LED color temps and wavelengths

[TimmyD16]

I posted this in the Lighting forum but maybe it'll get more traction here...

So first some quick background... I want to get into my first reef tank, I have a hexagon tank lying around, but I was afraid of using it due to it's awkward shape and relatively tall stance. My number one concern was lighting, I want to do a mixed reef and I was very worried I couldn't fit the power under the small hood to light a reef, but then I started researching LED Panels. But in shopping for them I've run into a few questions...

1.) I ran across one, labeled as an atinic, it is a 9 LED panel consisting of (3) 10,000K LEDs and (6) 453nm LEDs, which from my understanding would look relative to a 20,000K fixture...? maybe a bit higher... But then should I just go with the same panel just with (9) 20,000K LEDs?

2.) Should I go with one large power panel, my tank is 22 gallons so I was figuring about 80W probably 20,000K.... or should I go with (4) different 20W panels, I'm thinking the later would give me more customization, maybe do (1) 10,000K (2) 20,000K and (1) Royal blue... Just to get more spectrum... Or even a 6kK a 10kK a 20kK and a RB..?

3.) How important would some green or red light be? I want the best lighting for my reef, but I would be highly doubtful that small wattages would reach any significant depth, unless I step up the wattage and then I feel like it would affect my viewing pleasure... any thoughts?

Thank you all for your time and info...

[uncleof6]

http://reefcentral.com/forums/showthread.php?t=2231073

[TimmyD16]

Thanks for that thread, but I have already read it, and it does answer the question about reds and greens (albeit in a very argumentative way), but it does't answer whether the light would reach the bottom with the minimal wattage behind it... Which is what I'm asking...

[theatrus]

Exact power output is hard to determine without knowing the LED die types and any optics involved.

[gbru316]

Originally Posted by theatrus

Exact power output is hard to determine without knowing the LED die types and any optics involved.

Power is merely a function of voltage and current, and is measured in watts. Did you mean PAR?

OP, keep in mind that LED's are directional. 3 watts directional is FAR brighter than 3 watts omnidirectional. That being said, it would help to have link, or at least some info about the array's you are looking at. Without further info, it's hard for us to offer any advice.

[uncleof6]

Originally Posted by gbru316

Power is merely a function of voltage and current, and is measured in watts. Did you mean PAR?

OP, keep in mind that LED's are directional. 3 watts directional is FAR brighter than 3 watts omnidirectional. That being said, it would help to have link, or at least some info about the array's you are looking at. Without further info, it's hard for us to offer any advice.

You started out strong, then went off track. Watts is power-- a function of voltage and current. What needs to be determined in "meeting needs" is the PAR of a particular emitter, (or more precisely the PAR at depth, a particular grouping of emitters produces) not the watts. You can have a 100watts of emitters, very bright, with 0 PAR--unlikely, as green/yellow will contribute--just not as much, but UV will contribute probably zero and be harmful.

The OP's question is ambiguous, as he is talking about correlated color temperatures, and watts, with only 1 reference to the radiation itself: 453nM. It is mostly a discussion of aesthetics rather than meeting needs. Even if the emitters came with a Photosynthetic Photon Flux Density/watt specification, (Lumens/watt tells us nothing other than how bright we would perceive it to be) it would still tell us little about what it would do in salt water, at depth. Color temps are not necessarily related to wavelengths. However, I do know that the OP really wants to know two things: What is going to be good for his critters, and what is going to look good. Sometimes there is a gap between the two, and for aesthetics we are most likely going to sacrifice some PAR.

In simple terms, it is the quality (spectral distribution) of light that meets needs, equating to PAR, and filling in the gaps for aesthetics (CCT etc.) is secondary. That is what I linked to the other thread for.

[TimmyD16]

I totally agree with the fact that looks don't always equal healthy animals, I think it's more of an issue of whether I would get more benefits out of a single say 100W actinic chip (toned down as to not fry my tank) with 90d optics..

Or from say 4 various 20W chips of varying spectrums, maybe 1 RB, 1 20000K, 1 10000K, and 1 6500K, all with 90d optics?

Not even from a "pleasant to look at" perspective, but more from a wider wavelength perspective. One of the most common arguments I hear against LEDs is that they are too narrow of a bandwidth, would something like that setup alleviate that, or should I just go with blue/20000K because thats what someone else did and his tank didn't crash?

[uncleof6]

Without a spectral distribution for the emitters, and a PAR reading for the fixture, that is an unknown. The 445nM RB, corresponds to the peak in the photosynthetic curve, however, what benefit the others would be, depends on the wavelengths used to create those particular correlated color temperatures, and what we see is not necessarily what is actually there.

[TimmyD16]

I mean wouldn't a 6500K LED inherently have more yellow/red than a 20000K?

[BeanAnimal]

It may have no "yellow" or "red" depending on how the color is derived.

The K rating of a "white light" really says nothing about the spectral components of the light. It simply indicates the relative look, under certain conditions, compared a specific black body radiator of said temperature. Confusing... yup

[TimmyD16]

yes, but how else would you get a lower temp white other than by adding/subtracting different colors, yellow for example... How would you go from the blueish tint of even a 10000K light source to a "warmer" looking 5000K other than having a bit more yellow/red involved in the overall spectral picture?

[BeanAnimal]

In simple terms, it could be warmed up by adding more red in context to the "fixed" amount of blue, or by decreasing the amount of blue in context to the fixed amount of "red". That is the point: without looking at the actual spectral plot it is hard to tell what is really going on.

For any given appearant color (as seen by the eyes) there are an infinite number of spectral combinations that can make that color. Look up metamarism

[TimmyD16]

Very true, but given that the 6500K has a maximum lumen output of 1400 and the 20000K a max of 1550, 150 difference or a mere 12 spherical candle powers... I have to assume that decreasing 150 lumens worth of blue light wouldn't change a 20000K to a 6500K, there has to be something more there yes? some added spectrum somewhere, be it red yellow green or orange... and with this added spectral quality somewhere wouldn't it lead to a more full spectrum LED array? Resulting in something that more resembles the sun?

[Neogenocide]

What are the dimensions of the tank?

[TimmyD16]

Hex tank, each side is 9", 24" tall...

[Neogenocide]

diameter? I don't feel like breaking out my geometry tonight.

[TimmyD16]

about 18" roughly...

[TimmyD16]

So taking out color from the equation... my biggest question is this... traditionally people use a 2:1 ratio of blue to white... So would it be better to use one powerful chip of atinic, which has a 3:2 ratio of blue to white, or do I use three or four lower power chips to get potentially more spread (?)... as long as the net results are equal in power would it not really matter? would it just come down to personal preference?

[Dave Thebrewguy]

A lot will depend on how you plan to aquascape the tank. A single, high ouput emitter over the center of the tank may work wellif you have rockwork against the back wall of the tank, but it will produce a ton of shadows if you build a rock column up the center of the tank.
Re; Color.

Originally Posted by TimmyD16

Very true, but given that the 6500K has a maximum lumen output of 1400 and the 20000K a max of 1550, 150 difference or a mere 12 spherical candle powers... I have to assume that decreasing 150 lumens worth of blue light wouldn't change a 20000K to a 6500K, there has to be something more there yes? some added spectrum somewhere, be it red yellow green or orange... and with this added spectral quality somewhere wouldn't it lead to a more full spectrum LED array? Resulting in something that more resembles the sun?

The important thing is which wavelengths are increased or decreased when the color temp is changed. To go from 6500K to 20000K you can increase the output at 440nm, or at 460nm or even at 500nm. Or, you can reduce output at 670nm, or 600nm, etc. you could even reduce 660nm (a wavelenght that is often beneficial) and increase output at 530nm (not useful at all, IIRC).
It's not the color temperature alone, it's the combination of wavelengths that make up the temperature that matters most. It's not an entirely accurate way to look at it, but color temperature is like an average of the wavelengths. The average of 8 and 12 is 10, as is the average of 7 and 13 or 6 and 14. Your organisms need some 6 and a lot of 7, plus a little 10,11 and 13. You can add some 12 if you think it looks good but the organisms don't need it. Buying a lamp that states "average;9" may get you a great bulb that meets the needs outlined above, or it may be just equal parts of 6 and 12.

[TimmyD16]

Dave, I never thought of considering that, and that is a huge misstep on my part as I will be doing a column of rock around a central overflow, which makes me lean towards multiple lower power panels... Do you think I would still achieve comparable PAR?

As for light colors, your explanation truly makes sense, as I understand the logic behind your number representation... It's pretty frustrating how arbitrary the Kelvin scale is...

[TimmyD16]

^^ anyone on the question of PAR? will it be comparable as long as the total wattages are the same?