Sciencemadness Discussion Board - Anyone here have yellow LED light, and light-green paint?

Anyone here have yellow LED light, and light-green paint?

Neal - 18-2-2025 at 07:43

I wanna know what blue light + yellow light makes. Blue paint and yellow paint make green paint.

Many websites incorrectly say blue light and yellow light make white light. Because yellow light is 50% red and 50% green light, and they argue red blue and green light makes white light. But a light that is 50% blue and 50% yellow is actually 50% blue, 25% red, and 25% green, so that doesn't make white. It's not the same as 33/33/33.

This is where red light and blue light make magenta, rather than purple. But I can't find on the Internet for someone that made the blue and yellow counterpart. So looking for someone that has yellow light.

[img]https://www.sciencemadness.org/whisper/files.php?pid=0&aid=105108[/img]

Now, red light and green light make yellow light, but what does red paint and green paint make? Well, if you look at these 2 color pinwheels, the "spectral" green light and "green paint" are not the same paint. The green paint, is a much darker green. Spectral-green is more of a lime-green. So I'm wondering what color paint do you get if you mix red paint with lime-light green paint?

What's so ironic to me is how red light and red paint always seem to be the same red, and to some extent, same with blue paint and blue spectral light, but I've never seen green paint being expressed the same as spectral green light.

Has anyone experienced with liquids with dyes as colors? Do they follow the paint model or light model?

[Edited on 18-2-2025 by Neal]

Neal - 18-2-2025 at 07:45

Now, HTML has an answer to this, but I don't know how realistic it is to life.

HTML's RGB(255,0,255) is magenta. But, if you use RGB(128,0,128) that is a purple or dark purple.

ChatGPT claims the 1st is 100% red and 100% blue, whereas the 2nd is 50% red and 50% blue...

1 seems to be the light model, and the other the paint model.

I already tried with blue and yellow light. Which would be 50% blue 25% red 25% green. But neither turned out to be white-ish or green-ish.

Deathunter88 - 18-2-2025 at 13:09

The way we view color is much more complex than just the idea of pure hues and wavelengths. Our eyes also do not see color linearly, and the color we percieve is affected by a lot of other factors.

This video might be a good starting point:
https://www.youtube.com/watch?v=FTKP0Y9MVus

[Edited on 18-2-2025 by Deathunter88]

Twospoons - 18-2-2025 at 13:17

The first white LEDs were made using a blue LED and a yellow phosphor, making a rather bad 'whitish' LED. This only works because of the rather broad overlap between red and green color receptors. "yellow" around 560nm is picked up by both receptors, add in some blue and you have the perception of white. Color rendering is awful though, because of the way many objects reflect different colors - e.g. a 'red' object may look dark if if does not reflect 560nm yellow light.

For a deep dive into modern LED lighting :YujiLED
These guys make fantastic White LEDs - i used some to create the lighting for my wifes art studio.

[Edited on 18-2-2025 by Twospoons]

woelen - 18-2-2025 at 13:56

Quote: Originally posted by Neal

Now, HTML has an answer to this, but I don't know how realistic it is to life.

HTML's RGB(255,0,255) is magenta. But, if you use RGB(128,0,128) that is a purple or dark purple.

ChatGPT claims the 1st is 100% red and 100% blue, whereas the 2nd is 50% red and 50% blue...

1 seems to be the light model, and the other the paint model.

I already tried with blue and yellow light. Which would be 50% blue 25% red 25% green. But neither turned out to be white-ish or green-ish.

It all depends on intensity. Yellow light and blue light can be viewed as white light.

Full intensity yellow light on 8 bits per color is (r, g, b) = (255, 255, 0)
Full intensity blue is (r, g, b) = (0, 0, 255)
Mixing them is adding them and this gives (r, g, b) = (255, 255, 255), which is white.

The yellow light is brighter to our eyes than the blue light, but this is not because it has 2 times level 255 for a primary color. Our eyes are more sensitive to red and green light than to blue light, we perceive them as brighter at the same intensity.

A 50% red + 50% green can of course also be (128, 128, 0), but this is not bright yellow, it looks darker, more brownish. If you mix that with full intensity blue, I expect a pale blue color, not purely white.

Neal - 18-2-2025 at 17:55

Quote: Originally posted by Twospoons

The first white LEDs were made using a blue LED and a yellow phosphor, making a rather bad 'whitish' LED. This only works because of the rather broad overlap between red and green color receptors. "yellow" around 560nm is picked up by both receptors, add in some blue and you have the perception of white. Color rendering is awful though, because of the way many objects reflect different colors - e.g. a 'red' object may look dark if if does not reflect 560nm yellow light.

For a deep dive into modern LED lighting :YujiLED
These guys make fantastic White LEDs - i used some to create the lighting for my wifes art studio.

[Edited on 18-2-2025 by Twospoons]

Actually I wasn't going to believe what you said until Woelen brought up that our blue cones in our eyes don't work the same as red and green cones.

And in your map, shows the blue cone further out than the others.

The fact that we have blue cones, instead of violet, is interesting. For 1 thing, it explains:

-In the Doppler effect, shouldn't it be red and violet shift, not red and blue shift?

-And is it really red green and blue light to make white light, when it should be red green and violet light?

After all, green is halfway in the spectrum between red and violet.

1 thing I will later experiment with are liquids, and food dyes. In big chain grocery stores, sell dyes, that are red blue green and yellow. And dye it to water. And see if mixing equal amounts leads more to the paint theory or light theory.

[Edited on 19-2-2025 by Neal]

Sulaiman - 18-2-2025 at 18:29

may I suggest:
Cyan magenta and yellow printer ink
printers do a pretty god job at overlaying inks (colour printing)
the inks could be mixed in a solvent,
sometimes the solvent changes the optical properties
some liquids have a different colour with transmitted light than with reflected light.
(try water, alcohol, cooking oil, gasoline...whatever you can get)

an old crt colour monitor or tv tube could be used as an rgb light source
rgb led strip lights are a convenient bulk source of additive colour light

I learned with watercolour paints
then filament bulbs and coloured cellophanes.

[Edited on 19-2-2025 by Sulaiman]

Twospoons - 18-2-2025 at 18:57

This idea that red+green+blue = white is a very anthropocentric thing - because human eyes have receptors centered on the red, green and blue and thats how we perceive and interpret color. Bees would have a different idea of what makes white, as they can see UV. The Mantis shrimp has 12 different color receptors, ranging from deep UV to near IR. "White" for the mantis shrimp would need all of those colors.
True white light contains every visible wavelength, not just a few select ones.

Neal - 18-2-2025 at 19:32

You know I should ask this question while we're on it. It's about the color pinwheel. That what 1 absorbs for a color, reflects for another color.

But why is the color pinwheel only in the visible spectrum? I'd like to see someone make a color pinwheel that extends from say, IR to UV.

Twospoons - 18-2-2025 at 20:07

Oh, so you're really interested the difference between additive and subtractive primaries, wrt to color mixing?
UV and IR don't really come into play here, probably because you would not notice any real difference between UV at 360nm and UV at 400nm, as only the blue receptor is affected. Likewise at the red end 750nm and 700nm would look the same, as only the red receptor is affected. So the color pinwheel really only needs to encompass blue though red to represent the color space we can perceive.

I think you should read this : https://en.wikipedia.org/wiki/CIE_1931_color_space

[Edited on 19-2-2025 by Twospoons]

Neal - 18-2-2025 at 20:55

Quote: Originally posted by Twospoons

Oh, so you're really interested the difference between additive and subtractive primaries, wrt to color mixing?
UV and IR don't really come into play here, probably because you would not notice any real difference between UV at 360nm and UV at 400nm, as only the blue receptor is affected. Likewise at the red end 750nm and 700nm would look the same, as only the red receptor is affected. So the color pinwheel really only needs to encompass blue though red to represent the color space we can perceive.

I think you should read this : https://en.wikipedia.org/wiki/CIE_1931_color_space

[Edited on 19-2-2025 by Twospoons]

No that's not my question.

My question is what would stop something that absorbs IR or UV, to reflect visible light?

bnull - 19-2-2025 at 05:29

Quote: Originally posted by Neal

-In the Doppler effect, shouldn't it be red and violet shift, not red and blue shift?

It was Eddington or Hubble (I'd bet on the former) who gave these names to Doppler effect in light. Nothing special in them, except for blue being more noticeable than violet. I believe that's why Eddington chose blue over violet.

Quote: Originally posted by Neal

But why is the color pinwheel only in the visible spectrum? I'd like to see someone make a color pinwheel that extends from say, IR to UV.

The purpose of the color wheel is to present the visible colors; if they're not visible, you can't present them. We can see some short IR and long UV, so they end up grouped to the red and violet ends of the spectrum. As for the rest, what color is the color of IR? Red, say, but what shade of red? What is the color of UV? What about X-rays and microwaves?

Quote: Originally posted by Neal

My question is what would stop something that absorbs IR or UV, to reflect visible light?

Its molecular/atomic structure. Take water, acetone, and ethanol. They are transparent, yet you can distinguish them by their absorption spectra.

Or metals. Most metals are sort of gray because they reflect most of the visible light and absorb wavelengths that we can't usually see (IR and UV), and we don't miss what we can't see. Gold absorbs blue, so the reflected light is yellowish. Copper absorbs green/blue, hence its reddish appearance.

Newton reported many experiments with prisms and colors in his Optics that you would be able to repeat. Small prisms, like the ones used in binoculars, should do. Or you can make some yourself. Go to a glazier's, ask or buy some scraps of plate glass, make wedges with them and fill with water. Glue a cover in them so they don't leak. You can also pay the glazier to make them for you. Or use microscope slides.

Check out Fowles' Modern Optics, chapters 6 and 8, for the more technical side.

I have a bunch of spectra of LEDs and gas discharge lamps that I took for an Experimental Physics class in 2019. Fun stuff if you're into that kind of thing.

Edit: I grouped the LED spectra in a single plot. The Intensity scale is arbitrary. The spectra were measured using an OceanOptics spectrometer USB4000. The LEDs were the ordinary types. The plot was generated by Spectragryph 1.2, where I chose a grey background for better visibility. The gas discharge lamps will be attached later in PDF.

Notice that the white LED uses a phosphor, like @Twospoons said. We see a peak in the blue region and a hump in the green/yellow region.

Edit2: I processed the spectra of the gas discharge lamps.

Attachment: Gas discharge lamp spectra.pdf (483kB)
This file has been downloaded 164 times

[Edited on 19-2-2025 by bnull]

Edit3: Minor correction.

[Edited on 20-2-2025 by bnull]

Neal - 19-2-2025 at 08:50

Quote: Originally posted by bnull

Quote: Originally posted by Neal

My question is what would stop something that absorbs IR or UV, to reflect visible light?

So basically, it's just a coincidence that all the non-visible absorbance also reflects non-visible light?

Like absorbing near-IR, must reflect far-IR, and absorbing far-IR, may reflect near-IR, and vice versa with UV?

Twospoons - 19-2-2025 at 14:48

Quote: Originally posted by Neal

My question is what would stop something that absorbs IR or UV, to reflect visible light?

Nothing. Perfectly possible. BaSO4 is a great example. Reflective in the visible spectrum and absorbs long wave IR.

Neal - 20-2-2025 at 06:08

Okay, I got several answers.

Finally got ChatGPT to generate a color that is 50% blue and yellow, as well as paint color of red and lime-green. The 1st looked pretty gray, 2nd looked more sunset-orange.

Now when I Google what is the most visible color, I didn't get yellow. I got green. But I found 1 Youtube video that said chartreuse green, which is 555 nm.

And so, there is a misconception. The red cones are not actually red, their peak is actually more at yellow. And so, the halfway point between green and yellow, is chartreuse green, 555 nm.

And it makes sense that out red cones are not actually red because we can't see into the IR.

[Edited on 20-2-2025 by Neal]

Neal - 22-2-2025 at 19:25

Anyone have any thoughts on this? Not sure if this can be proofread.

I asked ChatGPT, specifically without saying "paint or light" to see if it would ask me, but it didn't.

Twospoons - 22-2-2025 at 19:57

One problem you will have with this is you are limited by the phosphors used in your LCD monitor. e.g an LCD can't generate true yellow, only fake it with red + green.

Sulaiman - 22-2-2025 at 21:27

you can buy led colours from ir to uv that do not use phosphors hence have a moderately narrow spectrum
(eg kingbright L-53SYC: 590 +/- 14 nm, a typical yellow led)

Typically, if an led forward operating voltage is below 3v then it does not use phosphors,
so it will emit near monochromatic light,
of a wavelength where the eV of photons is a little less than the operating voltage at low current.

bnull - 23-2-2025 at 06:14

Quote: Originally posted by Neal

I asked ChatGPT, specifically without saying "paint or light" to see if it would ask me, but it didn't.

Don't rely too much on ChatGPT to find answers. There's no easy way to tell if they're correct or not when you don't know beforehand.

I suggest you look it up in a color science or colorimetry journal, perhaps something by the Optical Society of America (it's now called Optica, https://www.optica.org/). Optica (the OSA magazine or journal) used to have something on the physical aspects of color sensing. The few numbers I have, at least (they're buried somewhere in my papers).

On a side note, you can see if an LED has a phosphor by hooking it up to a multimeter in the volts range and shining a flashlight on it. A reading close to the operating voltage (0.5-0.7 V below it, I guess) means no phosphor, zero or close to means phosphor. I had to sort out LEDs for a project that used them as photodetectors; the project flopped when pandemics came.

Neal - 23-2-2025 at 08:20

Quote: Originally posted by bnull

Quote: Originally posted by Neal

I asked ChatGPT, specifically without saying "paint or light" to see if it would ask me, but it didn't.

Don't rely too much on ChatGPT to find answers. There's no easy way to tell if they're correct or not when you don't know beforehand.

That's why I just

"Earlier when I asked you to send me 3 square colors, of light red dark blue and dark red light blue, you didn't ask me whether it was for the "light" model of colors, or the "paint" model of colors."

And ChatGPT replies that it was the light model. So it assumes light model when not asked.

Btw, here's something contradicting.

Let's go back to that color pinwheel in inorganic chemistry. It seems to use the paint model rather than light model (spectrum).

Remember I said if something appears yellow to us, we don't know whether it reflects yellow spectral light, or reflects red & green light in a 50/50 ratio? Well, let's look at the color pinwheel.

For reflecting yellow, is absorbing purple. Well guess what, purple isn't a color in the spectrum. Violet is. But the color pinwheel doesn't seem to have violet since it seems to be of the paint model.

That could be that reflecting yellow is never the same as spectral yellow and reflecting yellow is only due to red/green 50/50 ratios.

Twospoons - 23-2-2025 at 12:34

Quote: Originally posted by Neal

That could be that reflecting yellow is never the same as spectral yellow and reflecting yellow is only due to red/green 50/50 ratios.

The only way to now for sure is to use a spectrometer, diffraction grating or prism to analyse the light.

Neal - 23-2-2025 at 13:28

Wow check this out, I asked ChatGPT to output a square of 1000 pixels by 1000 pixels, that are red and blue pixels back to back. The 1st row to be red-blue-red-blue and the 2nd row to be blue-red-blue-red.

The interesting thing is, it looks like darker purple. But after 1 zoom in, it looks light purple. And doesn't seem to change upon zooming in some more, but after zooming in you can see that it is red and blue pixels. And it matches the 1st red-purple mix it made a bit.

[Edited on 23-2-2025 by Neal]

bnull - 24-2-2025 at 12:53

Quote: Originally posted by Neal

Let's go back to that color pinwheel in inorganic chemistry. It seems to use the paint model rather than light model (spectrum).

Remember I said if something appears yellow to us, we don't know whether it reflects yellow spectral light, or reflects red & green light in a 50/50 ratio? Well, let's look at the color pinwheel.

For reflecting yellow, is absorbing purple. Well guess what, purple isn't a color in the spectrum. Violet is. But the color pinwheel doesn't seem to have violet since it seems to be of the paint model.

That could be that reflecting yellow is never the same as spectral yellow and reflecting yellow is only due to red/green 50/50 ratios.

An easy way to find is to use color filters. Buy a green filter, a red filter, and a yellow filter. Some places sell sets for telescopes, nothing really fancy or prohibitively expensive. No filter is perfect but they should be good enough.

Are you taking into account that:

the effect you see may be due (of course it is) to the individual pixels that make up the screen? A better option would be to print the checkerboard on a sheet, pin the sheet to a wall and observe it from various distances (from 30 cm to 5 m or more);
the specific shades of colors that you see are not necessarily the same seen by someone else because of the way the cones are structured and wired to the brain? My blues may be slightly bluer than yours, for example;
the color wheel is arbitrary and has more to do with physiology than with physics?

Edit: Typo.

[Edited on 24-2-2025 by bnull]

Neal - 24-2-2025 at 15:15

Quote: Originally posted by Twospoons

Quote: Originally posted by Neal

That could be that reflecting yellow is never the same as spectral yellow and reflecting yellow is only due to red/green 50/50 ratios.

The only way to now for sure is to use a spectrometer, diffraction grating or prism to analyse the light.

I don't have experience with this, how does it work? If I have a yellow object, having an instrument measure the wavelengths, most likely would get wavelengths from something other than the yellow object.

Twospoons - 24-2-2025 at 17:02

That last statement makes no sense. Can you rephrase it?

Neal - 24-2-2025 at 20:34

Quote: Originally posted by Twospoons

That last statement makes no sense. Can you rephrase it?

The yellow object can be next to a red object, the red object can be next to the blue object. So wouldn't I get false positives?

Twospoons - 25-2-2025 at 12:47

Not if you just measure the yellow object. Just like if you want to measure a book you don't measure the entire bookshelf.
Optically this would involve imaging lenses, or a mask, so you are only looking at the object of interest, rather than the average of an entire room.

The attached image is of a fluorescent bulb as seen through a diffraction grating. You will notice there are 5 distinct images, as the bulb emits light on 5 distinct wavelengths, and the grating separates them out.
Applying this to your yellow object: if it were only reflecting yellow you would see one yellow image; if it were only reflecting red and green you would see two images , one red and one green. Does that make sense to you?

Most everyday objects reflect a broad band of wavelengths, and so would appear as a rainbow colored smear when viewed through a diffraction grating.

$30_fluorescent-lights_diffraction-grating-1024x497.jpg - 33kB$