Sunlight is not white
sublime tool
Hello.
I was looking into what's the "official" colour of sunlight. Is it yellow, is it white?
As it is, the sunlight that touches the surface of objects on earth can be slightly yellow or warm, depending on the time of day and the altitude of your scene.
This can be found through observation, but I was looking for a scientific reason why that's so.
- - - - -
The light beams that come from the sun are white. They are not perfectly white, but they're so close to white that we can consider them as such.
White light is the combination of several light colours (see Additive Colour).
The sun is yellow in this picture just for illustration, it does look white when seen from space.
When a beam of sunlight enters the atmosphere, that beam is separated in a particular way because of the particles in the atmosphere.
The atmosphere scatters ("distributes") the colder \ bluer hues of light in many directions. This is why the sky appears blue and the general exterior ambient light has a hint of blue. It's a diffuse light, it comes from all directions.
The warmer hues of light are almost not scattered at all, so they follow a very similar direction to what they originally had when they left the sun.
So you have the colder hues being scattered away from the sunlight beam to some extent, and this makes the beam look warmer. The final warmth of a direct sunlight beam depends on how much atmosphere it goes through.
The effect is more pronounced the "deeper" that a beam of light enters the atmosphere.
A light beam with a grazing angle, something that happens in a sunrise or sunset, goes through a greater distance in the atmosphere. It hits more particles, so more blue hues are taken away from it. This makes the beams arrive warmer than beams that happen at noon, for example.
At noon, a beam of light is coming more vertically and so it goes through less atmosphere. Less blue hues are taken away from it and it looks colder \ whiter in comparison.
TL;DR:
When you're going to paint or render a scene, the highest that the sun is in the sky (that is, the closer that you are to noon), the whiter the sunlight will be.
When you approach sunrise \ sunset, the sunlight is much more filtered by the atmosphere and the whole environment is tinted towards warmer hues, progressing from white to yellow to orange to red.
Altitude also influences this. When you're up in the sky, the sunlight that reaches you hasn't been filtered as much as when you're down on the ground, so the sunlight in high altitudes feels whiter -- more like it originally was when it left the sun.
This is explained in more detail here: http://www.physlink.com/Education/AskExperts/ae665.cfm
References:
- http://en.wikipedia.org/wiki/Rayleigh_scattering#Reason_for_the_blue_color_of_the_sky
- http://solar-center.stanford.edu/SID/activities/GreenSun.html
- http://scienceblogs.com/startswithabang/2010/10/14/why-is-the-sun-yellow/
- http://earthobservatory.nasa.gov/Features/EnergyBalance/page2.php
I was looking into what's the "official" colour of sunlight. Is it yellow, is it white?
As it is, the sunlight that touches the surface of objects on earth can be slightly yellow or warm, depending on the time of day and the altitude of your scene.
This can be found through observation, but I was looking for a scientific reason why that's so.
- - - - -
The light beams that come from the sun are white. They are not perfectly white, but they're so close to white that we can consider them as such.
White light is the combination of several light colours (see Additive Colour).
The sun is yellow in this picture just for illustration, it does look white when seen from space.
When a beam of sunlight enters the atmosphere, that beam is separated in a particular way because of the particles in the atmosphere.
The atmosphere scatters ("distributes") the colder \ bluer hues of light in many directions. This is why the sky appears blue and the general exterior ambient light has a hint of blue. It's a diffuse light, it comes from all directions.
The warmer hues of light are almost not scattered at all, so they follow a very similar direction to what they originally had when they left the sun.
So you have the colder hues being scattered away from the sunlight beam to some extent, and this makes the beam look warmer. The final warmth of a direct sunlight beam depends on how much atmosphere it goes through.
The effect is more pronounced the "deeper" that a beam of light enters the atmosphere.
A light beam with a grazing angle, something that happens in a sunrise or sunset, goes through a greater distance in the atmosphere. It hits more particles, so more blue hues are taken away from it. This makes the beams arrive warmer than beams that happen at noon, for example.
At noon, a beam of light is coming more vertically and so it goes through less atmosphere. Less blue hues are taken away from it and it looks colder \ whiter in comparison.
TL;DR:
When you're going to paint or render a scene, the highest that the sun is in the sky (that is, the closer that you are to noon), the whiter the sunlight will be.
When you approach sunrise \ sunset, the sunlight is much more filtered by the atmosphere and the whole environment is tinted towards warmer hues, progressing from white to yellow to orange to red.
Altitude also influences this. When you're up in the sky, the sunlight that reaches you hasn't been filtered as much as when you're down on the ground, so the sunlight in high altitudes feels whiter -- more like it originally was when it left the sun.
This is explained in more detail here: http://www.physlink.com/Education/AskExperts/ae665.cfm
References:
- http://en.wikipedia.org/wiki/Rayleigh_scattering#Reason_for_the_blue_color_of_the_sky
- http://solar-center.stanford.edu/SID/activities/GreenSun.html
- http://scienceblogs.com/startswithabang/2010/10/14/why-is-the-sun-yellow/
- http://earthobservatory.nasa.gov/Features/EnergyBalance/page2.php
Replies
Very cool info, thanks for sharing!
http://solar-center.stanford.edu/SID/activities/GreenSun.html
I would also argue that the effect is negligible when it comes to rendering because while blue light is diffused, it essentially adds to the ambient component, so what is received by a spectrograph pointing directly at the sun will not be the same as all the mathematical components that add up to the light that is illuminating a surface.
We also have to remember that our brains have an internal white balancing system. Thats why in a room illuminated by an incandescent light bulb, things appear white, but on cameras the truth comes out. (when the camera isn't using auto white balance anyway). Which means that our brain will attempt to treat the dominant light in the scene as white.
And lastly because blue light is scattered it means your ambient component will have a larger amount of blue in it, meaning that the shadows are cooler in comparison to the lit components, which exaggerates the warmth of the light.
I've heard that before.
I don't know if my brain is broken or something, but incandescent bulbs give everything an ugly yellow hue as I see it. I've never in my life noticed this effect.
http://www.moillusions.com/black-and-white-spanish-castle-in
Doing a bit more reading on it just now, turns out that while the brain has part of the blame, most of it is in the retina. the cones in our eye apparently become less sensitive to the colours it is detecting, and so it attempts to balance it out, which is why this optical illusion works.
So unless it doesn't work for you, no your brain is not broken.
"The iris (4-6-fold range change) and fast rod/cone adaptation (80-fold) occurs in a few seconds, but the slow rod/cone adaptations (another 40 fold) take o.t.o.o. 10 minutes. Iris control is CNS mediated, of course, and the other mechanisms are reserved for the photoreceptive cells. The slow mechanism is responsible for the slow onset of vision when you go from a well lit area to a rather dark area."
so I don't think it contradicts the OP at all
Everybody already knows that the sun turns orange through red at sunset.
However, that statement can be confusing when it's not accompanied with an image of the sun's radiation curve:
Taken from here.
It's known that the peak of the sun's radiation is approximately at the 0.5 nm wavelength mark.
A wavelength of 0.5 nm corresponds to a greenish colour -- so the strongest colour emitted by the sun is green indeed.
But by no means is that the only colour emitted by the sun, or else it would look green in space (and we've seen that it looks white).
Blue is one of the primary colours of the light \ additive colour model, and once it's scattered away from the direct sunlight beam, what remains is the mix of warmer hues that tends to ("goes towards") yellow, as written. I hope it's clearer now.
The warmth in direct sunlight is also made stronger when we compare it with shadows: shadows carry only the diffuse light from the blue sky and other faint light bounced by surfaces.
Wikipedia has an article on sunlight with an interesting graphic, described as the "spectrum of the visible sunlight at approximately sea level:"
According to this graphic, direct sunlight should feel warmer than the diffuse light from the blue sky, so your daylight scene has warm highlights and cold shadows\ambient term. The time of day will influence this difference, as explained in the first post.
There is no objectively correct white balance, our cones and rods are not a uniform sensitivity, and even if they were, as the wave length changes, the energy of the photons change.
So when we are working out how to treat our images that means a lot of it comes down to viewer discretion.
In fact, the object that surround the subject are going to affect the light quality far more than any perceived amount of the sun being yellow. This is due to light bouncing off objects.
In fact if you want to get philosophical about it. A white piece of paper we hold up in sunlight is probably the most white thing you will ever see(unless you personally go to space and see it with your own two eyes). Sunlight is what we base all of our white calculations off, when we compare light bulbs, the bench mark is to get as close to sunlight as possible.
So while we can analyze a light for it's core components, i would say that we aren't testing what temperature sunlight is, but more working out what our ideal white is.
I hope that makes some sort of sense.
Taken from these references:
- http://www.physlink.com/Education/AskExperts/ae665.cfm
- http://earthobservatory.nasa.gov/Features/EnergyBalance/page2.php
http://gurneyjourney.blogspot.com.br/2011/09/part-3-gamut-masking-method.html