cgardner
Offline
• • • • •
Upload & Sell: Off
|
 p.1 #7 · Basic Portrait Lighting Set-ups and goals in modeling faces | |
When lighting a face in a conventional portrait my goal is to flatter the subject. I don't as a rule find dark distracting shadows anywhere on the face or eyes half hidden in shadows to be flattering with regard to rendering a person as they might look ideally if you met them on the street or an office in flattering light.
The biggest variable in portraiture is the 3D shape of the subject's head. Monte Zucker, who I learned from, was a master at looking at a person's face and in a split second noticing all the imperfections which would make them look less flattering in a portrait. That informs you which facial angle / lighting patterns to avoid.
I learned to look at a face full, which reveals immediately whether or not it is naturally symmetrical. If not you'd want to avoid shooting them full face with a centered, symmetrical pattern (CS pattern). One of the more interesting optical illusions created by lighting patterns is that the highlighted side of a face on a darker background will seem larger than it actually is. That's due in part to the fact that on a dark background the brain will focus more on the highlighted parts of the face and mentally tune out the darker shaded ones. So if I see a person's face is wider on one side than the other when viewing it full face I know from experience that if use a CS pattern that will become obvious, but if I instead put the light at 45° to the side on the narrower side if the face in the full face view (FF_NS45° pattern - full face_narrow-side 45° key light relative to nose) the net effect of the symmetrical camera view and asymmetrical narrow-side highlighting will make the face seem better balanced in the photographic rendering than either a CS pattern or a FF_WS45° key light placement would.
Having done C&C on thousands of portraits it's obvious from the results I see that many photographers aren't aware of this cause and effect, or in many cases that their subject's face is asymmetrical. If you don't see the problem, correcting it becomes a hit or miss, trail and error process. If you try all three strategies, centered key light, key light on right at 45°, and key light at 45° to the left on a subject you will, by trial and error find which is the more flattering. When you learn from that experience how the asymmetry of the face predicts the outcome from those three basic strategies you will be able, like I am, to look at face when meeting a person and know where to set the key light (centered, right, left) to get the more flattering results for a full face pose, and which direction to turn the face in an oblique pose with the key light placed 45° to the highlighted side of the face with the camera shooting into the shadows.
A factor in the relative size illusion in asymmetrical (key light to the side) lighting patterns is the ratio of key vs. fill. Shadow tone is a function of fill intensity. The fill is under the highlights. If you where to set your lights by turning on and adjusting fill first you'd set the f/stop you wanted to shoot at, then adjust the fill light strength until the shadows were as light as you wanted them the be in the final result. If the reader were to actually try this some time they would notice how the fill, placed at chin level under the lens of the camera, models the face. Not sideways, front-to-back. The gradient you see from cheek-to-ear is caused in large part by the directional fall off of the fill front>back. If instead placed to the side aimed at the ear the gradient will be reversed and become back>front. Both work, they just produce different modeling of the face with the fill.
Natural skylight has two components due to the way the sun bounces off the sky and the ground surrounding a subject. The downward component will usually be brighter, which is evidence by shaded eye sockets seen in open shade. That's the "key" component modeling the face. You won't see any distinct shadows because the "fill" component from all the other directions is nearly as strong. But if you want the eyes in your portraits as bright as the cheeks if you see the eyes are shaded the first step is to have the subject raise their face so the "key" vector of the skylight reaches the eyes.
Natural light bouncing from the sky is so far way in terms of apparent source that here is no little visible fall-off front-to-back unless something is blocking the light from the sides. So the lighting you get on a face standing in the shadow 40' in front of the north side of a building will be different than the lighting you'd get on the front porch, doorway, or inside by a north facing window. The "key" light angle will be the same but the amount of side fill will be less. That's what is known as "subtractive" lighting.
If you were to shoot the person standing out in the open shade 40' from the building with nothing blocking the side fill there would be a pattern created by the downward direction of the skylight. By aiming the subject's nose at that brightest part of the sky you can get a variety of lighting patterns just when moving a face around under a fixed artificial source. But the lighting ratio will be very low.
What happens if tilt the face up into the light to get it into the eyes at a 45° angle then add flash so it also hits at a 45° angle as the brighter part of the skylight? SInce the modeling vector is the same the flash will create the same geometric pattern on the face. What the pattern will look like will depend on how the face was being modeled by the natural light, but it's character will be different because: 1) the flash source is smaller and will create more distinct shadows, and 2) the flash being much closer to the face than the sky will fall off much more rapidly.
For example when shooting a full face pose with a single speedlight on a bracket I will first note how the natural light is hitting the face, turning the face so the nose is pointing towards the brightest part of the sky. At mid-day where I live that means point the nose north because the sky is brightest opposite the sun high in the southern sky, then get the subject looking up. Seems perfectly logical to me to use the natural light to model the face in a flattering way by getting it into the eyes with CS_pattern BEFORE adding the flash.
No, you won't seen a dark shadow hanging down from the nose, in part because with the light at a 45° angle relative to the nose the nose, which projects from the face as about the same angle, winds up parallel to the key light's angle //. The flat front of the face and cheeks are modeled at 45° relative to the key light, but the ridge of the nose? That's at the same angle so unless the tip hangs down below the base the key light angle will not create a long downward shadow. A slightly upturned nose may not create any shadow at all.
What happens when flash is added AT THE SAME 45° RELATIVE TO THE EYELINE is that it will fall off more rapidly front>back than the skylight. The fall off of the flash creates greater contrast in the lighting pattern because when exposure is adjusted to keep the highlights correctly exposed it makes the sky lit shadow areas darker in the recorded image. Also because the size of the flash source is smaller than the diffuse skylight it will create more distinct and specular highlights which will make the lighting seem "harder" perceptually and change the perception of 3D shape from being more rounded in the flat skylight alone, to being more dimensional. Adding flash to an open shade shot will enhance the 3D modeling the lighting creates. Whether or not that 3D modeling flatters the face in front of the camera depends on how the face is turned into the combined sky and flash modeling vectors hitting the face at the same angle, small source in front of larger one.
Using a single flash to enhance the 3D modeling via highlights can create a new problem of shadows which are too dark and harsh looking. If the flash is hitting at a 45° downward angle relative to the eyeline there will be places it does not reach. The difference between using centered flash in open shade vs moving it to the side is that when centered the raised flash on a bracket hits 90% of everything that is seen in the foreground. The 10% that is seen shaded has dark sky-only filled shadows, but they aren't noticed much. The same is true for single flash shots taken with a bracket indoors. The perceptual cause and effect is the same — making the lighting more flattering by hiding the distracting shadows.
But when a single flash is moved off to the side and more of the face winds up in shadow the fact that shadows are darker than seems "normal" will be noticed and seen as a defect. The solution to that problem? Add more fill.
One of the differences between a raised 45°V flash assisted shot in open shade vs. the same shot taken in backlight is that the sun hitting the area surrounding the subject bounces much more fill. The environment becomes a variable. If you are shooting a backlit subject on a parking lot you will get less bounced fill than when you shoot at the beach or on a concrete sidewalk. At the beach or sidewalk you may get too much bounced up-fill. On grass the bounced fill will take on a greenish cast creating a new mixed color temp problem.
The point being that when shooting outdoors you need to have situational awareness of the direction and character of the natural light and then use that to guide the placement of any artificial lighting or reflectors you decide to use. If you like dark shadows hanging off noses and where shooting a portrait outdoors in open shade you'd need to know how to create that dark visible shadow by adding flash because you won't see it with natural light.
To me it is really irrelevant what a pattern is called. What matters most to me is:
1) There light in the eyes and on/in the mouth?
2) The nose shadow is not creating a distraction from eyes and mouth.
Keeping the key light at a vertical angle of about 45° —regardless of whether it is natural, artificial, or a combination of both — and using chin level centered fill to control the tone of the shadows and create front>back shadow gradient on the face accomplishes both. Dark nose shadow hanging off the nose aren't flattering so I avoid key light placements that cause it. When using a centered key light I try to angle the nose to the light so it reaches the eyes and the nose casts no shadow at all, or one that is hidden from view below the nose by keeping the camera above the nostrils. When using my key light to the side I position it so the nose shadow fall over the nose rather than hanging out (not possible with all noses, the goal I try to achieve). My centered fill strategy is based in part on physics — the closest thing to the fill will have the lightest shadow on it. Centered fill makes the nose the closest thing to it, so the nose has the lightest and least distracting shadow on the face.
My decisions are also based on the look I want, I just don't like dark distracting nose shadows in my portrait lighting.
|
