I've been studying shadows recently, and I've noticed something "contradictory"...

sk66

Advertiser
Messages
8,650
Name
Steven
Edit My Images
Yes
Due to an assignment I'm mentoring/critiquing I've been paying a lot more attention to shadows in general (shadows are the theme). And it seems to me that cloudy skies do NOT make shadows softer. Yesterday was a particularly good day to observe this as we had a low cloud layer which should be better than a high layer. It wasn't too thick and it was somewhat inconsistent. So I could watch a shadow as it was being cast from direct sun thru all of the transitions to no shadow at all.

In my observations the only thing clouds really do is make shadows less dense/dark and they have little/no effect on "softness." In essence, they really only create "flat" light when all shadows are gone.

There *is* a very brief moment where the shadows do seem to become notably softer, but I believe it is due to reflected light coming from surrounding surfaces and not because of the clouds themselves. This particular balance of sunlight diffused by clouds and reflected light is actually a very hard thing to find...especially with any consistency to it. A much more significant factor in softness seems to be how far the shadow has to travel.

So, am I nuts? Or is the "clouds produce soft light" really a misnomer/confusion of density -vs- softness characteristics?
 
Do you mean that there were some clouds about and that the sun was occasionally obscured?
Or do you mean that it was an overcast day, where the clouds obscured all of the sky?

I don't remember ever hearing or reading anyone claim that a few clouds in the sky and the sun behind one of them, or merely a thin layer of cloud through which you can see the sun, would cause soft shadows.
What I have often seen claimed, and proved for myself, is that an overcast sky (completely obscured) provides soft shadows.

If you can see the sun through a thin layer of clouds or single cloud then the shadow edges from the sun will not be much softened (as from a poorly designed softbox where there is a very clear hotspot, say). If the sun is completely hidden behind a cloud the shadow edges will be softened somewhat (I tried this this morning).
If you cannot see the sun through clouds that completely cover the sky (an overcast day) then the shadows will be soft (as from a big softbox with good and effective diffusion material).
 
Last edited:
The other day it was fairly heavy overcast with a few breaks in the clouds.

This morning it was a completely blue sky except for one slow moving cloud. As the cloud blocked the sun the shadows became quite "soft." As it was beginning/ending the shadows were "hard." The softness could not have been due to the "diffusion" of cloud cover. It must only be due to a reduction in the ratio of direct sunlight vs ambient/reflected light. Another way of looking at it is "if you cannot see the sun, where is the light coming from?"

I think the effect must be similar to using a small light source to light a very large silk from a short distance... If pointed directly at the silk (as the sun is during mid-day) there is little spread/effect other than a loss of intensity. But if pointed at an angle (as the sun is early/late) then the effect is a gradient. But in both cases the effect of the diffusion is "minimal" due to the size of the light source vs the size of the modifier. Most of the effect is due to a reduction in light ratio.
 
Last edited:
The sun is a constant. The only variable is the amount of cloud. If there is direct sun but there are still clouds in the sky, those clouds will add a bit of diffused light. If the sun is covered by cloud it will all be diffused.


Steve.
 
I don't really think so...
In the analogy above the clear blue sky above the clouds is the same as having an illuminated white wall behind the small light source (and it would be illuminated by the small light source bouncing off of the silk). Adding diffusion to it doesn't make much difference.

Another way of looking at it is to look at the shadow cast by a tall tree. The shadow from the lowest branches nearest the base will be harder and darker. The shadow from the tallest part of the tree and farther away will be softer and lighter. The difference is not the diffusion or it's size. The difference is in the amount of reflected/indirect light reaching the area (which is now your light source if you are in the shade). I believe that is really what is happening with clouds.

It is (I think) pretty well known that if you want to use shade you want to use the edge of soft shade (the tall part). That's because it will allow you to have a lighting ratio (i.e. not "flat") and the light will be "soft" because the sources are indirect/reflected.
 
This isn't complicated.
There are always shadows, because a shadow is an area that is deprived of light by the object that is between the shadow and the source of the light, that's a given. The variables are the intensity and direction of the shadow, and how sharply defined (or otherwise) its edges are.
Shadows are hard-edged and obvious when the light source is distant and small, relative to the size of the subject and relative to the distance that the shadow has to travel from the subject to wherever the shadow hits, and relative to the distance that the light has to travel to the subject.
This is all very easily tested in a studio, just using the modelling lamp for visual effect. Try the lamp at various distances from light to subject, with the subject at various distances from where the shadow falls, and with the light at different sizes (e.g. with and without different sized modifiers).
A light source that we would normally consider to be small, for example a standard reflector, that will cast a harsh shadow with clearly defined edges if you point it at someon'e head from 20' away, will produce an incredibly soft shadow if you point it at a pea from 2" away.

Outdoor shadows behave in exactly the same way. A cloud is just a type of diffuser, obviously some clouds will be more efficient diffusers than others. It isn't the cloud that produces soft light, it's whatever stops the sun from being visible and which disperses its light over a large area (which doesn't have to be a pretty cloud) and a totally grey sky, which totally blocks the sun, invevitably creates an incredibly soft light simply because it totally obscures the sun and makes the entire sky the light source..
So, on a foggy day for example, the whole sky becomes the light source, and because the visibility will be poor, the light source will be close too.

It's as simple as that.
 
Last edited:
Garry, I've thought the same thing for decades... it's what we've always been told and it makes sense if you correlate it to using a softbox with good ambient/spill control. But it really doesn't seem to work that way.

First, light doesn't bend, and a point of light doesn't spread. So if you block the sun with a solitary cloud what is causing the soft shadow? It's reflected light and not "the cloud."
I believe the analogy holds; at best a cloud layer is akin to lighting a silk with a small bare head from a relatively short distance... and we both know that doesn't work well at all. You can make the silk as thick as you want, but that won't really do anything but cut the light.

Obviously, there's a huge amount of variability. You can have multiple cloud layers which should work much better. And clouds can have various levels of moisture/water drops which will refract/redirect the light more/less (your foggy day). So yeah, there is *the potential* for clouds to create soft light. But it's really not as simple as that. And it's really probably not something anyone should depend on.

The main thing I'm seeing is that you don't *really* want clouds or shade. What you really want is a place where the light is coming from many directions and more evenly due to reflection (i.e. bounced fill). That will probably include using shade or cloud cover, but it's not the key factor.

I *am* differentiating between "soft light" where this a lighting ratio/shadows, and "flat light" where there are no shadows. If you want flat light, then heavier cloud cover and deeper shade is an answer. Sometimes that's better than hard light, but it's seldom what you really want. Now, if you combine that flat light with supplemental lighting, then you've got something to work with.
 
Last edited:
No different to putting a bit of tissue paper over your flash to diffuse it.

The shadows will be most distinct when the only light is directly from the sun. The sun is far enough away that it can be considered to be a point source.

As you add clouds, you add light coming from a larger area so that area of shadow which was created by something masking the sun's light now has a bit of light falling on it from other angles. At the same time, any cloud in front of the sun is reducing its intensity. Both of these make the shadows less didtinct.

More cloud = more of this effect. When there is enough cloud, just about all of the light is diffused through clouds and no direct light is eveident. At that point, there are no distinct shadows.


Steve.
 
Last edited:
No different to putting a bit of tissue paper over your flash to diffuse it.
Which is about pointless...
The shadows will be most distinct when the only light is directly from the sun. The sun is far enough away that it can be considered to be a point source.
It can only be considered a point light source in terms of the inverse square law.
As you add clouds, you add light coming from a larger area so that area of shadow which was created by something masking the sun's light now has a bit of light falling on it from other angles. At the same time, any cloud in front of the sun is reducing its intensity. Both of these make the shadows less didtinct.
If you add an even cloud layer, you haven't changed anything really... the lighting ratio from the various areas of the sky remain ~ the same. It requires a cloud that is darker to be blocking the sun... it can even be a solitary cloud.
More cloud = more of this effect. When there is enough cloud, just about all of the light is diffused through clouds and no direct light is eveident. At that point, there are no distinct shadows.
I think you are mixing light ratio with light "quality." I am trying to separate them.
 
Last edited:
It can only be considered a point light source in terms of the inverse square law.

I was thinking more in terms of the shadow having sharp, defined edges.

If you add an even cloud layer, you haven't changed anything really... the lighting ratio from the various areas of the sky remain ~ the same. It requires a cloud that is darker to be blocking the sun... it can even be a solitary cloud.

With no cloud, you just have the direct light from the sun. With cloud, you have diffused light from a much larger area and from all directions.

I think you are mixing light ratio with light "quality." I am trying to separate them.

I'm not sure if you can.


Steve.
 
Last edited:
What about a quick thought experiment where the sky is so overcast that the position of the sun is impossible to see.

Add an imaginary tree.

There will effectively be no shadow.

Now imagine the same scene with a clear sky, bright sunlight.

There will be a well defined shadow.

In between these two extremes what definition do the shadows have?
 
I'm confused.

Same building, shots taken not long apart, but with different cloud levels
Yes, you can get those extremes and everything in between...
And you can also get everything in between with just one cloud, the overcast and making the sun "larger/diffused" has little to nothing to do with it.

Part of the issue/confusion comes from the idea that "the sun" is the light source as if it is just a point source. But in fact, the entire sky (as illuminated by the sun) is the light source. It just has a really hot "hot spot." Adding diffusion to the sky ( a cloud layer) doesn't make it any larger, it doesn't make the hot spot any less hot proportionally, and it doesn't really make the hotspot (sun) any larger (you can see this yourself in any number of ways).
A single cloud serves to neutralize the hot spot. What a cloud layer does is much more involved than just "diffusion," and it is way different than "a softbox." A lot of it (more than "diffusion") has to do with the way light behaves as it travels thru our atmosphere, and why clear sky/shadows/shade/cloudy days are all "blue."
 
Yes, you can get those extremes and everything in between...
And you can also get everything in between with just one cloud, the overcast and making the sun "larger/diffused" has little to nothing to do with it.

Part of the issue/confusion comes from the idea that "the sun" is the light source as if it is just a point source. But in fact, the entire sky (as illuminated by the sun) is the light source. It just has a really hot "hot spot." Adding diffusion to the sky ( a cloud layer) doesn't make it any larger, it doesn't make the hot spot any less hot proportionally, and it doesn't really make the hotspot (sun) any larger (you can see this yourself in any number of ways).
A single cloud serves to neutralize the hot spot. What a cloud layer does is much more involved than just "diffusion," and it is way different than "a softbox." A lot of it (more than "diffusion") has to do with the way light behaves as it travels thru our atmosphere, and why clear sky/shadows/shade/cloudy days are all "blue."

Perhaps I a misunderstanding you, but a completely covered, overcast sky where you are unable to discern the location of the sun totally removes the hotspot.

Can you post some of your test shots?
 
The larger the light source, the softer the shadows. Effective size of the light source is relative to distance (from the subject), and proportional to the size of the subject. That's it.

The sun/sky/clouds etc follow the same rules, it's just that it's an infinitely variable light source, or multiple light sources of infinite variety, with even more variety added by reflective surroundings. I don't see any mystery here.
 
The sun/sky/clouds etc follow the same rules, it's just that it's an infinitely variable light source, or multiple light sources of infinite variety, with even more variety added by reflective surroundings. I don't see any mystery here.

During daylight hours it's really a fixed intensity light source with infinitely variable diffusion from the clouds - but the outcome is the same.


Steve.
 
During daylight hours it's really a fixed intensity light source with infinitely variable diffusion from the clouds - but the outcome is the same.

Steve.

Yes, the sun is fixed intensity and clouds etc obviously change things dramatically. But another set of variables is the height above the horizon and the influence that has on shadows, colour and brightness. The earth's atmosphere is another light source and there is plenty of brightness around even when the sun has dipped below the horizon. The earth's atmosphere is what we see as blue sky; beyond that, the sky is black.
 
Perhaps I a misunderstanding you, but a completely covered, overcast sky where you are unable to discern the location of the sun totally removes the hotspot.
Yes, but assuming the cloud cover is relatively uniform it is equally cutting the light from all other areas of the sky as well. Therefore, the hotspot remains relatively "hot" in proportion to the rest of the sky... the clouds have just killed all of it. And now you have "flat" lighting (not "soft" lighting).
 
Yes, but assuming the cloud cover is relatively uniform it is equally cutting the light from all other areas of the sky as well. Therefore, the hotspot remains relatively "hot" in proportion to the rest of the sky... the clouds have just killed all of it. And now you have "flat" lighting (not "soft" lighting).

Perhaps we only have these skies in wales... But you can get a cloud cover under which it is impossible to discern the position of the sun with a uniform brightness in all directions.

PS, what do you see as the difference between flat lighting and soft lighting?
 
The issue is how light behaves as it passes thru the air. Shorter wavelength spectrums (violet/blue/even green) are scattered as light travels thru the air. So all of the sunlight that is *not* landing on you directly *is* providing short wavelength illumination. This scattering is why the sky is blue and the sun is yellow... The scattered light is not "lost" and it can provide a significant amount of illumination. This is why snow tends to record as blue, not because it is reflecting the sky like a mirror, but because it is being illuminated by the blue light. But, short wavelengths continue to scatter and have a harder time traveling longer distances. This is why the sky graduates to a lighter blue towards the horizon, it's also why the sky can even turn green and the sun turn red at sunset.

These shorter wavelengths are going in every direction and coming from all areas. And what clouds primarily do is cause (more of) the sunlight to refract/scatter to the shorter wavelengths. This is why clouds, shade, shadows, cloudy days all record as blue. It is also why there can still be plenty of light before the sun has risen above the horizon and when it is completely obscured by clouds. And it is this short wavelength component that actually provides the "soft light." The other thing a cloud layer does (to a much lesser degree) is re-reflect the light that bounces off of the surface.
This is completely different from what happens when you use a light in a softbox... in numerous ways.
 
Last edited:
Perhaps we only have these skies in wales... But you can get a cloud cover under which it is impossible to discern the position of the sun with a uniform brightness in all directions.

PS, what do you see as the difference between flat lighting and soft lighting?
"Flat" lighting is completely even... it can be from hard direct light filling all shadows, or soft indirect light causing no shadows. "Soft" light has shadows, but the shadows have a gradation to the edges. Another way of thinking of it would be in terms of "sharp" and "dull" edged shadows.
 
Last edited:
Unless you are talking about gravitational lensing or peculiar laser things, light does not bend. It reflects/refracts, but that is *different.*

I'm talking about diffraction, light bends as it passes the edges of objects. While you're correct, it's not strictly bending, it does fan out in a curve from that point.
 
"flat" "soft" "hard" lighting terms are in fact all subjective and have no meaning that a physicist would understand, they are just different words used to describe our opinions of the shadow intensity and shadow transfer edges.

No, light doesn't bend, although like just about everything else in life, it can change direction when it hits a solid object and can also change direction when it hits an object that transmits some of it, such as glass, water etc.

But this is all just over-complicating what is really a non-issue. The reply below sums up the situation.
The larger the light source, the softer the shadows. Effective size of the light source is relative to distance (from the subject), and proportional to the size of the subject. That's it.

The sun/sky/clouds etc follow the same rules, it's just that it's an infinitely variable light source, or multiple light sources of infinite variety, with even more variety added by reflective surroundings. I don't see any mystery here.
 
But this is all just over-complicating what is really a non-issue.
Perhaps...
I was simply trying to resolve my observations that a diffused cloud layer does not necessarily result in soft light, when "in theory" it should. And why "soft light" can be created by a single small cloud obscuring the sun, when "in theory" it shouldn't.

Does it "change things?" Not if you look at it from the "multiple light sources of infinite variety, with even more variety added by reflective surroundings" perspective... but that is significantly different from simply "diffusion."
 
It might also be relevant to point out that the basic concept of "diffusion makes your light source larger" is also erroneous and misleading. Discounting small gains from highly reflective interiors, the diffused light is only as large as the originating spread of light is to begin with.
What diffusion actually does is break the light into "a larger collection of light sources." This small difference can make a huge difference in understanding/application...
 
Last edited:
It might also be relevant to point out that the basic concept of "diffusion makes your light source larger" is also erroneous and misleading. Discounting small gains from highly reflective interiors, the diffused light is only as large as the originating spread of light is to begin with.
What diffusion actually does is break the light into "a larger collection of light sources." This small difference can make a huge difference in understanding/application...

The confusing use of the word 'diffusion' is often exploited by manufacturers, leading people that purchase say a Stofen to believe that it's magically giving them softer light. It's not - it's just 'redirecting' the light all around so that it bounces off the ceiling/walls/surroundings. They then become a large source of soft light, but the component of light coming directly from the Stofen itself is still as hard as you get from a naked flash gun (because it's the same size).

This is one of the most common misunderstandings.
 
This is one of the most common misunderstandings.
And leads to the (to me) surprising popularity, and almost comical misuse of such devices... (the "light sphere" being the most prevalent it seems).
But it does also reduce the ratio of direct/hard light to some extent...
Which then leads to the (pretty dumb) idea/recommendations of putting tape or some other small form of diffusion over the built in flash.... you're better off just reducing the power.
 
Last edited:
I'm talking about diffraction, light bends as it passes the edges of objects. While you're correct, it's not strictly bending, it does fan out in a curve from that point.
Yes, but from a photography perspective, the light has to have already fallen on your subject (from behind) in order to see it.
 
Last edited:
Due to an assignment I'm mentoring/critiquing I've been paying a lot more attention to shadows in general (shadows are the theme). And it seems to me that cloudy skies do NOT make shadows softer. Yesterday was a particularly good day to observe this as we had a low cloud layer which should be better than a high layer. It wasn't too thick and it was somewhat inconsistent. So I could watch a shadow as it was being cast from direct sun thru all of the transitions to no shadow at all.

In my observations the only thing clouds really do is make shadows less dense/dark and they have little/no effect on "softness." In essence, they really only create "flat" light when all shadows are gone.

There *is* a very brief moment where the shadows do seem to become notably softer, but I believe it is due to reflected light coming from surrounding surfaces and not because of the clouds themselves. This particular balance of sunlight diffused by clouds and reflected light is actually a very hard thing to find...especially with any consistency to it. A much more significant factor in softness seems to be how far the shadow has to travel.

So, am I nuts? Or is the "clouds produce soft light" really a misnomer/confusion of density -vs- softness characteristics?

"flat" "soft" "hard" lighting terms are in fact all subjective and have no meaning that a physicist would understand, they are just different words used to describe our opinions of the shadow intensity and shadow transfer edges.

+1. I had a minor revelation a while ago when I realised that applying these terms to light sources makes much less sense than applying them to the shadows.

The clouds function as a diffuser. They cause light to strike the shadow region behind the subject from many different directions, thus causing the shadows to be less dense.

A softbox softens shadow edges because not all areas of the shadow zone are visible to all of the softbox; different regions of the shadow zone have differing amounts of fill. (Distances from the different parts of the softbox plays a part here, too). Thus some parts of the shadow region are made less dense than others, so giving them a soft edge.

Whether clouds have the same effect on your subject will depend on the size and positioning of the subject. It's one reason I sometimes position a model by a wall when photographing them under overcast skies.

Sorry if that only adds to the confusion.. I thought twice about posting in the presence of acknowledged experts. It helped me, though.
 
I think most more advanced/experienced *do* think of hard/soft in terms of the shadows... that's what it's actually referring to. The confusion arises in the correlation to the primary cause of it. Then you add in the term "flat" (a lack of shadows/definition) which can be caused by either form of lighting, but many just think of as being a characteristic of extremely soft lighting... as if it's "good."

I still have trouble w/ the correlation of clouds to softbox as there are a few other differences that haven't been mentioned.
One is that a lot of the longer wavelength/lower energy light is reflected off by the clouds and is largely lost... this *could* be correlated to a highly inefficient shoot thru umbrella I guess.

Another is that the size and distance(height) of the cloud(s) appears to be irrelevant. One might think that a cloud layer has already made such a huge change in "the size of the sun" that a difference of a dozen miles or so is minor. But when doing advanced lighting of small objects using very large modifiers from very short distances, you will find that even differences of a foot or less can make a significant impact. I would think a dozen or more miles would matter...

And then there is the fact that a cloud/cloud layer doesn't *seem to conform to the inverse square law... and it should (* due to the "infinite variability" of the situation(s) I can't put empirical numbers/measurements to it).

But, in the end it really doesn't matter much I guess... Light is light and you have what you have. It's what you do with it that matters.
 
Last edited:
I think that there's a real danger of over-thinking things here.
I sort of agree that the height of a cloud *should* make some kind of difference, but I have no way of accurately estimating (or even guessing) at the height so it isn't something that I'm going to concern myself with. And as the height isn't within my own control there would be no point in wasting my small number of brain cells on it even if I did know.

In the studio, it's different. Using modifiers at very close distances, a foot is an enormous distance and makes an enormous difference - hell, an inch can make a big difference, but this does matter because we are in control of that distance and relative size. But, it's all about ratio. If we move a modifier from say 24" distance to 12" distance then we are quadrupling its effective size, if the light from the sun travels 93,000,000 miles before hitting some kind of obstruction it can't matter that much whether it is travelling 10 or 12 miles more or less.

Over-thinking is one of my faults. My sport is clay pigeon shooting and I am capable of estimating distance extremely accurately. I have a set of figures in my head that tells me exactly how much the shot pattern will spread over the distance until it reaches the clay, how fast the clay is moving, how quickly it is dropping or climbing and how much the shot will fall over that distance. Therefore, I should stand a much better change of making contact with the clay than someone who doesn't have that knowledge - but does that help me? No, or at least it doesn't guarantee success. Some of the very best shots are barely literate, or at least they know none of these things, they just concentrate on the job in hand and that means that they get their shot off quicker than me, which really does help:)

Same with lighting. We need to have a basic understanding of its importance and how to use it to good effect, but too much theoretical knowledge can interfere with the creative aspects of the shoot. Knowledge is important, but skill trumps knowledge.
 
Back
Top