LED Color Aesthetics: The Emperor’s got no clothes. Or does he?

[Garage1217]

well without knowing the BIN that the "cool whites" are from, then there is no way to recommend anything to help out. Same with the blues. Definitely looks like you are running narrow optics even if it is up high, you can see spotlighting on several corals.

[sseaner]

Great photos and a great thread. I was just about to start a DIY but I put it on hold to see where this goes. Some questions:

1) Are you using any optics?

2) How high is the light above the water?

3) What percentage were you driving the blues and white?

[Neogenocide]

I think garage is on to something here. I'm seeing some major spotlighting in your pics. I would like to see some more specifics about your design, layout of led's, what optics your running, as well as how high off the water you are. I think you can definitely improve on the results your seeing right now. I also agree that the color of the whites are not where they need to be yet, but given time, it will come.


A little background on where I'm at. My tank has been up for 7 months, I quickly changed from MH's to LED's as I didn't like the look, heat, or amount of electricity they used. I ventured away from the norm with my number of led's, and layout, but have been very happy with my results. I did get quite the cold shoulder on the DIY thread, but oh well. Now comes the disclaimer, I bought my 175w Hamilton magnet ballast pendants second hand, and while I was told the bulbs were only 5 months old, I think they are at the end of their life. I sunk my money into LED's instead of new bulbs at the time. I just want you to keep that in mind when look at my pics. I LOVE the results I've gotten so far, and my brother in law also like them enough he is going all LED over his 180. I'm also in the process of getting a 180 together, and will go LED over it as well.

Scott

Pics:

Grrrr....photobucket isn't playing nice tonight. I will post them up tomorrow when it's hopefully playing nicer.

[Neogenocide]

Guess I could post up some specs on my setup,

-46G Bowfront
-37 LED's -18 Cree XR-E Royal Blue -16 Cree XR-E Cool Whites
-These are split across 2 10"x10" Heatsink USA heatsinks.
-Mounted approx 8" off the water
-60 degree optics on all LED's
-4 Mean Well LPC-35-700W constant current drivers

Scott

[jusmee]

Originally Posted by scolley

Hopefully by comparing the same subjects in a variety of lighting conditions, it will help clarify some of the color differences. To my eye, the Cool Whites do nothing but harm to the palate, except to temper the deep purple/blue of the royal blues. The pop - as we all know - from the Royal Blues it great. But if you're looking for a less blue look, there's GOT to be better color from "white" LED's out there than what these Cool Whites cast.

Lovely photos, and a great way to compare. Thanks for doing that.

Firstly, having seen way more coral in the wild (Great Barrier Reef) than under traditional lamps, I have to say the Cool Whites are closer to the wild, but I fully understand that people want a certain look in their tanks, and partly that has to be what you are used to, at least at first. Yes, the colours are more drab and less exciting.

Having read Cree's literature a fair bit, there is a whole range of Chromaticity Regions available for each group. For example, an XR-E Q5 cool white can be as follows
Group: Q5
Flux: 107
Chromaticity Regions :
WA, WB, WC, WD, WE, WF, WG, WH, WJ, WK, WM, WN, WP
WC, WD, WF, WG
WC, WD, WF, WG, WH, WJ, WN, WP


Maybe the key lies in knowing which type we are looking at? I know I read many posts from people who really like their LED lights. Maybe their LEDs are different.

[scolley]

Thanks for the replies folks. I definitely should have provided a bit more info on those pics, so here goes...

1. Height: 18-20" distance between emitters and water surface. That's an estimate, I did not measure at the time.
2. Water depth: 16". The blue/green acan and sps/clam are on the bottom. Zoa cluster 1 is 7" up from the bottom. red/blue acan and zoa cluster 2 is about 10" up from the bottom, though the acan is shaded by an overflow prefilter. And zoa cluster 3 is about 12" up from the bottom.
3. LED driving power: In all the LED pics I was using dimmable Meanwell ELN 60-48D's trimmed to approximately 750 mA (+/- 2 or 3 mA). The dimming circuit was powered by 9V batteries pushing 8.25 V. By Meanwell's specs that gives me a maximum of 80% power. In all LED pics I was running that that 80% power except the combined CW/RB pic. In that one I was running the RB at 80%, but much dimmer on the CW. How much dimmer? Don't know. Starting with 0% CW, I turned them up until until the moment that my white sand changed from purple to white, and left the CW's at that setting, whatever that was.
4. Optics: Every LED had 60 degree optics
5. LED Configuration: two rows of 12 stars. The two rows are 2" apart, and the LEDs in each row are also 2" apart, measured center to center. In each row the CW's and RB's alternate, and the two rows are running the opposite patter from each other. It looks like this:
   CW-RB-CW-RB-CW-RB-CW-RB-CW-RB-CW-RB
   RB-CW-RB-CW-RB-CW-RB-CW-RB-CW-RB-CW

I hope that helps. And as Garage pointed out I've got no idea what bin number my LEDs were from. But my guess is the Cool White came from bin number... hmmm... hmmm...

Which bin number do they put the crappy ones in?

[tuumi]

I think it comes down to expectations. When looking at any lighting system one has to identify the pros and cons of that system. Getting color that rivaled the best MH bulb/ballast combo was not something I was expecting. Electricity costs, bulb life, and the ability to ramp each color were something I had to have so I went for it. I'm very happy with those results. I'm even happy with the color I get in the tank. Though in retrospect I wish I would have done more experimenting with colors and ratios just to see the difference.
I went to the nanoreefers site to look at LED write up. In the benefits description there is no claim of "whiz bang" color. While everyone wants "whiz bang" color I realize that few things in life are perfect and for me LED get me close to my ideal lighting system for my situation.

[g8gxp]

Originally Posted by Garage1217

well without knowing the BIN that the "cool whites" are from, then there is no way to recommend anything to help out. Same with the blues. Definitely looks like you are running narrow optics even if it is up high, you can see spotlighting on several corals.

i don't think that's spotlighting. it looks like there's a center brace or something, and the corners look like the return pipes maybe?

[Koyaanisqatsi]

scolley, after seeing those pics, I can see why you don't like the LEDs much, compared to the Radium/actinic.

CW LEDs peak in the deep blue and yellow-orange through yellow-green bands, and they lack blue, aqua/green and deep red. This will make what looks like white-ish light, but it doesn't render the colors very well. The Radium peaks in several places through the spectrum, especially near aqua/green and deep red.

On your red and blue acan, the Radium/actinic pic shows more red but the LEDs look sort of burnt brown/orange. That's due to the LEDs rolling off before hitting the deep red spectrum. Reds will look brown and earthy under the LEDs.

At the same time, your GSP and Maxima are showing the LEDs to be deficient in green, but peaking in the yellow-green band. The Radium peaks a little deeper into the green, so it will pop the vibrant greens better.

My suggestion to anyone wanting to experiment is get yourself a few deep red (650-670nm), green/aqua (500-530nm) and blue (470-480nm) LEDs to work along side the cool whites. I expect an additional 25% of red and green especially will be a huge improvement.

[tuumi]

Good suggestion. I might add some more LEDS to my build. Could you provide links to specific LED that fall into the color range that you specified? I'm going to look for some now.

[Waterobert]

Could the lack of shimmer be a factor?

[thebanker]

I have to side with the T5 camp. At MAX this year there were a lot of LED tanks, some halides, but I thought the T5 tanks always looked the best. The only cool thing about LED is the "shimmer" you get that is unlike anything else, but it's not essential to me.

I bet you could create T5 shimmer using a series of fresnel lenses. Possibly.

JMO

[Garage1217]

I am not looking at the back wall or center section. I can see it on individual corals and even in the sandbed, I can see each led *LOL* Very very bad spotlighting. This may not fix the color problem, but will make the tank look funky for sure.

Originally Posted by g8gxp

i don't think that's spotlighting. it looks like there's a center brace or something, and the corners look like the return pipes maybe?

[Koyaanisqatsi]

Originally Posted by tuumi

Could you provide links to specific LED that fall into the color range that you specified?

I'll see what I can dig up. I did a quick check at Mouser.com last night and only found some odd-balls.

[scolley]

Originally Posted by Garage1217

...I can see each led *LOL* Very very bad spotlighting.

Horse hockey. If - by spotlighting - people mean poor overlap of the light cones from the LED's then that's incorrect. I've stated previously that the emitters were 18-20" above the water, which was 16" above the sand. And that the optics (snap on black cones) were 60 degree. And that the LEDS were 2" apart, alternating colors. You can do the math, draw diagrams, whatever, and know that by the time the light hits the bottom, it's all blended it. Up at the water's surface though it's not quite blended so well.

But so that you won't have to take my word for it, and so that I won't have to pull out the ugly light-cone diagrams I created before I stated building - I'll show you. At the moment my camera's battery is recharging. But I'll get pics soon.

What you'll find - as I can already see it - is that I've got pretty decent color blending. Even as close at 20" from the emitters. But you'll also see that the shadows are a totally different story. The occluding object's distance from the emitters, and its distance from the surface where the shadow is cast, makes a huge difference. But no matter what the distances (object from the light, object from the shadowed surface) the result is color bands. The distances determine the size of the bands.

In my tank the object - I'm fairly sure - is the water itself. The surface of my water stays pretty churned up. And I've got reasonable flow. Take a look at the GSP. That's flow. Not enough to rip polyps off. But it's decent flow. And it's somewhat random, and I believe that's what's casting shadows.

I'll get the pics later. You'll see. It's not spotlighting - I've got nice color blending at most every level. It's shadows, or rather the multiple shadows that any multi-point light source is destined to create.

Originally Posted by g8gxp

i don't think that's spotlighting. it looks like there's a center brace or something, and the corners look like the return pipes maybe?

Off to the right is an prefilter for my overflow box. It's over the red/blue acan. And along the back and left side are return pipes. They shadow the zoa cluster number 3, but nothing else. And there is no center brace.

[Neogenocide]

Originally Posted by scolley

Horse hockey. If - by spotlighting - people mean poor overlap of the light cones from the LED's then that's incorrect.

It's quickly becoming apparent that your are being fairly close minded about this whole thing. If your not getting the answer of "it's the LED's color spectrum", your not wanting to hear it. I'm bowing out of this conversation, but I'll leave you with a pics of my progression. Keep in mind you seeing the progression of my tank as well.

My 175w halide Pendants


After first attempt - pretty horrible huh?


Here is a 1/2 LED vs 1/2 MH pic after fixing the above problem for comparison.


And 2 shots of where I'm at today, first being taken on all Auto settings, the second taken with my amateur attempts of getting the colors correct using setting inside the camera. The second pic is darken that actual, but colors on the corals being closer. No photoshop.




I hope you get your LED's all squared away.

Scott

[kmu]

Originally Posted by larryfl1

i think this is why AI and others are heading towards the tri puck idea...

white, blue and royal blue...to get some more pop...

I was just at AI web site, they just show 16 white 6,5k and 8 blue 470nm per module...

Where did you get the info that they will include an extra color LED to the modules?

[Tang Salad]

Originally Posted by kmu

Where did you get the info that they will include an extra color LED to the modules?

Maybe here.

[tahiriqbal]

Originally Posted by kmu

I was just at AI web site, they just show 16 white 6,5k and 8 blue 470nm per module...

Where did you get the info that they will include an extra color LED to the modules?

Read this... http://www.aquaillumination.com/?p=623

From my understanding, tri-puck colour means blue/royal blue and white. Just to make it clear that both plain and royal blues colour differ from each others. They have only used royal blues and white LEDs for their previous versions and by adding plain blue means tri-puck which means latest version. I think they will update their website very soon.

[blasterman789]

went to the nanoreefers site to look at LED write up. In the benefits description there is no claim of "whiz bang" color

Again, for the billionth time, 99% of those DIY articles are written by the same guy, or copied by somebody else. There is only *1* color combination allowed in those articles; cool-white plus RB. Show me a single DIY article advocating using other color combinations or emitter brands.

I think I'm going to pick a single brand of halide bulb or T5 and write 'do it my way' articles with only that bulb. Fair?

At MAX this year there were a lot of LED tanks, some halides, but I thought the T5 tanks always looked the best.

Because 99% of them were running cool-white mixes. Doesn't it seem a bit stupid that the non LED tanks were all lit with a huge variety of bulbs, but the LED tanks are lit with with a uniform color combo?

My local reef shop hates every LED light they've seen and refuses to us them, but they did like my RB / 3500k mix. They beat me with their new radium / T5 combination lid, but it was close.

My suggestion to anyone wanting to experiment is get yourself a few deep red (650-670nm), green/aqua (500-530nm) and blue (470-480nm) LEDs to work along side the cool whites.

I have a better idea. Ditch the stupid cool-whites and use warmer emitters. The excessive green/yellow green in cool-whites is most of the problem, and there is no scientific reason to use cool-whites on your tank other than you own Cree stock. You can't produce deep warm colors whe you're blasting huge amounts of green, and a neutral or wamr emitter is not a cool-white with red in it. Another urban myth.

Try putting 6500k T5's on your tank and not complaining about color. Why the hell would you not expect 6500k bin emitters to not have the same problem?

But no matter what the distances (object from the light, object from the shadowed surface) the result is color bands.

If you're talking about 'disco shadow' problem it's the result of having evenly spaced emitters of different color. Again, we can thank the DIY crowd for pushing this bad design in their 'do it my way' articles. Also, every shot Neogenocide posted shows the problem to some extent, along with the sterile LED color we're complaining about.

To fix the problem from the start and not get into the overlap nonsense all you have to do is run emitters in clusters rather than even spacing, and this fixes a lot of it because the light sources are then closely collimated. If you have a 1:1 color mix then bump different colored emitters together in pairs - easy. I use a 2:1 mix (if I'm using kiddie 3watt emitters which I rarely do anymore) and so I have a two RBs clustered around a neutral. Eliminates most of the disco color problem in shadows.

Again, the DIY crowd is being stubborn about this because it means admitting they promoted the wrong design, but LED makers like Aqua Illumination have figured it out as well.

It's quickly becoming apparent that your are being fairly close minded about this whole thing.

No he's not. He sees the emperor in his birthday suit. Try getting this color with your cool-whites:

[tuumi]

Originally Posted by Tang Salad

Maybe here.

What about this combo? Scroll down until you see the reef heading. Looks like they are doing their homework trying to get the color just right for reefing.

http://www.cutter.com.au/products.php?cat=Cree+XPG

Any idea what kind of coverage a 7up would provide?

[jusmee]

Originally Posted by tuumi

What about this combo? Scroll down until you see the reef heading. Looks like they are doing their homework trying to get the color just right for reefing.

http://www.cutter.com.au/products.php?cat=Cree+XPG

Any idea what kind of coverage a 7up would provide?

Well that's confusing. They have a big graph showing 4 white and 1 Royal Blue XPE combined (note no blue, just Royal Blue), but none of the choices that you can buy actually implement that combo.

[Koyaanisqatsi]

Originally Posted by jusmee

Well that's confusing. They have a big graph showing 4 white and 1 Royal Blue XPE combined (note no blue, just Royal Blue), but none of the choices that you can buy actually implement that combo.

Naw, that graph is barely different than the graph of the stock cool white Crees. It's got a higher peak in the royal blue, as would be expected by adding the RB LED. But the problem doesn't lie in the RB. It's the missing blue to blue-green and the deep red that is causing so much dislike for LEDs.

And adding more RBs to get fluorescence is NOT a good direction. Some reds are not fluorescent, so this does not work for them. The light should look good without the RB, then add the RB to excite the glowy stuff.

blasterman789, your passion for the subject is appreciated and your knowledge is very appreciated. But you really need to be an expert without being a dick about it. I'd like to hear what you have to say, but please tone it down a bit. Thanks.

I have a better idea. Ditch the stupid cool-whites and use warmer emitters. The excessive green/yellow green in cool-whites is most of the problem, and there is no scientific reason to use cool-whites on your tank other than you own Cree stock. You can't produce deep warm colors whe you're blasting huge amounts of green, and a neutral or wamr emitter is not a cool-white with red in it. Another urban myth.

That _might_ be a good idea if you're buying all your LEDs now. But for those who have already invested a few hundred bux into CW LEDs, that doesn't work so well.

I have yet to see any product (or idea) address the lack of deep red. Steve's pics show it clearly. Even WW LEDs lack the deep red. But if you pair up deep red LEDs with the WW LEDs, you risk flooding the tank with orange light. And you'd need more RB and blue-green to make up for the spectrum of the WW LEDs.

The key to this problem is Color Rendering Index. And the only way to get it is to assure a smooth and complete spectrum across the colors you're interested in. The CW LEDs lack deep red, blue green and blue. The WW LEDs just add a ton of yellow/orange to the mix, which doesn't help much.

[scolley]

Originally Posted by Neogenocide

It's quickly becoming apparent that your are being fairly close minded about this whole thing. If your not getting the answer of "it's the LED's color spectrum", your not wanting to hear it.
Scott

I’m not even going to rise to that bait. Though I DO appreciate the inclusion of some good FTS before/after, including HM combo shots. Thank you.


But now I realize that I should have gone further in my request for pics… we need to see pics with corals in them. The more colors the better. Live rock shots are helpful, but SURELY there’s someone out there that can say… “Hey, look at my beautiful corals under my MH combo. And now look! Same thing under Cree LEDs! Nice huh?”
I wonder why we aren’t seeing those. We’ll have 2,000 views of this thread pretty soon. So the reason is not because no one is paying attention…

[scolley]

BUT ON TO A NEW TOPIC...

Ok, just so we can put this coning idea to bed…

Here’s a picture of coning. This is the same light, 60 degree optics, with the emitters a silly 12” off the white paper. Both the cool whites and the royal blues have been turned down pretty low so you can see the effect more clearly – distinct separation of colors.



If I were not tired of this topic, I’d have swapped the 60 degree optics with my 40’s. But we all know what we would have seen – the same picture, with more distinct color separation.

This next pic has the emitters raised to a whopping 17” off the paper. Color separation is getting REALLY hard to see. Good thing I wasn’t using 40 degree optics…

Now we see below the lights 22” off the paper. Nice blending of color. But in the pic below it, without raising the light, I stuck a small white sheet 4” off the bottom. And what’s that around the edges of the shadow? Coning? No. That’s the multiple shadow effect that is going to happen in ANY multi-point, tightly light source. It’s as Garage posted… this is an unavoidable artifact of the design.





Next we see the same thing, this time a close-up of the raised white square with the emitters 32” off the white paper. This would correspond to the very highest things in my tank, from the pics I took (see my last post with pics) with the emitters 18-20” from the water. The emitters are about the same distance from the paper square as from many of my corals in the pics I posted yesterday.



But I DO have to apologize for the fact that the white balance of these pics is ALL off. I had to adjust it because of the wacky lighting I’m trying to photograph in tonight. And what’s being lost is the REALLY important fact that those rectangular shadows are ALTERNATING in color, blue shadow, brown shadow, blue, brown, etc. If anyone doubts this, please speak up. Proof is just a few photographs away in daylight tomorrow, and I'm happy to provide it if needed.

This point is important however, because I believe it explains the clearly varying coloration in my tank that was mistaken for coning. With the water flow at the surface bending light, it created shadows… shadows of alternating colored borders – just like this paper.