Full Spectrum DIY guide

[BeanAnimal]

I think it is prudent to explain something that many of you may be missing, or not understanding....

A metal halide lamp (or the sun for that matter) outputs "full spectrum" light. That is, the photon energy is present in ALL of the visible spectra. Depending on the source, there may be peaks or dips at certain wavelengths, but it is "all" there.

When we talk about LEDs, we are talking about devices that emit light in a VERY NARROW spectral band. That is the have a single peak to which there is no other output. When we mix a few LEDs toghether, we are NOT creating NEW spectral output outside of the narrow bands (peaks if you will) that the LED emits. We are simply tricking the cones in our eyes into see the "average" (for lack of an easier term) between the peaks.

In other words, when you look at the light mixed together from RED, GREEN and BLUE led, you are NOT seeing "real" white lite. Your brain is being tricked (by the cones in your eyes) into seeing white. When you start mixing a whole bunch of single peaks, the output is simply not going to be as predictable as you think. More importantly, your corals will SEE the peaks and NOT the spectral illusion you are seeing.

Hope tha thelps.

 

[Nanofreak79]

I think it is prudent to explain something that many of you may be missing, or not understanding....

A metal halide lamp (or the sun for that matter) outputs "full spectrum" light. That is, the photon energy is present in ALL of the visible spectra. Depending on the source, there may be peaks or dips at certain wavelengths, but it is "all" there.

When we talk about LEDs, we are talking about devices that emit light in a VERY NARROW spectral band. That is the have a single peak to which there is no other output. When we mix a few LEDs toghether, we are NOT creating NEW spectral output outside of the narrow bands (peaks if you will) that the LED emits. We are simply tricking the cones in our eyes into see the "average" (for lack of an easier term) between the peaks.

In other words, when you look at the light mixed together from RED, GREEN and BLUE led, you are NOT seeing "real" white lite. Your brain is being tricked (by the cones in your eyes) into seeing white. When you start mixing a whole bunch of single peaks, the output is simply not going to be as predictable as you think. More importantly, your corals will SEE the peaks and NOT the spectral illusion you are seeing.

Hope tha thelps.

I'm not sure if it was here, or another thread where I explained that quickly I quess about mixing R,C and CB gives the impression of white , but deffinetly brings out colors in the corals. Thanks for posting though very useful info. The whole point of this thread is to say LEDs are not predictable, and let everyone join in and give there experience. Even if there doing something different or out of the box, unusal type set ups/DIY's, using various colors and blends etc. Thank you for your input, it should help explain why people might run into trouble!

 

[wackeeash]

Not specifically for DIY, but an informative article that includes some understanding concerning lights and corals.

 

[dsdaley77]

I think it is prudent to explain something that many of you may be missing, or not understanding....

A metal halide lamp (or the sun for that matter) outputs "full spectrum" light. That is, the photon energy is present in ALL of the visible spectra. Depending on the source, there may be peaks or dips at certain wavelengths, but it is "all" there.

When we talk about LEDs, we are talking about devices that emit light in a VERY NARROW spectral band. That is the have a single peak to which there is no other output. When we mix a few LEDs toghether, we are NOT creating NEW spectral output outside of the narrow bands (peaks if you will) that the LED emits. We are simply tricking the cones in our eyes into see the "average" (for lack of an easier term) between the peaks.

In other words, when you look at the light mixed together from RED, GREEN and BLUE led, you are NOT seeing "real" white lite. Your brain is being tricked (by the cones in your eyes) into seeing white. When you start mixing a whole bunch of single peaks, the output is simply not going to be as predictable as you think. More importantly, your corals will SEE the peaks and NOT the spectral illusion you are seeing.

Hope tha thelps.

Not to offend but where does your info come from? I just posted this response from an led company and it directly contradicts what you just posted, kinda confused :-/

 

[dsdaley77]

"In an email I recieved from an American LED company they state, " While true UV/Pure Blue LEDs are measured at 420nm you must also account for any Full Spectrum White LEDs you plan to install. The three spectrums of white you have selected will add somewhat to the overall total spectrum as being measured in individual nm's. In your example, the 10-12000k you are looking at are indeed a full spectrum white but with more focus on the Blue and green spectrums with some red range mixed in. The ammouts of visible light in most of these spectrums is very low but in similar fashion to a T5 420nm bulb, just because you can't see it doesn't mean it's not beneficial. While all white LEDs have total spectrum, you see differences in the visible light based on what end of the nm ladder the lights were designed for. Traditionally the 10-12000k has been so popular with aquarists such as yourself because the animals you keep, like and are easier sustained in home aquaria with lights that contain additional blue or Actinic.""

 

[Nanofreak79]

Not to offend but where does your info come from? I just posted this response from an led company and it directly contradicts what you just posted, kinda confused :-/

I guess I read it as we are searching for a color that we are never going to see? Maybe I'm confused as well? In reading again, I think he saying we're going to expose the corals to too many funky spectrums? And they won't like it but our eyes will? Am I wrong.......bean do you have any DIY experience to share? Pix, layouts or colors you suggest?:worried:

 

[BeanAnimal]

Not to offend but where does your info come from? I just posted this response from an led company and it directly contradicts what you just posted, kinda confused :-/

It is basic physics...

If you want more information you need to read about "additive" color mixing and the RGB color model. We are not "creating" light, we are tricking the brain into seeing a color of light. Take note that the R, G and B are not important, they are simply chosen because they offer the widest "gamut" (range) of percieved color when mixed. In fact, no set of RGB phosphors is the same, and that is why each and every RGB device renders colors differently or is more or less capable of differnet extremes.

See if this helps:
You have a SINGLE wavelength of RED and you put a filter over it that only allows a single wavelength of YELLOW through. What do you see through the filter? Nothing

Do the same with a GREEN wavelength and a YELLOW filter. You see nothing!

Now mix the two colors RED and GREEN and you see YELLOW. Now place the YELLOW filter in front of the mixed source. What do you see? Nothing, as there was no yellow wavelength to begin with. The "yellow" you saw was the brain being tricked to see additive light.

To have yellow light pass through the filter, the photons being emitted have to be in the yellow wavelength


I did not read the response from the LED manufacturer, but will either say that A) they are 100% full of crap or B) their context is "visible" light and color rendering, not true spectral output. That is, their business and context are "what will the light look like to the eye" and they are unaware of the photosynthetic context.

 

[bhazard451]

It makes sense. When you use a colored led like red, it focuses specifically on that narrow band of spectrum. Together mixed with blue and green/cyan, it may look like a white light, but it is not the same as a white led which has spectrum peaks within it, like a warm white with a red peak.

I usually try to mix different color temp whites to achieve this, as I've found "Cool White" leds to be lacking. Other than 420 being beneficial, I use colored leds strictly for visual appeal.

 

[BeanAnimal]

The goal is twofold. You need to get the spectral peaks correct for coral growth. Once you have that established, then anything you add is for visual appeal, which is subjective.

We know we need the blue and violet LEDs in a few peaks to grow coral. Some of the white peaks appear to help a bit. The hard part for many of us is getting the final color to not look overly blue or purple.

Sorry, I don't have any specific ratios that I know for sure work. To that end (and as mentioned above) each brand of emitter is going to use differnet combinations of phosphors to achieve the final "white light" you see. That means that (2) 20K chips that LOOK the same side by side may in fact be deriving the additive color using very different ratios of phosphors. That means mixing them with say a 420nm chip will result in a differnet look between the two chips when mixed with the 420nm chip. Hope that makes sense...

 

[bhazard451]

The goal is twofold. You need to get the spectral peaks correct for coral growth. Once you have that established, then anything you add is for visual appeal, which is subjective.

We know we need the blue and violet LEDs in a few peaks to grow coral. Some of the white peaks appear to help a bit. The hard part for many of us is getting the final color to not look overly blue or purple.

Sorry, I don't have any specific ratios that I know for sure work. To that end (and as mentioned above) each brand of emitter is going to use differnet combinations of phosphors to achieve the final "white light" you see. That means that (2) 20K chips that LOOK the same side by side may in fact be deriving the additive color using very different ratios of phosphors. That means mixing them with say a 420nm chip will result in a differnet look between the two chips when mixed with the 420nm chip. Hope that makes sense...

Yep, and its the exact reason why everything is so experimental right now. You need to know all the way down to what BIN your led is, or else even the same "Cree Royal Blue" led will have a different output than the same led that came from a different bin. Someone using the same ratio and leds, yet from a different company, can have different results.

You can let the manufacturers sort this out for you, but a lot of fixtures are lacking based on the price they are being sold for.

 

[dsdaley77]

I'm still confused. So are you saying that white is made by white? I understand that blues, reds, yellows, oranges etc emit a single spectrum based on the nm level of the light, but I don't think white falls into these same catagories. And it doesn't really help all of us to comment on something posted on the same page and to not have read it. If he's wrong I'd like to know why, not just a blanket " this guys full of crap". Our goal here is to learn not just blindly follow someone's words as end all. Just saying

 

[dsdaley77]

I guess from manufacturer's words and other experience not related to reefing I have developed an understanding that white light needs multiple spectrums to become visible as white. For instance, I "have a friend" who grows medical marijuana here in AZ and uses only warm and cool white LEDs to grow his crops. His medicine has been tested by labs and is perfectly comparable with growers medicines that were grown using MH and HPS bulbs. Now I learned in school long ago that they (plants) don't need every spectrum of light to flourish but do need multiple wavelengths to remain healthy and strong. So how can one achieve a healthy, strong, fully mature, viable plant from just white LEDs if they only contain one wave length? It would lead me to believe that white LEDs are indeed full spectrum.

 

[advancebc29]

I guess from manufacturer's words and other experience not related to reefing I have developed an understanding that white light needs multiple spectrums to become visible as white. For instance, I "have a friend" who grows medical marijuana here in AZ and uses only warm and cool white LEDs to grow his crops. His medicine has been tested by labs and is perfectly comparable with growers medicines that were grown using MH and HPS bulbs. Now I learned in school long ago that they (plants) don't need every spectrum of light to flourish but do need multiple wavelengths to remain healthy and strong. So how can one achieve a healthy, strong, fully mature, viable plant from just white LEDs if they only contain one wave length? It would lead me to believe that white LEDs are indeed full spectrum.

Look at any spectrum output of any white LED. Just about all stop around 450ish and also tend to have very little higher then 660ish so this is not true full spectrum. If we had single full spectrum LEDs that would solve a lot of issues. The LEDs your friend is using can grow plants very well as they tend to have a lot of orangish and redish light. They can also grow coral but are just not ideal.
Halide lights and other forms of lighting like T5s for coral have developed specific spectrum peaks that help corals both grow and color up nicely. For some reason all this was thrown out the window when LEDs came into the mix. Just within the last year or 2 have the DIY crowd started to figure this out and now the store bought fixtures will start to play catch up.
You are forgetting one thing when comparing plants to corals and that is water. Water absorbs the higher spectrum of light very quickly so corals tend to use more of the lower spectrum that can penetrate the water. Plants like the higher spectrum and my guess is can probably use both as they are readily available.

 

[bhazard451]

Think more simply. What is white light? Think of the "Dark Side of the Moon" prism album cover. The white light is refracted in the prism and reveals the color components akin to a rainbow.

 

[dsdaley77]

Exactly my point (I think), thank you for clearing that up. Would you then agree that a 6500k t5 puts out the same wavelengths as a 6500k led and 10-12k to 10-12k etc?

 

[bhazard451]

Exactly my point (I think), thank you for clearing that up. Would you then agree that a 6500k t5 puts out the same wavelengths as a 6500k led and 10-12k to 10-12k etc?

Similar, maybe. The same, no. The actual spectral plot could look very different between both to achieve what the manufacturer labels as 6500k.

 

[dsdaley77]

And based on the same picture that you posted that I had also posted it is clear to me that wavelengths outside of 417-655 really mean nothing to us aquarists in the big picture. So it really makes no difference to my corals if they can see the wavelength or not as long as its there. So what we are left with is this, if we provide all the different whites then the RGB etc become for our visual enjoyment and won't really provide much in usable light. It is believed that corals as a photosynthetic colony can only use so much before it is just there in excess.

 

[dsdaley77]

Bhazard, most of these are somewhat directed at bean and adv as I'm sure you can tell, I really do appreciate all input though we will get to the bottom of this

 

[BeanAnimal]

I'm still confused. So are you saying that white is made by white? I understand that blues, reds, yellows, oranges etc emit a single spectrum based on the nm level of the light, but I don't think white falls into these same catagories. And it doesn't really help all of us to comment on something posted on the same page and to not have read it. If he's wrong I'd like to know why, not just a blanket " this guys full of crap". Our goal here is to learn not just blindly follow someone's words as end all. Just saying

WHITE "LEDs" are "BLUE" Leds painted with a layer of phosphors. When a photon hits a phosphor, the energy is absorbed and then spat out as a photon of a different energy (wavelength). Different combinations of phosphors create different output wavelengths. You still have a very narrow peak (or peaks) that you are using to mix with the other peaks to trick the eyes into seeing a spectral range.

Also I did not say the guy was full of crap. I said, that I did not read his response, but if he was saying that you could create spectra that was not being emitted by the LEDs, then he was full of crap. You can't defy physics I did in fact give him the benefit of the doubt and said that he was likely speaking in terms of the perceived spectral output (what your brains sees as reported by the cones of your eyes).

 

[BeanAnimal]

Exactly my point (I think), thank you for clearing that up. Would you then agree that a 6500k t5 puts out the same wavelengths as a 6500k led and 10-12k to 10-12k etc?

No... I already explained this.

Your eyes can be tricked into seeing 6500K by any number of combinations of spectal peaks. 6500K is not a "color" that is defined by a single fixed ratio of spectral peaks.