Why sun always white

[eengelmann]

Does someone has an idea why the sun appears during sunsets always white or at least much lighter then the surrounding, even when it clearly appears red to the eye? Only exception when it is already almost completely under the horizon.
I thought it was a question of dynamic range, but I have doubts it is that, because the sun is the only part which behaves that way.
An example in the following image. I can guarantee the sun looked absolutely red!
I have even tried to fix my sight on the sun, to avoid the apparent increase of the dynamic range of the eye through adaptation to the different parts of the framing, but the sun always appears in the pictures much lighter.



Here, in the contrary, the sun appears with the correct color:

[Shadowman]

Is it really white or are your eyes just perceiving it as white? If you look at the RGB values for the brightest portion of your image; in this case the sun, do the values measure as white?

[xpatUSA]

Does someone has an idea why the sun appears during sunsets always white or at least much lighter then the surrounding, even when it clearly appears red to the eye?

As John hints, in the first image, all the channels are blown in the sun. When all channels are blown, the viewer can only show white:




The mention of dynamic range is relevant because, if the exposure was made for the sun, then the sea and the clouds would be too dark.

We get this all the time with outside shots especially with bright blue skies or side-lit clouds.

[george013]

Does someone has an idea why the sun appears during sunsets always white or at least much lighter then the surrounding, even when it clearly appears red to the eye? Only exception when it is already almost completely under the horizon.
I thought it was a question of dynamic range, but I have doubts it is that, because the sun is the only part which behaves that way.
An example in the following image. I can guarantee the sun looked absolutely red!
I have even tried to fix my sight on the sun, to avoid the apparent increase of the dynamic range of the eye through adaptation to the different parts of the framing, but the sun always appears in the pictures much lighter.



Here, in the contrary, the sun appears with the correct color:

Different metering values. In the first image your sun is clipping. Value 255,255,255, average 255. In the second image your sun is somewhere halfway the histogram, value 255,100,71, average 140.

There's no exif in your photo. But analyze your metering settings.

George

[NorthernFocus]

It is indeed a dynamic range issue. Your eyes/mind may not perceive it as such at the time, but the camera sensor simply records the light values at that moment in time. Next time you take such a shot try shooting multiple frames at different exposures. Later on the computer you can compare/contrast and determine how you want to shoot in future based on whether the sky or foreground are most important to your desired image.

[george013]

Beside the dynamic range and what you want to meter, there's another variable: the white balance.
And the posibility you don't know exactly anymore what you saw. That's normal.
Sun set colors are not the same every time. The lower the sun gets, the redder it can become. Google on "sunset red" and you will find more explanations.
Comparing your pictures with others, I think the wb is set to warm.

George

[Manfred M]

I totally agree about the dynamic range being the key issue. At that time of day there is simply no way of capturing the sun and the rest of the scene in a single shot. I tend to not shoot directly into the sun anyway, but if I did I would expose the scene the way I wanted and then (at the same focal length) grab a shot where I can capture the sun and then do a "poor man's HDR) and just clone the sun into the shot that was taken for the scene.

As for George's point on WB, I also agree as the colour temperature changes rapidly as the direct sunlight cuts through that much atmosphere. I seem to remember that at sunset, the colour temperature is around 2900K. That is something my camera's AWB cannot get down to and the colour of the light will be quite warm. In the meantime, much of the sky (other than any clouds) will have gone quite blue (well over 10000K) so you are shooting in a mixed light situation (the blue (and violet) light is refracted at 90 degrees, so that is why the direct sunlight looks red because the blue / violet component of the light has been redirected), and you can do whatever you want with the colours because you have such a wide colour temperature range to work with.

[Dave Humphries]

Hi Enrico,

Yes, this scene has a vast dynamic range.
You cannot capture it all; i.e. retain detail in sea, sky and expose properly for the sun.

However, the main reason the sun is white in your first image is because the meter has probably led you to over expose for it (as distinct from the scene as a whole).

When the sun is too far above the horizon, it may not be possible to actually avoid blowing all channels, even at 100 iso, a very narrow aperture and your fastest shutter speed. If it looks red to the naked eye (and you can view without damaging your eyesight), then it should be possible, but you may not get the clouds as you'd wish, which is where Manfred's suggestion comes in to play.

Cheers, Dave

[bje07]

No-one speaks about bracketing, which is a solution to enlarge dynamic range, .... or get sun at right exposure (on one shot) and use multiple layers to change the wrong white dot over exposed.

[Shadowman]

No-one speaks about bracketing, which is a solution to enlarge dynamic range, .... or get sun at right exposure (on one shot) and use multiple layers to change the wrong white dot over exposed.

Or using a camera with the widest dynamic range; however there are limits of course. Also, no mention of ND filters.

[george013]

Or using a camera with the widest dynamic range; however there are limits of course. Also, no mention of ND filters.

A nd filter doesn't change the dynamic range. It just causes a lower ev.


But the question was how to explain the difference in colors.

George

[dem]

Luminance is used to characterize emission or reflection from flat, diffuse surfaces. The luminance indicates how much luminous power will be detected by an eye looking at the surface from a particular angle of view. Luminance is thus an indicator of how bright the surface will appear.

The sun disc at the horizon is about 300 times brighter than the average cloudy sky. The sun disc at noon is about 1,000,000 times brighter than the sky:

https://en.m.wikipedia.org/wiki/Orde...de_(luminance)

[IzzieK]

Good stuff, Ted!

As John hints, in the first image, all the channels are blown in the sun. When all channels are blown, the viewer can only show white:




The mention of dynamic range is relevant because, if the exposure was made for the sun, then the sea and the clouds would be too dark.

We get this all the time with outside shots especially with bright blue skies or side-lit clouds.

[Pippan]

Hi Enrico, I have little experience in photography compared with those who've already commented but I have a suggestion that may be of practical use to you. Yes your white sun is totally blown out because the dynamic range is greater than your camera (or, I believe, any camera) can handle while producing a reasonable exposure for the sky. The red sun on the horizon is not blown out because even just that little bit lower, the sun's light is passing through much more atmosphere, which absorbs more of the sun's light, reducing the difference in brightness between the light from the sun and the light from the surrounding sky, thus reducing the dynamic range of the scene to something your camera can handle. Also, I don't believe your white balance is too warm because your sunset photos are so much like sunsets I see nearly every day from Nightcliff, in Darwin, Australia (except tonight, unfortunately, the night of the supermoon, sadly very overcast). In fact I took very similar photos to yours just a few evenings ago and have attached them. They were taken just a few minutes apart.

This is what I do in this situation. I use a raw converter called Photo Ninja, which is very good at 'recovering' blown highlights. I'm sure other programs can do this too, I'm just not familiar with them. Photo Ninja does not really 'recover' blown highlights but samples colour from surrounding areas and fills in the blown highlights with the colour. It doesn't always get it right but it usually does and it looks better than white. Sometimes, especially with blown suns and fire, the colour is too saturated and looks a bit unreal. However, you can easily adjust the level of saturation with a slider from 0-100%, so I just play with it until I think it looks OK. In the first picture below, the part of the sun above the cloud was as blown and white as your sun but I think this treatment (backed off to 55% 'highlight recovery') makes it look acceptable. In the second picture the darker orange part of the sun on the horizon was not blown, but the lighter area in the top half was blown, and in this case I left the highlight recovery saturation at 100%. It looks OK but I'd have liked it a bit darker orange. Probably some Photoshop magic would do the trick but that's way beyond my skill level at present. Anyway, something like this may be an easy way to improve your sunsets. I'm assuming you shoot raw as I don't know if jpegs have this much flexibility. BTW, these are the first images I've ever uploaded to a forum so I hope it works.




On white balance, I never shoot in AWB. I usually set a custom white balance or light profile from midday sun and use that later for sunsets. As someone above said, white balance changes rapidly around sunset. However, if you set a 'correct' white balance for that orange light, everything would look like normal daytime--you'd lose all that wonderful colour from the sunset. You don't want that! Anyway, yours are great sunset photos and I hope my suggestion helps.

[eengelmann]

I fear I haven't expressed my question clearly enough!
My question should have maybe been: if it is just a dynamic range problem, why in the case of sunsets (and only in that case) does the eye see something completely different?
In other cases of too large dynamic range situations I can forsee exactly already from what I see what will be too dark/burn out (a lamp/a room with a window in a bright day, two persons dressed one in black one in white, etc...)
But when I shot the first picture (which is already a fusion of two pictures separated by 2 stops) the sun wasn't that bright at all! I could look directly at it without feeling it disturbing for the eyes and seeing it CLEARLY orange or bright red!

In other words, what I see and what I get in the picture is usually perfectly correlated, apart from the fact that the perceived dynamic range of the eys is much larger than that of the sensor.
In the case of sunsets, in the contrary, the correlation is broken!

So I suspect it is not just a dynamic range problem.

[davidedric]

A bit late to the party, but I don't see it as only a dynamic range question.

The sun acts near enough as a black body radiator with a surface temperature of around 5900K. Above the atmosphere, we would perceive that as white. If we were there to take a photograph, it would appear white or grey, but in any event neutral.

Below the atmosphere, scattering takes place from a number of different sources (Rayleigh scattering giving us a blue sky, for example), so the actual colour of sunlight varies with conditions and time of day, as we all know. So yes, dynamic range is important, but so is the colour temperature of the light we are working with.

Dave

[dem]

More on this topic:

http://solar-center.stanford.edu/SID.../GreenSun.html

If we manage to expose for the sun, we will see white colour minus the wavelengths that got scattered in the atmosphere - more blue gets scattered when the sun is lower, so it appears more red.

However, if the sun is not at the horizon, it will be so bright that it is going to blow all three colour channels. There is a specialist sun filter (no the Big stopper will not be enough, not even 10 of them for the midday sun) that looks like kitchen foil but is slightly more expensive...

http://petapixel.com/2013/02/13/phot...-his-backyard/