LIGHTING STRATEGIES: Soft Light vs. Hard Light
Hard lights are great for textures, but soft lights are great for defining spaces.
By Art Adams | December 21, 2011
In a previous article I described how to place a hard key light. In this article I'll look at why soft sources are a bit more realistic in color cinematography, require less precise placement, and can help define the volume of a space.
In my last article we talked about how to light faces: where shadows should fall for "classical" modeling and what to look for when struggling to create a classical portrait of someone who has a non-classical facial structure.
In this article we're going to talk about soft light, the most forgiving light we'll ever use on a face. This doesn't mean that we should always use it--the story and emotion driving the project you're shooting should guide you in your lighting strategy--but soft light is most forgiving of facial oddities. It's also a style that we easily accept because, in the real world, any setting that is not lit by direct sun is usually lit by some sort of bounced or diffused light.
Hard light works wonderfully in black-and-white filmmaking because the medium is already a bit abstract: hard shadows play better when the image is one step removed from reality, and the lack of color certainly creates a different world for the viewer. Hard shadows also reveal textures well, and our eyes love texture. Soft light, though, generally plays better in color because it's closer to how we see the world. Hard light can look fake in color photography because multiple hard sources can make a space feel "lit": the quality of the light is not what we'd expect in that environment, and that pulls us out of the story. There are exceptions to this (the first three Raiders of the Lost Ark films are great examples) but it takes a lot of practice and a good eye to get away with hard light in color cinematography.
But enough about theory. Let's look at some virtual faces, created in Poser 9:
In my last article I presented this image as an example of classical Rembrandt modeling with hard light. The trick with hard light, though, is that it is quite face dependent. What works on one face may not work on another:
This example isn't as bad as it could be because Poser doesn't supply me with challenging facial models. As you can see, though, her face looks very different from his. His lighting makes him look ruggedly handsome whereas hers is acceptable, but maybe not optimal.
This is what happens when I diffuse the key light significantly. This look is reminiscent of what might happen if I placed a frame of Lee 129 (a very dense diffusion material) two to three feet to the side of his face, between him and the lamp. The camera right side gets all the light, but the camera left side gets only a little bit of the light. If you look in his eyes you can see the soft key strongly in his camera right eye but very weakly in the left one. You can also see the fill light I placed near the virtual camera as a pinpoint in both eyes.
Let's see what happens to his female counterpart under the same lighting:
Wow, that's beautiful.
There are three key differences that I see between hard light and soft light:
(1) Hard light creates hard shadows that enhance defects and create very dramatic modeling on a face, whereas soft light's broader gradations conceal facial blemishes and don't have to be placed as precisely because they aren't as distinct.
(2) Hard light usually creates two tones on a face: a bright side and a dark side. Soft light creates an almost infinite number of tones in the transition between the key side and the fill side, and our eyes like tonal complexity.
(3) Soft light comes from a source that's large in relation to the subject. As healthy skin has a little bit of shine to it the source will reflect in the skin, adding a very soft but beautiful highlight. Hard lights will not do this unless the face is very, very shiny, and even then the highlight will be a small one.
Putting diffusion material between the light and the subject doesn't automatically soften it. The size of the diffusion and how close or far it is from the subject make all the difference. When you diffuse or bounce light the lamp itself is no longer lighting the subject; rather, it is lighting the diffusion or bounce card, which is glowing and lighting the subject with the light that it is emanating. The lamp makes the diffusion or bounce surface bright, and that radiating brightness lights the subject.
The softness of a light on a subject depends on how big that light source is in relation to the subject. For example, a 4'x4' frame of Lee 129 diffusion placed 3'-4' from a person's face will create very soft shadows, whereas the same frame placed 30' away will cast very hard shadows. From the subject's perspective the close frame will appear to be a huge source, while the far one will be very small.
The example above shows what happens when the diffusion is placed to the side of a face. Let's see what happens when we bring it to the front of the face, closer to camera but still to one side:
This, too, is a very interesting look. The shadow on the camera right side of his face near his ear is interesting but won't happen in real life; it's a side effect of how Poser renders light and has no bearing on reality. The rest, though, is very realistic. Note how the nose shadow is minimized but still gives the nose shape due to the high contrast of its decreased shadow. That shadow isn't hard and harsh but soft with a lot of gradation over a short distance. The front surfaces of his face reflect the large source and appear to glow: I see this in his forehead, just above his smile lines and on his lips. This effect is exaggerated here by the software but you should see some of this every time you use a soft source close to a person's face.
She looks great as well. Notice how soft the highlight from the key light is in her eyes. The larger the source the less bright any part of it will be, so it creates a larger but dimmer eyelight. This is why an intentional eye light is usually created using small, dim, specular sources, but that's a subject for a future article.
Let's move the light dead on:
This is classic beauty lighting, although on a man it doesn't feel as glamorous as it does on a woman. The light is coming straight down the axis of the lens. If I were to do this with physical lights I would cut a hole in a 4'x4' piece of foam core, stick the lens through it, and then light the front of the foam core as evenly as I could with two lights, one from either side. Placing the camera and the rig about 4' away from the subject will produce this look, where the center of the face gets all the light from the card but the sides of the face receive less light as they curve away from the opposite side of the bounce card.
The look is similar to that of a ring light, although a ring light is smaller and casts sharper shadows for a more dramatic look.
This look can also be accomplished by placing two 4'x4' Kino Flos horizontally on either side of the lens, exactly at the lens height and as close to the lens as possible. The look isn't exactly the same but it's similar.
Here's what this setup looks like on a woman:
This is an idealized face lit with an idealized light, but you've seen this look in a hundred makeup and fashion commercials. The center of the face appears to glow due to the soft reflected highlight of the source while the sides drop off because they only receive light from one side of the source and are blocked from the other.
I've found that building a source that's as wide as the distance to the subject works best at achieving this look. If the subject is 8' away from the camera I'll build an 8' wide by 4' high source. (For this look the width of the light is more important than the height. More on light shape in a future article.)
Let's look at soft lighting from below the lens and using soft light to define spaces on the next page...
Get articles like this in your inbox: Sign Up