Open letter to the LED industry

Maybe I didn't explain my question well enough.

What I was wanting to know was how much of the color blend you get from different lights are created in our heads and how much is real. For example, if you take a blue lamp and a red lamp and project them onto the same area of a white surface you get purple light. Waves do affect each other but to what degree did the properties of the light actually change? Is it only purple to us or is it actually some new wave lengths of light in there and to what extent.

No new wavelength are generated by "blending" the light this way. In other words, lights don't blend like you described. They are not waves riding on the same medium.
 
Hello Ryan,
They are not smoothed.
This readings was taken on following config(on Ocean Optics spectrometer):
- integration time: 35ms
- scans to average: 2
- boxcar widh: 2
- electric dark correction: enabled
I think that our spectrum is most credible in compare to other manufacturers charts - we have measured some of competitors lamps and few of them have "little" other looking spectrum charts(as I remember I posted that kind of "sample" even in this thread).
We dont have to worry about our charts - we are taken full responsibility for them. I'm curious to know that all other manufacturers also..
Thanks for clarifying that. So LEDs with their broader peaks have the ability to provide the complete spectrum between 400 and 500 nm if you choose the right mix.
 
It helps to have a fancy spectrum analyzer doesn't it. :D

I had worked with one years ago when I worked in quality control for a lighting manufacturer. And again when I took an optics class at a university simply to have access to theirs. Yes it is nice to have one but it not a piece if equipment that is affordable to even some smaller companies. For a quality instrument that gives you good resolution your talking a considerable investment.

However there used to be some companies that rented these instruments out. I know when we used to send ours in for the annual calibration we used to rent one out. But I wonder how many places you can find that do rent these out today.

For the application that most people need one thought there is a lower cost optional approach. It is a matter of using a light box with a prism and photographing the rainbow created. Yes it is very crude and I doubt that it wll give you any accurate means of reading the wave lengths but it can give you an idea of what the spectrum from a single light source looks like. There are some people that actually think these can be accurate and have a web site dedicated to this method.
 
Last edited:
I've been following the various discussions about LEDs and have been interested about what "it" is that many of the LED fixtures are missing. In doing so, I started looking at European reef forums since there are some extremely successful reefs there that use LED. For the most part the common denominators both there as well as here seem to be the following:
1) More emitters to improve coverage, no lenses used
2) Total wattage is closer to traditional lighting
3) Use of full spectrum but not too much yellow or white.
4) Use of Violet LEDs(~420-430nm)

Then I found the best LED lit reef I've ever seen and a contender for top 10 in my book, the Triton Shop reef, using Lani LEDs made there. There is no secret to what they are using, but what they are doing as far as spectrum goes against almost everything Ive seen regarding LEDs:
1) Per 16"X16" square, the base model uses 32 Cree cool white and 32 Cree Blue(not royal blue)
2) They also use 16 true violet LEDs, exact spectrum not known, but I'd guess 420-430 since they refer to them as "purple"
3) They cover the tank almost completely with emitters, and run them at max 1watt each for efficiency. Heatsinks with no fans, no lenses used.

The Lani Pro adds 8 red, 8 deep red, and 8 royal blue to the above set of emitters for "color rendition" which makes the corals look even better from what I can gather but not much else.

In looking at a bunch of threads from there, people who see these tanks say that the corals look electric, and rave about the color(the pictures look good too). Corals grow like weeds. 1Meter deep tanks have no problems with penetration. No ramping of LED outputs are done for acclimation, corals don't seem to get burned like with some LED systems.

Have we been missing something here in the US when it comes to what emitters we choose? Most LED experts warn to not use Blue LEDs except as accents to the normal royal blue, and almost everyone seems to have moved away from cool white LEDS these days. Is having a massive peak of 420nm light enough to overcome the issues with these LEDs or is there some other synergy that happens?
 
I had worked with one years ago when I worked in quality control for a lighting manufacturer. And again when I took an optics class at a university simply to have access to theirs. Yes it is nice to have one but it not a piece if equipment that is affordable to even some smaller companies. For a quality instrument that gives you good resolution your talking a considerable investment.

However there used to be some companies that rented these instruments out. I know when we used to send ours in for the annual calibration we used to rent one out. But I wonder how many places you can find that do rent these out today.

For the application that most people need one thought there is a lower cost optional approach. It is a matter of using a light box with a prism and photographing the rainbow created. Yes it is very crude and I doubt that it wll give you any accurate means of reading the wave lengths but it can give you an idea of what the spectrum from a single light source looks like. There are some people that actually think these can be accurate and have a web site dedicated to this method.

there are many places even universities that will do testing on contract for companies at very reasonable prices, so it is not unatainable and really inexcusable for even a small company not to get spectrum analysis of their systems, even if they have to send off for third party testing (which personally I find preferable as the testing agency is then non biased) though you still have to then wonder what actually gets published.....
 
I've been following the various discussions about LEDs and have been interested about what "it" is that many of the LED fixtures are missing. In doing so, I started looking at European reef forums since there are some extremely successful reefs there that use LED. For the most part the common denominators both there as well as here seem to be the following:
1) More emitters to improve coverage, no lenses used
2) Total wattage is closer to traditional lighting
3) Use of full spectrum but not too much yellow or white.
4) Use of Violet LEDs(~420-430nm)

Then I found the best LED lit reef I've ever seen and a contender for top 10 in my book, the Triton Shop reef, using Lani LEDs made there. There is no secret to what they are using, but what they are doing as far as spectrum goes against almost everything Ive seen regarding LEDs:
1) Per 16"X16" square, the base model uses 32 Cree cool white and 32 Cree Blue(not royal blue)
2) They also use 16 true violet LEDs, exact spectrum not known, but I'd guess 420-430 since they refer to them as "purple"
3) They cover the tank almost completely with emitters, and run them at max 1watt each for efficiency. Heatsinks with no fans, no lenses used.

The Lani Pro adds 8 red, 8 deep red, and 8 royal blue to the above set of emitters for "color rendition" which makes the corals look even better from what I can gather but not much else.

In looking at a bunch of threads from there, people who see these tanks say that the corals look electric, and rave about the color(the pictures look good too). Corals grow like weeds. 1Meter deep tanks have no problems with penetration. No ramping of LED outputs are done for acclimation, corals don't seem to get burned like with some LED systems.

Have we been missing something here in the US when it comes to what emitters we choose? Most LED experts warn to not use Blue LEDs except as accents to the normal royal blue, and almost everyone seems to have moved away from cool white LEDS these days. Is having a massive peak of 420nm light enough to overcome the issues with these LEDs or is there some other synergy that happens?

Interesting. Do you have links to these tanks?

I have seen a nice SPS dominated tank in a local aquarium store which has been running for over 2 years under the LED fixtures only - I believe it was from Vertex - that did not have any lenses.
 
The triton shop tank:
picture.php

The triton reef tank, shop, and some stuff on the LEDs:
http://www.youtube.com/watch?v=aY-qhFgNVA8
A Lani Reef from Germany:
20131116_113137_2.jpg

Vinces' reef in UK:
20131116_142553.jpg

20131116_142215.jpg

and one more:
IMG_0709_zpsa7b339a7.jpg


There are more, but these are what I was able to find first thing this morning. I think the results are hard to dispute. The Lanis are Freakishly expensive but there isn't anything exotic about the diodes they use.
 
The Lani Pro seems like an interesting concept

16 Watts Cool White
16 Watts Warm White
16 Watts (Violet?)
8 Watts Red (620nm?)
8 Watts Deep RED (680nm?)
8 Watts Royal Blue (455nm)
72 Watts total.

Assuming that a 120 gallon tank could run on three of these you have 216 Watts of LED's which is less than many people here use.

My systems consist of mainly Royal Blues with Neutral Whites which is a much bluer combination that they are running. Sanjay is still a heavey user of 6,500K metal hides and his systems are known for there fantastic SPS coral growth.

Makes me wonder if we are pushing the blue part of the spectrum to the point where we are getting a negative return on it? Corals can only absorb so much light at any given frequency and providing excess can not only be a waste but also a detriment to the corals.

As far as the violets it would be a wid guess what they are using if they do not state what they are using. Technically UV is anything with a wave length shorter than 400nm it is only an "American Aquarium" reference that we refer to 410 nm and even 430 nm LED's as either UV's or near UV, Technically Violets are anything between 400nm and 440 nm.

The other issue is they are running a 430nm LED's I would love to know there source. Every 430 nm I have tried were very poor quality and did not last long. However I was running them at 500ma or 700ma which is usualy at a higher wattage then you claim there are.

Thinking for a future experimental build
3 5 watt Cool Whites
3 5 watt Warm Whites
3 5 Watt Royal Blues
6 2 Watt Violets
4 2 watt Reds
4 2 Watt Deep Reds
73 Watts total for a 30 Gallon tank?
 
The Lani Pro seems like an interesting concept

16 Watts Cool White
16 Watts Warm White
16 Watts (Violet?)
8 Watts Red (620nm?)
8 Watts Deep RED (680nm?)
8 Watts Royal Blue (455nm)
72 Watts total.

Assuming that a 120 gallon tank could run on three of these you have 216 Watts of LED's which is less than many people here use.

My systems consist of mainly Royal Blues with Neutral Whites which is a much bluer combination that they are running. Sanjay is still a heavey user of 6,500K metal hides and his systems are known for there fantastic SPS coral growth.

Makes me wonder if we are pushing the blue part of the spectrum to the point where we are getting a negative return on it? Corals can only absorb so much light at any given frequency and providing excess can not only be a waste but also a detriment to the corals.

As far as the violets it would be a wid guess what they are using if they do not state what they are using. Technically UV is anything with a wave length shorter than 400nm it is only an "American Aquarium" reference that we refer to 410 nm and even 430 nm LED's as either UV's or near UV, Technically Violets are anything between 400nm and 440 nm.

The other issue is they are running a 430nm LED's I would love to know there source. Every 430 nm I have tried were very poor quality and did not last long. However I was running them at 500ma or 700ma which is usualy at a higher wattage then you claim there are.

Thinking for a future experimental build
3 5 watt Cool Whites
3 5 watt Warm Whites
3 5 Watt Royal Blues
6 2 Watt Violets
4 2 watt Reds
4 2 Watt Deep Reds
73 Watts total for a 30 Gallon tank?

They use more of the whites and blues, so that each 16" square consumes 80 watts for the regular and 104 watts for the PRO.
The "Lani One":
32 Cree Blue
32 Cree cool White XT-E High bin(139lumen/watt)
16 Violet

The Lani Pro:
32 Cree Blue
32 Cree cool White XT-E High bin(139lumen/watt)
16 Violet
8 Red
8 Deep Red
8 Royal Blue'

I'm very intrigued to try building a small one for my naughty tank. The only thing holding me back is that pricing out the components comes to ~$450 for a 8"X16" DIY "pro" setup, the cost goes down a bit to ~410 bucks if building the basic model. If you have any suggestions to bring the price down lower, I'm all ears.
My tank's area is 36"X28" and right now its covered by a 316 watt Sunpower fixture. To cover the same area with one of these it looks like I could get away with 16"X32" of panels using either 160 or 208 watts max. To me that is a realistic wattage to properly replicate the current wattage of the T5s. Its not a huge power saving but there doesn't seem to be a performance compromise like a lot of other LED systems I've seen to it's piquing my interest...
 
Last edited:
As far as the violets it would be a wid guess what they are using if they do not state what they are using. Technically UV is anything with a wave length shorter than 400nm it is only an "American Aquarium" reference that we refer to 410 nm and even 430 nm LED's as either UV's or near UV, Technically Violets are anything between 400nm and 440 nm.

The other issue is they are running a 430nm LED's I would love to know there source. Every 430 nm I have tried were very poor quality and did not last long. However I was running them at 500ma or 700ma which is usualy at a higher wattage then you claim there are.
There was an article on another site that interviewed the builder and the article speculated that the UV LEDs were Edison, but after a lot of searching I couldn't find any Edison Diodes that seemed to match. The issue with frying the UV LEDs seems to be a big issue, but I did find some newer LEDs that run at higher current - stevesleds has a dual core UV led that runs at 1000mA for example. If using the 700mA UVs I'd just keep them on their own driver.

I dug for quite a while to figure out what "white" they were using and finally figured out it was a cool white(they gave the bin codes and lumen outputs). No such luck on the UV, however. I would expect that either 420 or 430 has to be used to hit the photosynthetic peak since they originally didn't use royal blue LEDs. Normally if you use that many UVs with royal blue it would look very purple, but I wonder if the combination of UV with regular blue gives a more balanced blue presentation, especially when the blue portion of the cool white is mixed in.
 
They use more of the whites and blues, so that each 16" square consumes 80 watts for the regular and 104 watts for the PRO.
The "Lani One":
32 Cree Blue
32 Cree cool White XT-E High bin(139lumen/watt)
16 Violet

The Lani Pro:
32 Cree Blue
32 Cree cool White XT-E High bin(139lumen/watt)
16 Violet
8 Red
8 Deep Red
8 Royal Blue'

I'm very intrigued to try building a small one for my naughty tank. The only thing holding me back is that pricing out the components comes to ~$450 for a 8"X16" DIY "pro" setup, the cost goes down a bit to ~410 bucks if building the basic model. If you have any suggestions to bring the price down lower, I'm all ears.
My tank's area is 36"X28" and right now its covered by a 316 watt Sunpower fixture. To cover the same area with one of these it looks like I could get away with 16"X32" of panels using either 160 or 208 watts max. To me that is a realistic wattage to properly replicate the current wattage of the T5s. Its not a huge power saving but there doesn't seem to be a performance compromise like a lot of other LED systems I've seen to it's piquing my interest...

They are running 80 to 96 LED's probably on a 350ma driver circuit. Yes at $4.00 per LED you have $320 to #84 in LEDs alone. But if you go with higher wattage leads for example simply running everything at 700ma you can decrease the number of LED's in almost half. Yes you have a drop in effeciency the higher the current your running but with a decrease number of LED's it is a big cost saver even if you have to add 20% more power to the fixture to make up the difference.

7 Cree Royal Blue
7 Cool White
Run at 1,500 ma would give the equivalent light. to theirs running a 1 watt.
10 Violets
5 deep red
5 True Blue
Running at 700ma would give you the equivalent light to theirs at 1 watt.

Now we dr4opped the chip count down to 34 and and the chips should be under $136 saving $184 on the build cost.

Interestingly the build you list here is not the same as the one earlier mentioned with warm whites , reds and deep reds.
 
My custom built build My led 48" bar fills in spectrum gaps well and is also a great evening light.

Lighting consist of 4 48" bars on a 120g long
1 x 65 watt build my led Custom 120 degree lense
24uv 18 deep blue 18 royal blue 8 warm white 4 660 red

1 x 10w 450nm
1 x 10w 12k white
1 x 50/50 1w lenses 60w

Build my led light only ... great evening light
10kt5%252Bled2.jpg
 

Attachments

  • color.jpg
    color.jpg
    45.3 KB · Views: 1
  • color2.jpg
    color2.jpg
    36.4 KB · Views: 0
  • reef.jpg
    reef.jpg
    98.3 KB · Views: 1
  • top1.jpg
    top1.jpg
    75.3 KB · Views: 2
  • top2.jpg
    top2.jpg
    78.6 KB · Views: 1
They are running 80 to 96 LED's probably on a 350ma driver circuit. Yes at $4.00 per LED you have $320 to #84 in LEDs alone. But if you go with higher wattage leads for example simply running everything at 700ma you can decrease the number of LED's in almost half. Yes you have a drop in effeciency the higher the current your running but with a decrease number of LED's it is a big cost saver even if you have to add 20% more power to the fixture to make up the difference.

7 Cree Royal Blue
7 Cool White
Run at 1,500 ma would give the equivalent light. to theirs running a 1 watt.
10 Violets
5 deep red
5 True Blue
Running at 700ma would give you the equivalent light to theirs at 1 watt.

Now we dr4opped the chip count down to 34 and and the chips should be under $136 saving $184 on the build cost.

Interestingly the build you list here is not the same as the one earlier mentioned with warm whites , reds and deep reds.
I think you might be mixing me up with someone else(i just checked my original post and didn't mention warm whites). The Lani runs 3 watt LEDs at 1 watt max for best efficiency and longevity, and one of their stated goals is to mimic the spread and saturation of T5, so lowering the emitter count changes the dynamic of the light. The folks that have these state there is very little shimmer, and very little shadowing. Also, these are not a name brand in the traditional sense, Lani LEDs are a hand built, custom deal made in the Triton shop. I don't think you can buy the real thing here in the US at all at this time.

I may try building the basic model and then add the reds and royal blues later if I like it. In bulk, maybe the emitters can be purchased a bit cheaper.
 
I think you might be mixing me up with someone else(i just checked my original post and didn't mention warm whites). The Lani runs 3 watt LEDs at 1 watt max for best efficiency and longevity, and one of their stated goals is to mimic the spread and saturation of T5, so lowering the emitter count changes the dynamic of the light. The folks that have these state there is very little shimmer, and very little shadowing. Also, these are not a name brand in the traditional sense, Lani LEDs are a hand built, custom deal made in the Triton shop. I don't think you can buy the real thing here in the US at all at this time.

I may try building the basic model and then add the reds and royal blues later if I like it. In bulk, maybe the emitters can be purchased a bit cheaper.

The eveness verses the sparkle is something that is debatable. People that loved the MH's and went to T--5's missed it others do not like it. I personalty like a little sparkle that you do not get with t-5's. So to me using more slightly higher powered LEDS is the key to get some sparkle.

As far longer LED life yes but I have run CREEE LEDs 15% over there maximum rating for 4 years now without any issues. However on several unknown brand Violet LED's even running them at there rated current produced issues in less than 6 months.

As far as the warm whites yes you did not mention them. However on another board they were mentioned on a Dutch LED system that used 16 Warm Whites, 16 Cool Whites and 16 Royal Blues. Sorry if I got the two confused. But they also offered a premium system that added 8 Violets, 8 True Blues, 4 reds, and 4 deep reds. The basic system ran the LED's at 350ma and the premium ran them all at 500 ma. They claimed the basic system was 50 watts and the advances was 120 Watts.
 
Build my led light only ... great evening light
10kt5%252Bled2.jpg

Thios is where personal color choice is an issue. I like most of the coral colors but noticed the red tint that even extended onto the substrate. But it could be that I'm used to looking at bluer tanks. Or it could be the way the camers interrupted the color data.
 
There are 4 deep red leds in the fixture ... they are only apparent when the build my led fixture is on by itself ... I may black out 2 or have them changed to green /cyan 500nm in the future ... but they don't seem to be adding algae and are not annoying or as visible in the pic and just add a purple tint to the tank ...
 
Back
Top