Aquatic Plant Forum banner

Experiments with light

7K views 30 replies 8 participants last post by  AaronT 
#1 · (Edited)
So I got interested in experimenting with light that REALLY covers the wavelengths that are best for the photosynthesis (400-700 nm). I don't doubt that any strong bulb will grow plants just fine, but let's see what I find with a mix of crazy colored bulbs. I will mix the weird color bulbs with bulbs that have more green so the light looks natural.

For starters, here's a picture of how the spectrum of my favorite bulb (Giesmann Midday) looks like:


That spectrum provides very little light around 640-660 nm which is where the second peak for the photosynthesis is. But the light looks very pretty (because of the greens in it). Keep in mind that there is a view that red light is being absorbed by the water so aquatic plants don't really need it. That is why, by the way, ADA bulbs have blue (good for plants) and green (good for people's eyes) but not really red. Well, that is one reason to want to experiment with a true Full Spectrum light setup. But also there is another very interesting thing - there is a way to boost the photosynthesis by using a wavelength that we normally ignore. Read on.

Here's what bulbs I got:

Giesemann Aquablue: Got these to balance the weird colors of the next two bulbs. They have both blue and green in the spectrum. In real life the light doesn't really look that blue though. Side by side with a Giesemann Midday this bulb does look blue but something like 9000K, not too strong of a blue:


Korallen Zucht Fiji Purple: Got this one because of the broad coverage of the red spectrum. The bulb apparently gives out a horrible purple color. Must be mixed with at least 3 bluer bulbs, some folk say even 4. Note the red peak. Such high red peak is not found in any other fluorescent bulb. It is about 2 times higher than what the photosynthesis uses in the red area.


Giesemann Lagoon Blue:


Note the 700 nm peak. That's pretty rare to see. Apparently there is an interesting phenomenon with the area around 700 nm. If plants are exposed to both the blue/red light (the normal range that we consider good for the photosynthesis) AND ALSO exposed to light around 700 nm and higher there's a boost of the photosynthesis:


Besides the rare 700 nm peak this bulbs also has huge blue/green wavelengths. This produces light that supposedly looks like a green, clear tropical sea (turqoise):


So. I am interesed to see if I can manage to concoct a normal looking light from all these bulbs. Just in case I have two other bulbs that are mainly blue. From a small experiment I did the last few days I can see that a specialty grow LED light (containing only red and blue LEDs) seems to boost the photosynthesis (could not figure out if the boost was from the strong directed light or the spectrum). But the color of the light is horrendous. So the mixing of the light to achieve someting normal is important.

Also what I'm very interested in now is to get the club PAR meter. Since it measures the amount of light falling inside the range 400-700 nm I should have some huge PAR readings by combining all these bulbs.

One thing that is important to note is that having funky wavelengths and exotic bulbs will do nothing if the intensity is not good. I will be using these bulbs over a 30 gallon tank. 4x24W (96W total) or 5x24W if I have to should be strong.
 
See less See more
6
#2 ·
interesting study.
 
#3 ·
Niko, if you figure out how to make it that teal tropical white/ blue in T8 or T12s or T5s let me know! I want that for my large cichlid/ moss tank. Hey by the way, long time no see, good to see that you guys are still doing cool projects :)
 
#4 · (Edited)
The Fiji Purple will need to be used with bulbs that have a lot of green not blue to make the visual color more pleasing. Many bulbs in the 5000K - 6700K range have lots of green. Just about any Coralife bulb will have lots of green.

400nm to 700nm is the human visual spectrum not photosynthetic. The middle of that range doesnt do much of anything for photosynthesis.


A university study in the Journal of Plant Physiology found that many plants subjected to infrared light causes photosynthesis to stop.
 
#5 ·
fiji purple with blue will do good for saltwater but will look weird with plants.
 
#6 ·
Ah! Nice suggestions. I didn't think about the obvious - red needs green to turn toward white. Well, thank goodness green is what a "pretty to the eye light is" and such bulbs won't be that hard to find. Thank you both for the advice!

My hope with buying all these bulbs is to have time to get the tank running and knowing well how it reacts to fertilzers, light period, etc. Then try to change the bulbs without changing anything else. And see if I can catch any differences. Not too scientific, but if you have a feel for the tank you will be able to catch the changes.

I don't expect some revolutionary findings. But I suspect that there will be interesting observations - like for example some plants REALLY love this or that bulb (wavelengths). This has to do with what Joey said in another thread - it is not only the PAR, but the PUR that matters (that part of the 350-750 nm spectrum that the plants actually use). Apparently different plants use different parts of the spectrum. So it will be interesting to use different bulbs and see what they do.

Now, keep in mind the spectrum graphs that we all find could be misleading. From what I understand the Fiji Purple has only one single graph and it may not be reliable. Beats me why Sylvania, the company that makes the Fiji Purple bulb or Korallen Zucht, the company that markets it are not more upfront about the spectrum. In any case - that casts a shadow of doubt on all other graphs I think. I just believe that they are close enough to what the light actually is. And as long as the bulbs come with consistent quality it should be all good for our hobby purposes.

FossilGuy,

Hi to you too. I remember you moved away a couple of years ago. I too hope to be able to get the amazing green color from that light. After what Newt and Joey said it looks like there will be some serious mixing of bulbs going on. At least the intensity should be good I guess, haha.
 
#7 · (Edited)
............Now, keep in mind the spectrum graphs that we all find could be misleading. From what I understand the Fiji Purple has only one single graph and it may not be reliable. Beats me why Sylvania, the company that makes the Fiji Purple bulb or Korallen Zucht, the company that markets it are not more upfront about the spectrum. In any case - that casts a shadow of doubt on all other graphs I think. I just believe that they are close enough to what the light actually is.............................................
Ain't that the truth. I was wondering that about that Giesemann Lagoon Blue bulb that had the 700+nm spike - is the red really that high in the nm range. What I find hard to believe about the Fiji is not the area the blue and red show up in but rather the strength of the emissions. Its fantastically powerful. Keep in mind that the Fiji spectral graph is a normalized spectral plot while the Giesemann Lagoon Blue is a relative energy spectral plot. Apples and oranges.

Yes, Sylvania does make the Fiji Purple. (Too bad I don't work there anymore. Those lab guys were very friendly.) In addition to the normalized plot of the Fiji there is the ATI comparison chart.


Keep in mind that Fiji won't appear that bright to you as it has little green and humans eyes aren't that sensitive to blue and red.

I know of a common T5HO bulb that has a good amount of green and I may have some in the cellar. Its a GE Starcoat 5000K.

Another point from a university study showed that blue light optimizes plant leaf development (makes for more compact and bushy) and red light stem elongation (can make plants long and gangly).

The shorter wavelength blue light penetrates water better and more quickly than red, which is slower and absorbed more quickly. Chlorophyll traps blue and red light and a photon of each provides the same photosynthetic energy to the plant but is more efficient with red light at 650 - 675nm. Blue is used at the same rate as red and it is more available for reasons mentioned above.
 
#9 ·
Newt,

Now is the time to explain what is "normalized spectral plot" and "relative energy spectral plot". That indeed sounds like an important distinction. Maybe that super high the red peak of the Fiji bulb is not something so off-the-wall as I think.

Also do you have any counter intuitive suggestions on how to test all these bulbs? My idea is to run just the Middays for starters until I get a feel for the tank. Then add one of the exotic bulbs and see what changes. Then swap exotic bulb 1 with exotic bulb 2 and see what happens. Then add both exotic 1+2. But there maybe a more reasonable way.

Also would it make a difference to run this or that exotic bulb longer than the other?
Or the judge the effect of the red spectrum in a shallower tank because of the issue with light penetration?
Should I try to always keep the wattage the same no matter what mix of bulbs I have (otherwise the increased intensity may result in better growth)?
 
#10 · (Edited)
Fluorescent bulbs marketed for aquaria are often more expensive and not necessarily better than generic versions. They are also not necessarily marketed correctly. Many bulbs offer spectral output graphs. However, many of these graphs are measured in relative power on the Y-axis rather than a known reference like watts per nanometer per 1000 lumens. All that 'relative power' lets you know is that 100% is the highest peak at a given nanometer and all other peaks are relative to this. So, don't be fooled by nomenclature and packaging (marketing hype).

The normalized plot of spectral power output is an apples to apples comparison of the amount of energy a bulb is outputting in Watts/nanometer per 1000 lumens. The Fiji has a 0.1 microeinstiens on both the blue and red peak. Compare this to a Sylvania GroLux Std T12, the red peaks at a little over 0.02.



In the normalized plots below the T8 Philips Aquarelle (one of the most powerful T8s) the red peaks at a little over 0.05.
The Fiji peak of red is twice the intensity.


So, back to Relative Power> the marketing dept. looks for the highest peak. That becomes 100% or 1.0 or similar. All other peaks are 'relative' to the highest peak. It does not let you know the intensity of the peaks.

Giesemann has not released their data. I even requested it. A spectrometer will give you the info you need so a normalized plot can be made. I will see if I can find a web based site on how to do a normalized plot.

I will give the test parameters some thought and post later.
 
#12 ·
Wow. That was eye opening!

So we discuss what not and really don't even know what we are looking at when looking at a spectrum diagram. And all along assuming that the graph is true on top of that.

I think I understand the bulb graphs and how one can show the actual (normalized) emitted wavelength and another graph assumes that the highest peak is the max and everything else is measured as a % of it (relative). If I got that right please explain to us how it all works with the spectrum for photosynthesis.

Is this normalized or relative?


The link to the DIY spectrophotometer is really good. I had gotten it in my head to make my own a couple of weeks ago. Your link seems to show every step. Nice!

Also could you discuss this situation: So the bulb manufacturers and marketers play tricks with the graphs. So doesn't it make more, much more, sense for me to just get a bunch of LEDs with known wavelengths and concoct my own light? What are the pitfalls here? Maybe the spectrum will not be as full (too naroow peaks)? Anything else for or against making your own light with "known" spectrum?
 
#13 · (Edited)
I don't think the bulb manufacturers can outright lie about the products; intentionally providing wrong incorrect information. They can take advantage of spectral plots by shifting the spectrum a little off a tad on the X=axis or having the color in the spectrum 'off' and could easily blame it on printing 'mishaps'.

The only way to compare bulbs is with a normalized plot IMO. PUR-efficiency tells you in one number how efficiently the bulb outputs it's energy in the area of the photosynthetic action spectrum. I think the Chlorophyll graph is relative as no indication of energy(absorption) is given. What's important for a bulb to do is hit those chlorophyll spikes as close as possible. There is also the other plant components involved in growth and health>

I really don't think LEDs are there yet for the hobby. Horrible visual effects for low power consumption with intensity and poor control over the spectral output.

Keep in mind one photon of red or blue provides the same amount of photosynthetic energy and red is more easily converted but is also absorbed and slower thru water.

As far as your comparison test. I would select a bulb like your Middays as the 'always in place bulb' (base bulbs) in the same amount and swap out the test/subject bulbs. You should keep everything the same such as photo period, water changes, Ferts (except when/if you find one bulb caused more consumption), etc. If you want to test the Middays as a subject bulb too then find an off the shelf bulb like the GE Starcoat 5000K. I can get you the exact mfg number if you want it. Or just go to Home Depot and get any 5000K bulb for the base bulbs.
 
#16 ·
The bulbs arrived allright. But the Lagoon Blue bulb (the one that will supposedly make your water look turqoise green) has a different spectrum from what I thought. It does NOT have the huge 700 nm peak. Meaning I won't be able to try to see if the Emerson effect will grow some out-of-this-world plants.

Here's the spectrum I thought the Giesemann Lagoon Blue had:


And here is the spectrum from the original Giesemann packaging:


So unless I get some kind of LED that emit in the 680-750 nm range I won't be able to see any kind the Emerson effect. Now, I read a few posts on some reef forums that LEDs are not what we all like to think of them. Some folk have gone back to T5HO bulbs and they say their coral growth is again nice. Newt here also said that LEDs are not really the substitute for everything as we like to think. So this is how things seem to be with T5HO and LEDs.
 
#20 ·
I am going to make a fixture today. It will have 4 bulbs total - 2 on one switch and 2 on another switch. Two Giesemann Aquablue bulbs on one switch and one Giesemann Aquablue + an exotic bulb (Fiji Purple or Lagoon Blue). Each bulb will have an individual reflector so the fixture will put out quite a bit of light.

I can test the fixture on two tanks - a 30 gallon Oceanic cube (18" water colum) and a 30 gallon standard (14" water column). The fixture will be very close to the water surface - 1" or 2", not farther.

I too want to see how the turqoise colored Lagoon Blue looks. As we all know pictures will hardly show the color of the light but I wil post as soon as I get something going on.
 
#21 ·
the fiji purple I've used before and it gives of the 420nm purple/blue look more of the antic saltwater blue then I nice cripst white.
 
#22 ·
Allright, here are some first hand observations about using exotic T5HO fluorescent bulbs. No pictures - they won't show anything realistic unless I really take very high quaility pictures.

I made a fixture that holds 5 bulbs. There are 3 switches - for a single bulb, for 2 bulbs, and for another 2 bulbs.

I got the Fiji Purple, Lagood Blue, Giesemann Aquablue, and two 5000K generic bulbs.

Using the Fiji Purple was a complete disaster. By itself the bulb makes Java Fern look brown. Colors are way off.
Mixing the bulb with 2 bluer Aquablue bulbs does not result in anything close to good looking light. Mixing it with 3 AquaBlues + 1 5000K bulb also produces a very ugly color light.

The Lagoon Blue indeed produces turqoise colored light. But it is not as intense as you may think. It is like aquagreen color - very pale. By itself it gives the green plants a bad coloration. Mixed with AquaBlues and 5000K bulbs it produces the most beautiful light I've ever seen. Turning off the Lagoon Blue and leaving the other 4 bulbs really shows how much the Lagoon enhances the color. It adds a kind of aquagreen-blue crispness that I've never seen before.

Mixing the Lagoon Blue with the Fiji Purple + 2 AquaBlue bulbs produces a very bad color too. Adding 1 5000K does not fix the color.

Overall I can say that the Fiji Purple will probably look good with 4 AquaBlues + 2 5000K bulbs. But I don't think that justifies buying the Fiji bulb. With so much light you can grow anything you want without the "special" spectrum from the Fiji. Unless it creates some unheard of kind of quality growth I don't think the bulb is useful for the planted tank hobby. The Lagoon Blue is a different story - the light color is gorgeous.
 
#24 ·
Yes - Mix#1 (1 Lagoon + 2 5000K bulbs + 2 AquaBlue) gives a very, very unique colored light. Turning off the 1 Lagoon bulbs takes the color from "amazing" to "great, but ordinary".

I was concerned with the reds too. The red color of the neon tetras is very deep red and their blue backs glow very bright blue too. The red areas of the leaves of the Amazon that I got stay looking blood red with the above mix of bulbs. The light is both naturally looking (not exagerating any color or washing off the colors) and it has this slight aqua-turqoise tint.

I actually am running a bit of a different Mix#2 now - 1 Lagoon + 2 5000K + 1 AquaBlue + 1 Fiji Purple. That combo is just a tad less good looking than the mix#1 described above. The slight aqua-turqoise tint from the Lagoon is not so pronounced. Also the tank does not look as bright as with Mix#1 because the Fiji bulb is 24 watts but the light intensity seems very low to the human eye.

Strangely enough adding a 6-th bulb - a Midday 6000K - did not change a single thing about the color of the light with mix #1.
 
#25 ·
I ordered a Lagoon blue to try adding to my mix.

I'll be running...

1 x Giesemann Aqua Flora
1 x Giesemann Midday
1 x Giesemann Lagoon Blue
1 x ATI Aquablue Special

Right not I have a mix of aqua flora and midday so I'm curious how the blues will look. I'm hoping if anything it will increase PAR at the substrate level.

Also, FYI for anyone interested in trying the Lagoon blue I tried ordering from fishtanksdirect.com and they called me to tell me it is discontinued so you'll need to find it in stock somewhere. I found them at ReefGeek, at least I hope.

Where did you order from Niko?
 
#26 ·
Reefgeek.com

The Lagoon bulb will add something like a white cast to your tank. It is VERY washed out aquagreen. That makes the tank have a spark that no combination of the other bulbs achieve. It is indeed like a shallow tropical sea water. But do not expect your water to be colored aquagreen if you run the bulb by itself.
 
#27 ·
Thank you Niko!!! I was tired of my 67k lights on my shrimp set up! Took my time reading your experiment and decided to give it a try!
Added. 2 lagoon, 2 aquablue and 1 ATI purple (I got this one from a friend free) +3 67k that I already had.
Now I actually enjoy looking at this tank! Shrimp colors come out very nicely! Here is a before and after pics


Sent from my iPad using Tapatalk HD
 

Attachments

#28 ·
I got my new bulbs in too.

I tried running one of each and the color was a bit too blue....

ATI Aquablue special
Giesemann midday
Giesemann aquaflora
Giesemann lagoon blue

I'm replacing the aquablue special with another midday to see if that balances out the blue a little better. I may need to find a 5,000k bulb to put there.
 
#29 ·
I got my new bulbs in too.

I tried running one of each and the color was a bit too blue....

ATI Aquablue special
Giesemann midday
Giesemann aquaflora
Giesemann lagoon blue

I'm replacing the aquablue special with another midday to see if that balances out the blue a little better. I may need to find a 5,000k bulb to put there.
I'm running 3 67k, 2 G aquablue, 2 G laggon, 1 ATI purple (I dont like this one, too purple)

I did notice that the ATI aquablue is a bit more blue than the geisemann!? I'm looking some images online.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top