Can we discuss this video on lighting and answer a question on inverse square law?

[jwbryson1]

This is a good video by Mark Wallace on shooting white backgrounds. Part of the video I am not 100% sure I understand.

The substantive discussion starts around the 2:20 mark but what I have questions about are his comments starting around the 3:10 mark forward.

He indicates that the background is not white because of light falloff. I get that. But then he indicates that some people think moving the light closer to the model will help but in reality it will actually make the background even darker because of the inverse square law. That's what confounds me...

Isn't light falloff caused by (or the "result of") the inverse square law? If yes, then it seems that moving the light closer to the model (and necessarily closer to the background) would illuminate the background because the light has a shorter distance to move and therefore will be more powerful. In other words...before moving the light closer to the model, let's assume that it needs to travel 10' to hit the background. So, it "falls off" for a distance of 10' and is not powerful enough to blow out the background. But, if he moves the light closer to the model and the light only has a distance of 5' to travel (and to "fall off") before hitting the background, why isn't the background exposed more?

Does that fact that the light actually HITS the model have an impact on the inverse square law? Is that what I am missing? Does the light hitting the model further increase light falloff?

Here is the video...

 

[Village Idiot]

The light being blocked by the subject doesn't affect the inverse square law. If the subject was not in the way it would be brighter. Light not hitting the background because of an obstacle doesn't change the fact that moving the light closer to the background will make it brighter.

 

[jwbryson1]

Light not hitting the background because of an obstacle doesn't change the fact that moving the light closer to the background will make it brighter.

This is contrary to what he says in the video and is the exact question I posed....

So, I'm still looking for an answer.

 

[Derrel]

As he states at 3:52, the Inverse Square law is, "totally counter-intuitive." Let me explain a bit. Whjat happens when a light sourece is moved close to the subject is that the light on the subject becomes brighter, yes. However, the rapidness of the falloff of light, over a short distance, is very "steep".

A light placed very close to a subject will fall-off rapidly for the first few feet, and then the rate of falloff decreases. At longer distances, like say 20 to 30 feet, the rate oif falloff is negligible. And for looooooooooong distances, like say from the Sun to Earth, fall-off is a non-factor.

When a light is placed very close to a person's face, like say 2.5 feet away, the highlight side of the face will be bright; the fall-off is sooooo rapid at 2.5 feet that the other side of the face, just a few inches away, will be in deep,heavy shadow. However, if the light is moved back to 10 feet, the lighting on both sides of the face will be...almost equal in intensity.

Now, if the woman in that studio setup stands at the same distance from the background, and he moves the light closer to her, he will need to close down the lens f/stop, from f/5.6, to f/8. And the background will be much darker. The rate of fall-off of light is VERY rapid, the close the light is to the subject. The farter away the light source is, the more "even" the output is over distance. Somewhere I have a link to an example of that.

 

[jwbryson1]

As he states at 3:52, the Inverse Square law is, "totally counter-intuitive." Let me explain a bit. Whjat happens when a light sourece is moved close to the subject is that the light on the subject becomes brighter, yes. However, the rapidness of the falloff of light, over a short distance, is very "steep".

A light placed very close to a subject will fall-off rapidly for the first few feet, and then the rate of falloff decreases. At longer distances, like say 20 to 30 feet, the rate oif falloff is negligible. And for looooooooooong distances, like say from the Sun to Earth, fall-off is a non-factor.

When a light is placed very close to a person's face, like say 2.5 feet away, the highlight side of the face will be bright; the fall-off is sooooo rapid at 2.5 feet that the other side of the face, just a few inches away, will be in deep,heavy shadow. However, if the light is moved back to 10 feet, the lighting on both sides of the face will be...almost equal in intensity.

Now, if the woman in that studio setup stands at the same distance from the background, and he moves the light closer to her, he will need to close down the lens f/stop, from f/5.6, to f/8. And the background will be much darker. The rate of fall-off of light is VERY rapid, the close the light is to the subject. The farter away the light source is, the more "even" the output is over distance. Somewhere I have a link to an example of that.

Derrel, I have seen the videos from Mark on the inverse square law, and I am familiar with what you are talking about. I understand the light fall off will occur very quickly and will decrease the further away it gets from the source.

So, I guess what he hasn't explicitly said in this video is that if he moves the light closer to her, he will probably have to stop down from 5.6 to 8 to keep her in correct exposure. If he does that, and the light falls off quickly the background will be darker.

Makes sense. I get it. Thanks.

 

[amolitor]

Inverse square law just says that when something is 2x further away, there's 4x less light. If it's 3x further away, there's 9x less light.

This is exactly the same thing as aperture stops, interestingly, which may help you get your arms around it. When I move the light back 1.4x, I get 1 stop less light falling on to whatever it is. When I move it 2x back, I get 2 stops less light. When I move it 2.8x back, I get 3 stops less light, and so on.

Let's pretend I have a light aligned with the camera, on axis, to make stuff easy. Let's say it's 10 feet from the model, and 20 feet from the backdrop (so the backdrop is 10 feet behind the model).

I have 2 stops less light falling on the backdrop than the model, right? I have 1/4 as much light falling on the backdrop as the model.

Now move the light closer to the model. Halve the distance to the model. We're now at 5 feet and 15 feet. The backdrop is now 3x further away than the model, NOT 2x further away like it used to be. The lighting ratio is now more radical, I have 1/9 as much light falling on the backdrop as on the model.

What if we move the light back insteadm to 20 feet away from the model? Now it's 20 feet from the model, and 30 feet from the backdrop. There's only about a 1 stop difference between the two, now (1.5x the distance is close enough to 1.4x the distance for government work..).

 

[Derrel]

This is NOT the example I wanted to show, but it's showing the same effect in an illustration, but not in the photographs of the example I've been looking for. Note the illustration and the the text. Rules for Perfect Lighting: Understanding The Inverse-Square Law

"There’s a 75% drop in light from 1 meter to 2 meters, but only a 5% drop in light from 4 meters to 10 meters."

See...those are approximately the kind of distances we find ourselves actually working at in-studio...

 

[TCampbell]

You can think of light falling and the inverse square law in a way similar to how you'd think about the depth of field.

Use depth of field (ignore light for a moment) as an example... it's normal to realize that the closer our lens is to a subject, the narrower the depth of field. The farther we are from the subject (using the same focal length lens and f-stop) the broader the depth of field. If you had a subject standing 5' in front of a background and your 50mm lens was somehow jammed at f/4 and couldn't be changed, but you needed to get a shot with BOTH the subject and background in focus... you wouldn't get closer... you'd get much farther. Because you realize that would naturally increase the depth of field and allow everything to be in focus.

Ok... back to light. Light fall-off works very similar. Think instead of focus, we're worrying about correctly lighting. There's an zone or range of distances from the light in which the light will be "just right" -- not too much, not too little. You can think of this as a "depth of correct light" instead of "depth of acceptable focus".

Each time the distance from the light increases by a factor of 1.4, the amount of light falls off by half (or if decreases by a factor of 1.4 it becomes twice as bright) BTW, we use 1.4 as the "rounded" value but the actual value is the √2 (1.4142...).

If your light is 5' away from the subject and it's providing a perfect exposure, then it ONLY 7' away (5 X 1.4 = 7) the lighting is actually just HALF of what is falling on your subject 5' away. At 10' away it's half again (or 1/4). At 14' away it's half again (1/8th). At 20' away it's half again (1/16th of the light) -- hence the dark background.

BUT... suppose your light is 50' (using a more extreme example) and the background is 5' behind them. The light wont fall-off by a full stop for 20 more feet (50 X 1.4 = 70). That means a background merely 5' back will only be slightly less lit than the model standing it front of them. The light fall-off would be well under 1/3rd of a stop.

The fall-off is happening based on a ratio, not linearly. It falls off the most rapidly in the first few feet from the flash, and the farther the light is from the flash, the more the fall-off slows down.

By putting the flash at a much greater difference to begin with, you've ignored that range of distances at which the light fall-off is happening rapidly and now you're using the area where the light fall-off is much more gradual.

 

[hirejn]

But then he indicates that some people think moving the light closer to the model will help but in reality it will actually make the background even darker because of the inverse square law. That's what confounds me...

Isn't light falloff caused by (or the "result of") the inverse square law? If yes, then it seems that moving the light closer to the model (and necessarily closer to the background) would illuminate the background because the light has a shorter distance to move and therefore will be more powerful. In other words...before moving the light closer to the model, let's assume that it needs to travel 10' to hit the background. So, it "falls off" for a distance of 10' and is not powerful enough to blow out the background. But, if he moves the light closer to the model and the light only has a distance of 5' to travel (and to "fall off") before hitting the background, why isn't the background exposed more?

Does that fact that the light actually HITS the model have an impact on the inverse square law? Is that what I am missing? Does the light hitting the model further increase light falloff?

It's a bit confusing, but Mark is right. Quantity of light falls off very fast initially. It's not about quality but quantity. So having the light next to the subject means there will be greater falloff from front to back and a background near the subject will still be dark. After a certain distance, the falloff is less severe and thus having some distance between subject and background is less significant. Thus moving the light further away from the subject places both the subject and the background in a zone where the falloff is almost insignificant and thus where they'll be more evenly lit.

You can test this with a meter. Start one foot from the light and move away. The first few feet it will fall off significantly, but as you move away the change will be less dramatic. If you expose for a subject near the light, you'll get the quick falloff. If you expose for a subject away from the light, you'll get the area of less falloff. The subject will be exposed correctly in both exposures; it's the falloff that will change.

Don't think about the inverse square law. I don't think many people think about inverses of squares when they shoot. You just have to visualize how it works. Just understand that the closer the subject to the light, the faster the quantity will decrease from front to back or side to side. The further the subject from the light source, the less it will decrease from front to back. If you try to fiddle with numbers and calculations and inverse squares of distances, you're missing it.

Here's another test. You have a background, a subject, and a light that doesn't move. The light stays where it is and the power doesn't change, and the background stays where it is, and the camera stays where it is. Let's say the background is 15 feet from the camera. You place the subject one foot from the background and meter the light, and then expose on meter. Both the subject and background will be fairly evenly lit because even though the subject is a short distance from the background, the light falling off within that foot is almost nothing that far from the source.

Then you move the subject about four feet from the camera, meter, and expose. The only thing that has changed is the position of the subject. With the same light at the same power, the background is darker. How can that be? That's because when the subject moved closer to camera, you exposed for the light closer to the camera vs. the light further from the camera, which decreased in quantity.

You could also visualize it as a graphic curve. The falloff will rise steeply within a few feet, and then it will level off. Or think of the sun. The light a few miles from the sun would be so bright you wouldn't be able to see anything else. But millions of miles away it falls fairly evenly on the Earth. This is why incident metering works. The light falling on a scene in mid day is the same light falling 10 miles away because it has already traveled so far from the sun, so if you take an incident reading in front of you, that reading will enable you to properly expose a subject two miles away in the same light. The light has already fallen off over millions of miles so it won't fall off any more a few miles away.

 

[Helen B]

I like amolitor's reply.

This is NOT the example I wanted to show, but it's showing the same effect in an illustration, but not in the photographs of the example I've been looking for. Note the illustration and the the text. Rules for Perfect Lighting: Understanding The Inverse-Square Law

"There’s a 75% drop in light from 1 meter to 2 meters, but only a 5% drop in light from 4 meters to 10 meters."

That is really misleading, or you could say that it is just plain wrong. If there is 100% light at 1 m, then there is 25% at 2 m - ie 75% of the 'light' that there was at 1 m has gone at 2 m (of course it hasn't gone, it has only spread out). If there is 6.25% of the 1 m light at 4 m and 1% of the 1 m light at 10 m, then the drop off from 4 m to 10 m is 5.25/6.25 or 84%. If you want to call it 5% then you run into a problem of choice of datum. The only comparison that is independent of choice of datum is to say that there is an 84% drop off.

 

[amolitor]

If you want to call it 5% then you run into a problem of choice of datum.

I am totally going to work "choice of datum" into all my technical conversations from now on. What a great way to compactly express that idea. I know it from navigation, but never heard it anywhere else and it never occurred to be to apply it anywhere else.