LED spectrum/visual perception

[Jase4224]

Hi guys, does anyone know if there is a chart that shows what any specific LED spectrum actually looks like to the human eye?

I have been using this tool: http://spectra.1023world.net/

It has a number of graphs but I don’t understand them all.

I am trying to design my Orphek led bars.

 

[Big G]

@saltyfilmfolks

 

[oreo54]

Interesting question..
Think using the overall "tone" swatch and hex color is pretty close..

This is what Orphek " blue sky" looks like..
https://www.color-hex.com/color/3358ff

 

[Dana Riddle]

I've just received a couple of Orphek strip LED luminaires for testing. The Reef Daylight Plus has an interesting spectral composition ranging from UV (~390nm peak) through infrared (~740nm peak.) I need a few days to get analyses done, charted, and pieced together in a coherent manner. In the meantime, PM me and I can discuss what I have.

 

[Dana Riddle]

This is the spectral characteristic of the UV LED.

 

[Jase4224]

Interesting question..
Think using the overall "tone" swatch and hex color is pretty close..

This is what Orphek " blue sky" looks like..
https://www.color-hex.com/color/3358ff

I suspected that the little square in the top left corner was the colour, just wasn’t quite sure.

Beautiful tank btw!

 

[Jase4224]

I've just received a couple of Orphek strip LED luminaires for testing. The Reef Daylight Plus has an interesting spectral composition ranging from UV (~390nm peak) through infrared (~740nm peak.) I need a few days to get analyses done, charted, and pieced together in a coherent manner. In the meantime, PM me and I can discuss what I have.

The problem I have with the daylight plus is that when you have so many colours on a strip light I can’t imagine they would blend well. Secondly, having certain colours so far apart makes me wander if there is any benefit to having them since only parts of the tank are receiving those wave lengths. The daylight plus to me seems to have a weird distribution of the colours along the strip.

I’m not dissing the product just discussing what I see by looking at the layout.

That UV diode has an interesting little bump in the far red spectrum!

Danna I would love to see your results for this light.

 

[Dana Riddle]

The problem I have with the daylight plus is that when you have so many colours on a strip light I can’t imagine they would blend well. Secondly, having certain colours so far apart makes me wander if there is any benefit to having them since only parts of the tank are receiving those wave lengths. The daylight plus to me seems to have a weird distribution of the colours along the strip.

I’m not dissing the product just discussing what I see by looking at the layout.

That UV diode has an interesting little bump in the far red spectrum!

Danna I would love to see your results for this light.

I'll take a look at color mixing when I get a chance (which might be a while - I cannot believe how busy I am in retirement. I had concerns that I might be bored. No way! My schedule are the next 8 days might take more energy than I have left.)

 

[Jase4224]

The problem I am having is that I want a balanced spectrum of UV/Violet, Blue and Cyan. That is easy but I don’t want the tank to be too BLUE as I like a Blue/Cyan look of the natural ocean let’s say 485nm would be nice!

This is easy BUT this is where it gets complicated ..
1) I don’t see any point having less than 4 of any LED colour otherwise the tank is not getting an even spread of that colour.
2) Also for photosynthesis the peak wavelength needs to be around 450-460. And this needs to be evenly spread.
3) as you can see by the bottom pic I have used a lot of Cyan but the colour chart in the top pic still shows a dark blue light.

This is confusing to me as to my knowledge the human eye is much more sensitive to the green wavelengths. So to my imagination this light should not be blue/purplish to look at. Confusing.

*this is a design of my own utilising 42 LED’s as per the Orphek or2 150*

 

[oreo54]

I suspected that the little square in the top left corner was the colour, just wasn’t quite sure.

Beautiful tank btw!

Thanks.. not mine.. Was borrowed from Orphek...
Just wanted to show (well w/ a few caveats) that the swatch and actual had relevance..

you don't have ALL those diode choices at Orphek btw...
There is a LIST somewhere.

You know, it's easy enough to build one yourself and get the added advantage of dim-ability..
Is this supplemental or primary lighting?

LED UV (410nm) [120°] x3
LED Blue (460nm) [120°] x2
LED Cyan (490nm) [120°] x5
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 280 lm
Radiant flux : 2,915 mW
PPF : 10.8 umol/s
TCP : ‑ K
CRI : ‑
λp : 412 nm
Color : #338EFF

Found the list:
Each light bar has 36 - 3 watt LEDs. The available LED colors are:
6300K
12,000K
410
430
460
490
595
630
660
730

 

[Jase4224]

Thanks.. not mine.. Was borrowed from Orphek...
Just wanted to show (well w/ a few caveats) that the swatch and actual had relevance..

you don't have ALL those diode choices at Orphek btw...

Thanks. The choice of diodes I am using is a list of what Orphek uses, they put it in the or light bar thread a few weeks ago. I understand that the program I am using also has colours Orphek doesn’t have

 

[Jase4224]

Thanks.. not mine.. Was borrowed from Orphek...
Just wanted to show (well w/ a few caveats) that the swatch and actual had relevance..

you don't have ALL those diode choices at Orphek btw...
There is a LIST somewhere.

You know, it's easy enough to build one yourself and get the added advantage of dim-ability..
Is this supplemental or primary lighting?

Sorry I forgot to mention this is my primary lighting. I’ll probably get 3 x bars with two being my Blue lights and the third being a white channel. When my blues are on without the white in the morning and arvo I don’t want it to be TOO blue.. that’s what I’m trying to figure out.

I considered DIY and I’ve made a few LED units myself, but for the quality Orphek bars are cheaper and just more convenient.

I am making my own fuge light though that’s easy because you can focus just on photosynthesis. For that I used 8 x 660nm, 6 x 450nm and 2 x neutral white.

 

[oreo54]

Whites have a lot of Royal blue.. esp high K ones..
Wouldn't really need it in the blue bars..
It overpowers the tone anyways..

not getting into "best" nm's here ...

Manuf. high K whites..
MIXING LIST
----------------------------------------
LED RoyalBlue (440nm) [120°] x1
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 1,569 lm
Radiant flux : 5,710 mW
PPF : 24.6 umol/s
TCP : 11630 K
CRI : 72
λp : 453 nm
Color : #B4B2FF

THEN design 2 bars to supplement ..
you need to think as a system not separate bars..
1 white bar 12 diodes (well 13 w/ the blue adjustment) 2 color bars 12 diodes
* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x10
LED RoyalBlue (450nm) [120°] x1 (part of white to shift K)
LED Cyan (490nm) [120°] x14
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,279 lm
Radiant flux : 12,732 mW
PPF : 50.3 umol/s
TCP : ‑ K
CRI : 56
λp : 412 nm
Color : #4BB1FF
https://www.htmlcsscolor.com/hex/4BB1FF

you want that "green blue" look?
Or just more natural blue look?


Dimming whites by 1/2


Sorry had charts but they got scrambled and don't feel like rebuilding them atm
but a new one 12 120000-ish K and 24 cyan

Adding 410nm violets doesn't really change the overall tone much, at least a s calculated

LED UV (410nm) [120°] x8
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x16
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,375 lm
Radiant flux : 12,478 mW
PPF : 49.9 umol/s
TCP : ‑ K
CRI : 60
λp : 412 nm
Color : #43BAFF

you can shift it more green w/ a different cyan..but unavailable..
510 cyans added
https://www.color-hex.com/color/33ffca

 

[oreo54]

Back to the available diodes..
* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x4
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x20
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,572 lm
Radiant flux : 11,770 mW
PPF : 48.4 umol/s
TCP : ‑ K
CRI : 70
λp : 493 nm
Color : #33D6FF

adding 1 more violet to each keeps the tone about the same.
Remember this is a mockup of 3 "bars" of 12 diodes each.

don't think true UV is available w/ the bars btw.

* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x6
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x18
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,473 lm
Radiant flux : 12,124 mW
PPF : 49.1 umol/s
TCP : ‑ K
CRI : 65
λp : 493 nm
Color : #3AC7FF
1:3 ratio of 410 to 490 seems ok.
This is all much easier w/ every color on it's own channel..
As to blending ect.. height and lenses will have an impact..
If you are going w/ a more shallow tank and just a bit off th water line 120 degree lenses produce cones of about 14" at 4"...at the water line (roughly)
34.6" cones at 10" from light emitting surface (really the lens)

Point is for blending and or worrying about spectrum "distribution" that has to be kept in mind..

Refraction and reflection off the glass also play a part of this.

 

[Dana Riddle]

Here's the spectral quality of the Orphek Reef Daylight Plus luminaire. Measurement taken with a Ocean Optics USB2000 spectrometer. Distance from LED to Spectralon 99% reflectance standard was 30 inches. Since there are only 2 LEDs producing radiation at ~740nm, my instincts told me there wouldn't be a peak at that LED's bandwidth - but there it is.

Here is the IR LED's output. This measurement was taken just a few inches from the spectrometer's fiber optic cable.

 

[Jase4224]

Back to the available diodes..
* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x4
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x20
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,572 lm
Radiant flux : 11,770 mW
PPF : 48.4 umol/s
TCP : ‑ K
CRI : 70
λp : 493 nm
Color : #33D6FF

adding 1 more violet to each keeps the tone about the same.
Remember this is a mockup of 3 "bars" of 12 diodes each.

don't think true UV is available w/ the bars btw.

* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x6
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x18
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,473 lm
Radiant flux : 12,124 mW
PPF : 49.1 umol/s
TCP : ‑ K
CRI : 65
λp : 493 nm
Color : #3AC7FF
1:3 ratio of 410 to 490 seems ok.
This is all much easier w/ every color on it's own channel..
As to blending ect.. height and lenses will have an impact..
If you are going w/ a more shallow tank and just a bit off th water line 120 degree lenses produce cones of about 14" at 4"...at the water line (roughly)
34.6" cones at 10" from light emitting surface (really the lens)

Point is for blending and or worrying about spectrum "distribution" that has to be kept in mind..

Refraction and reflection off the glass also play a part of this.

Basically I’m going to have two blue bars, front and back with a single white channel in between.

The blue bars on there own need to have UV/violet, blue and cyan. I want the light to look blue/cyan rather than just blue. This is the challenge..

The white bar when it’s on will brighten things up but this is easy to design around the blue bars once they are sorted.

 

[Jase4224]

Whites have a lot of Royal blue.. esp high K ones..
Wouldn't really need it in the blue bars..
It overpowers the tone anyways..

not getting into "best" nm's here ...

Manuf. high K whites..
MIXING LIST
----------------------------------------
LED RoyalBlue (440nm) [120°] x1
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 1,569 lm
Radiant flux : 5,710 mW
PPF : 24.6 umol/s
TCP : 11630 K
CRI : 72
λp : 453 nm
Color : #B4B2FF

THEN design 2 bars to supplement ..
you need to think as a system not separate bars..
1 white bar 12 diodes (well 13 w/ the blue adjustment) 2 color bars 12 diodes
* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x10
LED RoyalBlue (450nm) [120°] x1 (part of white to shift K)
LED Cyan (490nm) [120°] x14
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,279 lm
Radiant flux : 12,732 mW
PPF : 50.3 umol/s
TCP : ‑ K
CRI : 56
λp : 412 nm
Color : #4BB1FF
https://www.htmlcsscolor.com/hex/4BB1FF

you want that "green blue" look?
Or just more natural blue look?


Dimming whites by 1/2


Sorry had charts but they got scrambled and don't feel like rebuilding them atm
but a new one 12 120000-ish K and 24 cyan

Adding 410nm violets doesn't really change the overall tone much, at least a s calculated

LED UV (410nm) [120°] x8
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x16
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,375 lm
Radiant flux : 12,478 mW
PPF : 49.9 umol/s
TCP : ‑ K
CRI : 60
λp : 412 nm
Color : #43BAFF

you can shift it more green w/ a different cyan..but unavailable..
510 cyans added
https://www.color-hex.com/color/33ffca[/QUOT

Back to the available diodes..
* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x4
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x20
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,572 lm
Radiant flux : 11,770 mW
PPF : 48.4 umol/s
TCP : ‑ K
CRI : 70
λp : 493 nm
Color : #33D6FF

adding 1 more violet to each keeps the tone about the same.
Remember this is a mockup of 3 "bars" of 12 diodes each.

don't think true UV is available w/ the bars btw.

* MIXING LIST
----------------------------------------
LED UV (410nm) [120°] x6
LED RoyalBlue (440nm) [120°] x1
LED Cyan (490nm) [120°] x18
LED CoolWhite (8000K) [120°] x12
----------------------------------------

* SIMULATION DATA
----------------------------------------
Luminous flux : 2,473 lm
Radiant flux : 12,124 mW
PPF : 49.1 umol/s
TCP : ‑ K
CRI : 65
λp : 493 nm
Color : #3AC7FF
1:3 ratio of 410 to 490 seems ok.
This is all much easier w/ every color on it's own channel..
As to blending ect.. height and lenses will have an impact..
If you are going w/ a more shallow tank and just a bit off th water line 120 degree lenses produce cones of about 14" at 4"...at the water line (roughly)
34.6" cones at 10" from light emitting surface (really the lens)

Point is for blending and or worrying about spectrum "distribution" that has to be kept in mind..

Refraction and reflection off the glass also play a part of this.

how do you get the colour #??

 

[Jase4224]

Here's the spectral quality of the Orphek Reef Daylight Plus luminaire. Measurement taken with a Ocean Optics USB2000 spectrometer. Distance from LED to Spectralon 99% reflectance standard was 30 inches. Since there are only 2 LEDs producing radiation at ~740nm, my instincts told me there wouldn't be a peak at that LED's bandwidth - but there it is.

Here is the IR LED's output. This measurement was taken just a few inches from the spectrometer's fiber optic cable.

Interesting. If you move the spectrometer along the length of the bar I bet the peak will change.. I still don’t fully understand why the 740nm are there and if they really benefit.

 

[oreo54]

how do you get the colour #??

Program calculates it...

 

[Dana Riddle]

Interesting. If you move the spectrometer along the length of the bar I bet the peak will change.. I still don’t fully understand why the 740nm are there and if they really benefit.

740nm will relax Photosystem I (it absorbs at a max of 700nm thus called P700.) This will could help prevent damage to Photosystemm II .