OK, sorry for a stupid question but is the Inverse Square law also affected by the size of the modifier or only distance from the lightsource?
Sure it's affected by the size (and shape) of the modifier.
If you're far enough away, you'll always get an inverse square law.
In most cases, for example if you have a diffuse modifier, you'll get an inverse square law if the distance from subject to the modifier is more than a few times the size of your modifier.
If your modifier is a large parabolic reflector with your light source at its focus, or some other kind of beam-forming optics, you might get fairly constant illumination over a distance many times the size of the optics.
If the distance between the subject and the modifier is small compared with the size of the modifier, the fall off will likely be slower than inverse square, and in some cases, there may be almost no falloff at all.
For example, if you have a large diffused LED panel, or a double-diffused softbox, so the light output is very even across the surface, the illumination will be nearly constant for distances up to something like half the size of the diffuser.
OK, so...if I put my model....say 3 meters away from
A: a 100 cm octa
B: a 190 cm octa
the fall-off - or to put it in another way; working area for my model with the same light output/exposure will not be the same?
By "working area" you mean the space in which your subject can move without the lighting changing significantly ?
You want to make the illumination of the model less sensitive to the position of the model ?
Yes, exactly this. But still keep my light soft and pleasing
Using a bigger softbox (for a given design), or moving the softbox closer to the subject, will do that.
A 1m softbox at 3m will make shadows, and the illumination will be more sensitive to small changes in the position of the subject than a 2m softbox at 3m. A 2m softbox at 1m would be much less sensitive again, but at that point the big softbox can interfere with other lighting, and with the freedom of movement of your subject.
Some experimentatation with inanimate objects may be helpful.
Thanks for that! Although I am not sure about the "moving the softbox closer" part in this situation. Yes, the light would be softer but isn't it the other way around about sensitivity to movements? The falloff is more rapid closer to the lightsource?
Not for a uniformly radiating flat surface.
I'd been trying to think of a good way of explaining this. Without a load of calculus.
...
Suppose we have a sphere with a uniformly radiating (outside) surface.
The intensity of the radiation (outside the sphere) falls as 1/r^2, but r is measured from the
centre of the sphere,
not the surface of the sphere.
What's the difference between the light intensity 1m from the surface of the sphere and 2m from the surface?
If the sphere is has a radius of 1cm, the relative difference is (very close to) 1^2/2^2 = 1/4. The distance from the surface and the distance from the centre are practically identical.
If we do that sum more accurately, we get 101^2/201^2 = 0.2525
If the sphere has a radius of 100m, the relative difference is (very close to) 100^2/100^2 = 1.
If we do that last sum more accurately, we get 101^2/102^2 = 0.980
A large, flat, uniformly radiating surface behaves like a patch of a very big sphere, so long as the distance from the surface is much smaller than the size of the surface. And there is very little light falloff with distance from the surface.
At distances much larger than the size of the surface, the surface behaves more like a point source, with the familiar 1/r^2 point-source falloff.
...
Another way of looking at it is: starting very close to the surface, only a small part of the surface effectively lights up the subject. As we move the subject away from the surface, the area of the surface lighting up the subject grows as the square of the distance from the surface, exactly balancing the 1/r^2 falloff from each point of the surface.
...
A symmetry argument:
Suppose we have a horizontal (just for concreteness) uniform flat radiating surface, radiating upwards, extending infinitely in all directions.
Consider a transparent rectangle in a vertical plane, above the surface. Call the sides left and right.
What is the net light flow through the rectangle?
The light flow from left to right and right to left must be the same. So there can be no net light flow through the rectangle.
Suppose we make a transparent rectangular box, above the radiating surface, with four sides all in vertical planes, and the top and bottom ends horizontal.
What's the difference between the light flow into the bottom, and out of the top?
We've already found that there can't be any light flow through the sides, so the light flow into the bottom, and out of the top must be the same.
So there is no light falloff, however tall the box is.
(Similar arguments lead to a 1/r law for an infinite rod radiator, and a 1/r^2 law for a spherical radiator).
...
Any other folk here who have a better explanation ?
...
As I said, some experimentation with an inanimate subject may be helpful.
So the optimal solution would be a big 2m softbox say...3 meters away from the subject. And perhaps use flags on the sides to shape the light and give the model some sculpting since I suspect the light will be quite flat and boring right out of the gate.
[Edit: fix thinko confusing radii and diameters of spheres. Result unchanged.]