Is it true that a soft light source (=large, even source of light) becomes harder when you move it further away from an object? I mean it becomes relatively smaller when you look at it, so it should become a harder light source with more distance, right?

The refection will appear harder but because soft light has such a steep roll off, by the time you move it far enough away to make it 'seem' hard it likely won't light you subject correctly.

Softboxes are soft because the effective area of light transmission is physically larger. When you move the box away, the surface area of the light become smaller in relation to the subject so it acts harder.

It's most a problem with hard reflections on say a guy's bald head.

That's right. I've heard people say this before and I can't believe it. Soft light falloff is so steep it won't make any difference and will be so weak the farther back you put it you won't notice any "hard" effects because the lighting won't be adequate.

I think that we could all agree that the sun is a hard light source (assuming a sunny day without clouds). It's pretty big though, we're just far enough away from it that it turns into a hard light source.

Just illustrating the point, but I think in practical lighting terms the falloff would be so dramatic that you would never be able to turn a soft light source on the other side of the room into a hard source because of falloff.

Just to be clear - Light falls off at the inverse square of the distance, NO MATTER WHAT THE SOURCE. This is true for hard, soft, medium whatever. It's a simple matter of physics. This is true of 'intensity'.

The 'quality' of the light source ie:hard or soft is dependent on the SURFACE AREA OF THE GENERATING SOURCE. So.. a light that is 'generated' from a large surface IE: A 4x4 silk or bounce board will appear 'softer' than a light generated from a smaller 'direct' source such as a six inch open faced light or PAR or six inch fresnel. (That's why we bounce these sources)

SO, when you move the large soft source further away from the subject, the intensity of the light source falls off at the inverse square of the distance, and the 'relative size' of the generating surface also becomes 'smaller' hence the quality seems to be generated from a 'small' source... like our six inch fresnel.

So, two things to remember that controll the quality of light. The DISTANCE and the apparent size/shape of the generating SURFACE.

(And yes, the SUN is a 'hard' source, who's quality is alterered when passed through the 'silk' of an overcast day... same thing, just on a cosmically larger scale.)

I think there are some wrinkles regarding extended sources though that will play into this.

Let's consider a point source at 4 feet from a subject. As you move the point source to a distance of 8 feet, the illumination on the subject will fall to 25% of what it was.

Now lets consider an almost infinite rectangular plane of light emitting material. As you move the light source (the entire plane) directly away from the subject, the light emanating from the point closest to the subject (let's call it the "center" of the infinite plane) will fall off as we expected from the first statement.

However, the distance from a point on the edge of the almost infinite plane (I wanted to talk about the edges and corners, so am using the concept of an almost infinite plane instead of a really infinite plane) is only infinitesimally further away from the subject that it was before, so there is an infinitesimally small fall off in the admittedly low level of light striking the subject from this point.

Therefore, the total amount of light (considering that there are an almost infinite number of almost infinitely distant point sources) falling on our subject will have fallen off somewhat less than would be predicted by the point source model.

Therefore it is safe to say that the fall off of an extended source will be less abrupt than the fall off of a point source, ie somewhat less than would be predicted by the inverse square law.

The more closely the source resembles an infinite source (ie is large compared to the subject) the less the fall off, and the more it resembles a point source (ie is small compared to the subject) the more the fall off for the same change in distance.

Try this experiment: Hold a light meter 1 inch from a table lamp and measure the light intensity. Move the light meter to a distance of 2 inches (twice as far). Do you really think the illumination will be only 25% of what it was for the first measurement? Or do you think it will be practically unchanged?

In other words, to sum this all up, the falloff of an extended source is less abrupt than that of a point source.

The falloff of a fluorescent lamp is less abrupt than that of a halogen bulb. The laws of physics as we know them are convenient abstractions applied to abtract objects. Their application to real world objects as we know them in a real environment is approximate.

A marshmallow two feet from the surface of the sun will melt just as fast as one that's only one foot from the surface of the sun. A marshmallow two feet from the tip of a halogen bulb will probably melt slightly slower than one that's one foot away

Your mileage may differ!

Have fun!

Sorry, you can't change the laws of physics. The intensity of light falls off as an indirect square of the distance.

Period.

However, the 'quality' of softness... that is, the percieved harshness due to the focal point of the light emitting source and it's size, is ALTERED by it's apparent size. No arguement about that. There is a difference in the 'quality' of light that comes from a focused fresnel, a difuse softbox, a set of Kino-Flos... yes. This is apparent for a number of reasons related to the reflector backing the bulb, the material the 'difuser' is made of, etc. etc.

The apparent 'hardness' of the light source will alter with it's distance and percieved size relative to the subject... but the intensity of the source follows the laws of physics.

And yes, I've checked it with a light meter. Remember to factor in the COVERAGE of the ammount of light.

Yep - inverse square law applies 100% to uniformly radiating point sources in a vacuum.

How about the falloff in a laser beam?

If you stand 1 inch from a large lamp will there be 25% as much illumination at 2 inches? Will there be 1/10000 as much illumination at 8 feet?

I don't think so

To expand on Richard's comments a bit, he introduced to factors: inverse square law and surface area.

One point of clarification re inverse square law is that it is defined for a point source. With a reflector or broad source you won't see the light intensity fall off at the same rate.

You can prove to yourself with a light meter by measuring the incident exposure at various "f-stop feet" from both point (bare light bulb) and broad (reflector) light sources. If the inverse square relationship is in effect then the exposure should double when going from one f-stop distance to the next (e.g. from 5.6 feet to 8 feet).

Mathematically we can show why this is so. Consider a linear light source that is 4 meters long. Place the subject at a position that is along a line perpendicular to and passing through the center of the light source. The distance from the center of the light source is D1 and the distance from a point L along the light source to the subject is D2. By the Pythagorean theorem the relationship between D1 and D2 is:

D2*D2 = D1*D1 + L*L

Note the D2 is always greater than D1 when L > 0.

The intensity of the light from a point light source as it moves from point d1 to d2 follows the inverse square law:

i2 = i1/(d2-d1)^2

E.g. intensity at 4 feet = intensity at 2 feet / (4-2)^2 = i1/4

Jim,

Thanks - exactly my point. I was actually just about to type up the same example of an elongated fluorescent that you used. You saved me a lot of typing.

And the trick with a laser is that the energy does not radiate uniformly in all directions, therefore falloff is not governed by the inverse square law.

Not because we're changing the laws of Physics, but rather because we're being clear about the definitions underlying the laws and which laws apply to which kinds of objects in which domain.

A real room is full of reflecting surfaces, dust particles, whatever and all real physical sources are extended sources. The concept of a point source is an abstraction that is very useful in some but not all situations.

You guys don't need to worry about all these equations. Just do what I do, get a bunch of lights and a monitor and start moving the lights all around the place and turning them on and off until it looks good on the monitor.

Shhhh! Don't tell anyone!!!!! It's a secret!!!!

That's what I do too!

Who knows what would happen if this information got out!

According to Wikipedia:
Electromagnetic radiation:
For another example, let the total power radiated from a point source, e.g., an omnidirectional isotropic antenna.
++++++++++

So, in other words, the light source has to be omni directional to have the law apply.
If the source is a focused light, like a laser or leko, the law does not apply.

And it has to be a point source - which doesn't exist in the real world (ie size = 0 Really zero.)

And to be really up to date about it, it should be outside of a gravitational field as well!

But as Brian said, it isn't necessary to worry about it unduly - just make it look right.

The point about point sources is mostly of use when the light is close. If you do the math in the previous article you'll find the ratio of intensities i1/i2 approaches 1 as the distance becomes great. Something else also happens, the angle subtended by the light source becomes smaller, so the broad source effectively becomes a point source at large distances, i.e. the light becomes less soft and shadows become harder & more defined.

Back to the original question for a second. Imagine you are shooting a product shot of a golf ball on a tee, a few inches away from a background surface. Let's also imagine you have a 12" square Chimera (flo, whatever--doesn't matter what the source, it's 12 x 12 of evenly distributed light).

Position the light 10 feet away and fire it up. We would expect the golf ball to throw a fairly defined shadow onto the background. Now move it to within six inches of the golf ball (and knock the output down so that it exactly matches the level from 10 feet away). The shadow will be all but nonexistent.

The issue of output level is somewhat irrevelant to the discussion, as it is entirely possible with even humble lighting sources to significantly alter the quality (hard to soft etc) of a given source depending on where you place it relative to the subject. The example above is extreme of course, but even if that same source was placed at a more realistic six feet away versus 2 feet away, there would still be a noticeable difference in the quality of light/shadow and the exposure difference could easily be contained by stopping down the camera and/or applying scrim/nets/dimmer etc to the unit.

Thank you Charles.

That was my point in mentioning the sun earlier. If you're 2 inches away from the sun (and figure out a way to not melt/burn) there would be virtually no shadows because of the immense size of the sun compared to you. Now here on terra firma, because of the distance of the sun and the relative size of the sun because of that distance the sun becomes a "hard" light source. It's all relational - between the object and the light.

Another example would be if you took a standard softbox (pick your favorite brand and shape) and move an object close to it, the light will have soft characteristics.

Now let's move the same object a substantial distance from the light (let's be absurd - let's talk miles, not feet). At this distance, the light (if we had a camera sensitive enough to pick up the light and everything else is dark) will have hard characteristics.

This is because at this absurd distance all of the light that is reaching the object is traveling in near parallel. And the near parallel light will cast hard, or sharp, or distinct shadows. The reason for the soft shadows in the first part of the example is that the light is striking the object from many different angles.

Yes, we are saying the same thing. I just thought that it would be helpful to include an example that could possibly be duplicated here on earth (since few of us have access to super-insulated spaceships etc, although I would not be surprised if Jim Jannard had something along these lines).

On a more practical level...

You'll likely find softboxes almost useless for anything other than static product shots, faces, etc. a 6' x 6' Butterfly, backlit with a sheer fabric, can generally be good for full-frame people shots. If you want to move around a room, you need a BIG soft source; bounce off an entire white wall or scrim up a wall of bedsheet-size fabric on stands with several open face lights behind them (work lights can do this well, but the rectangular open faced lights with barndoors are better -- and there are usually several on eBay any given week). In this case, you need enough room to get your lights a few feet back from the scrims, and aim them towards the top for a more natural look. Bouncing off the ceiling at the same time can add to the natural factor. You'll get to the point that when you look at a given set, you'll know what you need to get the look you want. All the pricey gear in the world can't beat experience.

How far would I need to be away from a neutron star to get a hard light source?

Are black holes good negative fill or does the Hawking radiation negate the black?

Seriously, I would explain a "soft" source as simply one relatively large compared to the subject. Think of every tiny point of a light source as an individual ray of light that casts it's own shadow. Draw an imaginary line between all points of a light source and the subject and you will realize that a large source will make infinite shadows in all directions from the features of a subject. All those little shadows in random directions blur the line of the shadow. A hard light like a tungsten filament inside a clear glass bulb will make a very distinct shadow. It is possible to make a soft source of light from tungsten filaments by having lots and lots of them spread out over a few feet. Get a bundle of rope light and put it near a subject. All those points of light spread out the shadows.

Cute... .

I actually agree that using the sun is a good point. It illustrates that relative size and distance are important to how soft a light source appears. I visualize soft sources as individual points of light all generating their own shadow to make it simpler to comprehend.

If a source is to maintain the same character of light & shadow as it is moved away from the subject the ratio of the diameter to distance must remain constant. As the ratio decreases the light becomes harder.

I wouldn't get too hung up on theoretical concepts about point sources; an approximation is good enough for understanding. The idea of a source consisting of a bunch of point sources originating from various points does a good job of conveying how the character of light is affected by their distribution.