ratio, it's not a problem. When run @ 100% intensity, in most common fixtures, it is a problem, problem being that there is too much yellow-red. Not individually, but as a complete light source, there is to much.

I'm going to make another prediction: When/If Logan can make some tests I think we will see that a 1:1 Blue or Royal Blue to White will give a very bad spectrum, and by bad, I mean far from proven T5 tubes spectrum, like previously shown by Pacific Sun RD. Even if you add UV will the spectrum show wrong intensity when run as it was made. An easy solution for this would be to manufacture LED fixtures with a toned down "white" channel, this way people won't have a lot of the problems we see with LED reefs today. The LED industry would also benefit by creating a product that actually has ease of use and good results. At the moment, T5 is not a superior source, it's just fool proof to run with good results. At least I have not heard of anyone frying their coral under it.

 

[penfold2]

No. When run at correct ratio, it's not a problem. When run @ 100% intensity, in most common fixtures, it is a problem, problem being that there is too much yellow-red. Not individually, but as a complete light source, there is to much.

Just go to the website of any popular fixture and look at the spectral output. This will be the output with all channels at 100% (including white). There is no spike in the yellow area. The only way you could get this is by turning the whites way up, and everything else down. The whites are very well balanced by the other colors in the ratios that manufacturers are using. There is a spike in red, but this is due to red diodes purposely added in order to hit the large 660nm peak of chlorophyll a.

 

[Vannpytt]

Just go to the website of any popular fixture and look at the spectral output. This will be the output with all channels at 100% (including white). There is no spike in the yellow area. The only way you could get this is by turning the whites way up, and everything else down. The whites are very well balanced by the other colors in the ratios that manufacturers are using. There is a spike in red, but this is due to red diodes purposely added in order to hit the large 660nm peak of chlorophyll a.

No, if this was true, you would see everyone and their dog running LED fixtures over their tank, from cheap china products or whatever else with a 1:1 ratio run @equal A (since everyone is using Cree it seems).

A typical and fairly good quality LED fixture (Maxspect 16000 Kelvin):

A high end LED fixture (Vega):

Another high end LED fixture (Hyperion):

A typical and fairly good quality tube:

A typical and fairly good quality MH:

 

[Vannpytt]

Now, further, if you remove the white channel from the AI Vega spectrum, you would be left with something quite useful. (Or leave it on 10-20% white). This is true for the Maxspect also, but not for the Hyperion, who I recon, would preform better than both of these fixtures. At this point we are back to what I stated previously. Stop wasting space on useless chips.

I just want to make sure this doesn't get up. I do not work for, nor am I in any way affiliated with anyone in the industry at any level. I'm a simple hobbyist who want to succeed with LED lighting over my tank, and feel I can accomplish this with proper information.

 

[jda]

What do you suppose the output of these lights would be with no white, or 10-20% white? I know that PAR is not everything, but any guess if the output would be good or substandard?

 

[Vannpytt]

That's a twofold answer ime. One answer would "be more than enough". From what I gather, "White" LEDs are more counterproductive run at their maximum intensity. The other answer would be, "Why the heck don't they add more LEDs in the blue spectrum instead of wasting space on the white ones, which we actually don't want many of and need to dim down anyhow?

I figure, since there are tanks with good/decent success with LED, myself included, they run their white LEDs around 50-70%, depending on fixture and type (cluster/spread), then the blue that are there is doing most of the job them self, and don't fall in the trap of forgetting that blue light is far more powerful in salt water than white is. Blue light penetrates water better, violet even better. It does not seem visually bright to the human eye, simply because we don't see blue light very well, but it is, in PAR/power just as intense for all coral purposes, even more so. This is not scientific at all, but I would suspect a 3:1-5:1 range for everything:"white" is plenty. Point is, we get MORE PUR/PAR WITH LESS WHITE AND MORE BLUE

Corals can handle 100% of Blue intensity, and even more for better spread. I have actually never seen anyone prescribe "less blue" for a coral tank with bleached colors.

 

[jda]

I get that, but I am wondering, what these specific fixtures that people already purchased can be expected to achieve if they turn their whites down to 10%, or off. How then, do these fixtures compare to T5, VHO or MH? I guess that we won't know for sure unless somebody pops the meter out, but if they drop the PAR in half with half of the bulbs not running, or barely on?

The whole point of my question is: when do these become "done" and not changing all of the time. I totally get that somebody could make a 5:1 ratio new fixture and things will probably be better, but what happened to the many-thousands-of-dollars units that are on tanks now that are not even a year old? Suppose that all of this analysis is 100% right - is a manufacturer's promise of the holy grail one generation away from a garbage spectrum light worth much?

 

[penfold2]

The Maxspect is a rather old design with few colors. It really doesn't compare to current "full spectrum" designs. The Vega image didn't show up. And the Hyperion S doesn't even use white diodes.

Here are a couple "full spectrum" fixtures.

Ecotech Radion XR30w Pro

AI Hydra

There are no spikes in the yellow portion of the spectrum. Just a level output that is relatively low compared to the massive blue output and the moderate red output. If you still think this is too much high end output, you could reduce the number or output of the respective diodes, but people are deliberately aiming for these ratios. They're happy with the appearance and results. I've never heard anyone else claim that white leds in particular are burning their corals, and it doesn't seem to make sense, as natural sunlight has far more high end output than any of the spectra we're talking about. Iwasaki 6500K MH's grow coral like crazy, and they also have far more high end output.

 

[Vannpytt]

I get that, but I am wondering, what these specific fixtures that people already purchased can be expected to achieve if they turn their whites down to 10%, or off. How then, do these fixtures compare to T5, VHO or MH? I guess that we won't know for sure unless somebody pops the meter out, but if they drop the PAR in half with half of the bulbs not running, or barely on?

The whole point of my question is: when do these become "done" and not changing all of the time. I totally get that somebody could make a 5:1 ratio new fixture and things will probably be better, but what happened to the many-thousands-of-dollars units that are on tanks now that are not even a year old? Suppose that all of this analysis is 100% right - is a manufacturer's promise of the holy grail one generation away from a garbage spectrum light worth much?

First, the many thousands-of-dollars units already sold are profits already; -the whole reason why there is an industry in the first place (which is good).

Second, if you turn down your white LEDs to 20%, without having any Yellow/Red/Green chips your tank would look very blue, but your corals would enjoy it. Problem is, you probably would not.

I'm fairly sure I can create a fixture, that would end all problems related to LED and corals (acropora/montipora). Bold statement for sure, but so could the LED industry. I'm doing some testing of my own, but I do not have access to fancy high tech spectrograph (or whatever fancy shizzle they use) and PAR meters and all the marine biology science to understand the requirements of corals, although most have been mapped already.

Would you have to buy another fixture if they are done and works?

I think I know what is going to be the next generation and the one after already. Don't worry, there will be a lot more coming to replace whatever fixture you buy. There is a reason why the industry, with exception of Pacific Sun who actually did their homework, does not participate in this thread. Either they have a) a marketing strategy in place for selling fixtures a long time into the future, b) are wrong, or they c) don't know. I would assume option a. Again, a bit over the edge. I don't want to sound like an absolutist, but unless they want to communicate with the end user something else than marketing and commercial messages, then I honestly would not trust them at all.

1-2 will be setting the tech standard, the rest will follow suit with different marketing crews.

 

[nickdo]

White leds can be used as a more continuous wavelength source compare to other leds. In a small led cluster(like vertex/PS did), all leds within a cluster are regarded as a point light source. In my opinion, what you care about is a the final spectrum. ex, 5"uv", 5 rb, 4b, 1white, 1green, 1red, sounds good, right? Maybe you want to 100% driving the light, but some people want a more natural look. Fixture manufacturers must give an adjustable ability on it, just a few leds and that won't increase the total cost much.

no optic, all leds build in a very small cluster. I've this same idea last year in my DIY led, and this time PS give me a professional prove. PS RD did a very good job.

but I have a question for PS. Leds build as closed as possible is better for color mixing and Par, why dont u just stick all leds together and use under layer PCB for circuit?

 

[Vannpytt]

The Maxspect is a rather old design with few colors. It really doesn't compare to current "full spectrum" designs. The Vega image didn't show up. And the Hyperion S doesn't even use white diodes.

Here are a couple "full spectrum" fixtures.

Ecotech Radion XR30w Pro

AI Hydra

There are no spikes in the yellow portion of the spectrum. Just a level output that is relatively low compared to the massive blue output and the moderate red output. If you still think this is too much high end output, you could reduce the number or output of the respective diodes, but people are deliberately aiming for these ratios. They're happy with the appearance and results. I've never heard anyone else claim that white leds in particular are burning their corals, and it doesn't seem to make sense, as natural sunlight has far more high end output than any of the spectra we're talking about. Iwasaki 6500K MH's grow coral like crazy, and they also have far more high end output.

I'm kinda done :deadhorse1: with you. It seem you don't want to understand what "full spectrum" is. Read this Cree XT-E Datasheet and let's continue when you understand that when you combine that LED in the fixture, you get a lot of unwanted spectrum. Compare it to the T5 or the MH I posted and you can easily see the difference. Yes, 6500 Kelvin Iwasaki will grow coral "like crazy", but they won't color up. There are other particulars I won't go into discussion with you on here, but just because you call it 6500 Kelvin, does not make it yellow like a "white" LED.

That said, I see AI has found out that having a lot of White LEDs are a waste after all. They are currently running 1/5th of their LEDs as white.

WONDER WHY

 

[penfold2]

You seem to think that anything more than trace amounts of yellow light is undesirable and even harmful to corals. I'm just pointing out the fact that corals grow quite well under high levels of yellow light, and that the relatively modest amount of yellow light in most current led fixtures is there by design and considered desirable by many. If you think you know better than virtually all the led builders out there, then go ahead and experiment. But there are people in this thread who are now wondering if their current fixtures are worthless, and if their white leds are going to burn up their coral. Current led fixtures have proven themselves able to grow healthy, colorful coral. I'm sure there are further improvements to be made, but the current fixtures already do an excellent job.

 

[Vannpytt]

Industry comes up with a new "these next generation fixtures are now the new standard in LED light" nonsense. You should realize, just because they vewrite so, does not make it true, and they had said the same thing for a long time. When are they going to stop wasting consumer resource and corals on experimenting and testing. They should darn well do their own R&D and deliver a product that works. I won't go into how many people have failed with LED or how many corals have died due to wrong advice and product design in LED lighting, but the numbers are substantial, even AI Vega and Vertex users are "complementing" their fixtures with T5, and they should. For the value they put in their fixtures, they sorely want to make it work.

I'm not sure if people would pay premium for a fixture that delivers in terms of acropora growth and coloration, seeing so many buying unbranded eBay products to save a few bucks, but I really do wonder how many corals they need to kill before that cost is more than a premium fixture would cost them. (No, AI, Vertex and Ecotec, while marketed as a premium brand, are nothing special, but man!! that's some good marketing)

Premium LED fixtures are still 2-3 generations away, although the Hyperion S and some DIY fixtures is future generation fixtures that delivers just that. I'd like to see long term use of these (6 months minimum).

All the current fixtures can work and produce good results, when people learn to use them. Main points being these:

• When corals brown out under LED, it's not because of "too little light", it's because of too much yellow.
• When you see hard contrasts to the color from top/side/below, you don't need more light, you need diffused (Not blended) light. Remove optics
• Fixtures for tanks >70cm needs some optics or a higher energy LED, and a lot more to give proper penetration and spread.
• Corals (Lets narrow it down to acropora/montipora) don't grow well under high levels of yellow&red light. (In a tank less than 5-10 meters deep)
• Users who want to sell their fixtures and go back to MH/T5 are doing so for a reason, they feel they did not get what they wanted.
• There are no long term successes with the new fixtures yet, they are barely 1-2 months old, and from what I've read, the users don't run white at 100%
• The "white" LEDs are reduced in the current fixtures for a reason, they are not needed; -they can be reduced even further, and indeed people are doing so by dimming them down, reduce their current by default.
• 1-2 Yellow for each standard fixture is enough; -why put more there when it's not needed?
• No, there is not any evidence that acropora thrives in terms of coloration and growth under the current "white" LED spectral bump, simply because they don't.

I would love to see any of them coming to this, or any other thread and discussing these concerns with the current low CRI LED chips at 2500-8200 Kelvin (That's a pretty darn wide range with the CRI probably limiting that specter to the lower end). I'm not going to tell you why they are building these yellow light fixtures, but I'm quite convinced, it's to leave room for upgrades like we have seen, trending towards what I'm advocating: Less yellow. They should also start to think about less optics. Frosted acrylic lenses don't offer diffusion, but blending. It seems you believe that by putting the Blue LED curve on top of a white LED curve do you remove the unwanted yellow, but that's not what happens at all. You would want to reduce the yellow to reduce the yellow, by running the white channel on less current (less intensity) and reducing the amounts of white needed. I think it's more like White 1:15-20 Blue in intensity needed for a coral in that spectrum, but I do not have access to test equipment, so I can only ask the industry.

 

[wayne in norway]

This seems to keep going in circles. Yellow light is the work of satan. Then there is a discussion and seeming agreement that it isn't, and that a small amount is ok, and then, we suddenly jump back to the start, the 'full spectrum' isn't full spectrum and yellow is a bad thing.

There is not much yellow in the vega spectrum. If you were to look at the spectral plots at 10m, 15m of seawater then they would contain just as much yellow. What is wrong with it?

Working on the assumption that some small amount of yellow is a good thing, but we're going to take the white chips out, how do you propose to put the yellow back in to achieve a natural spectra?

 

[Vannpytt]

To not repeat myself yet again, I think we have gone through your question already.