Upload & Sell: Off
I look at lighting problems in terms of goals and solutions. One goal of mine in a portrait is putting attention on the eyes and mouth by highlighting them because they are the area on the faces which trigger the viewer's emotional reaction. What sits in the middle of them and distracts? The nose.
Unlike the rounded cheeks, chin and forehead the nose is tall and angular as a result of its different shape a light source which has a family of angles large enough to wrap around cheeks and forehead will not wrap around the side if the nose. There will be a wrap gradient on the ridge of the nose, but the side of the nose and the side of the face opposite the light will remain as dark. That's what I was referring to when saying you can't wrap light around to the shadow side....
The wrap gradient created by the key light is the result of the umbra (dark core)/ penumbra (feathered edge) effect. The shadows on the shallow curved surfaces wind up being mostly lighter penumbra shadows because light from outer parts of the source hit at enough different angles to cancel the darker umbra created by the more parallel "line of sight" rays from the center of the source. But the nose too tall for the outer edge source vectors to reach the far side so it's shadow is all umbra and as dark as if a collimated source was used as key light. Wrapping will lighten most shadows on the face but not the most distracting one — the nose shadow.
The net result of wrapping the key light placed to the side of the face is an overall shadow pattern where the nose will winds up being the darkest and most distracting shadow on the face. I see that a lot anytime there is a "one light is better" thread.
Photographers who use one light typically wind up putting their reflector on the opposite side of the face where it winds up closer to the shadow side of the face than the nose. That fill strategy exacerbates the problem of the darker nose shadow created by the wrapping key light because the fill source is furthest from the nose. The fill lightens the side of the face and ear more, so in terms of relative contrast the nose shadow looks even darker by comparison to the other shadows because the otherwise dark shadow on the side of the face has been eliminated.
The dark nose problem can, and often does get worse in single light + reflector shots. When a reflector is placed further back than the front plane of the face the cheek bone will wind up between the reflector and shade the fill. When side fill is shaded you will see much darker shadows in lower places: around the base of the nostril, smile lines, corners of the mouth and back teeth, etc. That's not to say single "wrapping" key lights and reflectors don't work, you just need to take care to keep reflector as far forward of the face as possible.
WIth some key light strategies it's nearly impossible to keep a reflector in front of the face and put it where it can catch the key light. It is also problematic to create a 2:1 ratio with a reflector. A 2:1 ratio requires the key and fill to be the same intensity, which is physically impossible unless the reflector is a mirror. So all things considered when an artificial key light is used it's simpler and more effective to also use a second artificial fill light.
You can guess I don't like dark nose shadows much. The nose is the least attractive feature on the face and I want it to not be consciously noticed by the viewer when looking at the face. I want eye-lock.
One of the reasons I prefer butterfly patterns for full face views is that there are very few shadow clues about the shape and size of the nose created by the centered key light. The nose still casts a dark mostly umbra shadow down under the nose, but by raising the camera it is hidden by the top of the nose and not seen. That's what makes single flash full-face shot taken with a flash bracket so flattering - no harsh shadows anywhere on the face.
Fill is, I find, the least understood aspect of lighting cause and effect. For example, beginners thing bigger modifiers are need to "wrap" the key light and create smooth highlight > shadow gradients when the cause and effect of shadow gradients is a function of fill angle, distance and intensity.
Consider for a moment what happens when a fill light is placed directly over the camera lens. The light would be flat and uniform if it was illuminating a flat wall, but a face is three dimensional. With the fill centered the nose winds up closer to the fill than the cheeks, and the cheeks closer to the fill than the ears. The net result is a smooth front > back gradient of tone on the face with the nose shadow winding up being the lightest and least distracting shadow on the face. The gamma of the gradient is controlled by fill distance and inverse-square fall off. For very soft gradual gradients from cheek to ear put the fill further away. For a more dramatic roll-off from cheek to shaded ear move it closer.
With a centered "neutral" fill source before you even turn on the key light you have most of the qualities you want in portrait lighting:
1) Light in eyes and mouth
2) Light non-distracting nose shadow
3) Tonal gradient which makes the front of the face contrast from the darker sides.
What is missing is natural 3D modeling. But by starting with separate centered fill source, versus other fill techniques, there is no compelling need for a large wrapping light source to create the illusion of 3D. The fill is already creating the desired gradients so all the key light needs to do is create some well placed highlights on the highest parts of the face from a downward angle. That's the way natural light works.
I've always understood that big modifiers are not needed to create natural looking lighting because I started with artificial lighting in high school on a low budget using 100W photoflood bulbs in shop reflectors. I had my camera in one hand and a Kodak "How To Light Portraits" book in the other. It recommended placing fill directly behind and over the camera. I tried it and as you can see it worked pretty well — no harsh unfilled shadows and smooth transitions.
A couple of years later when I went to work for Monte Zucker he handed me a Rolleiflex with a Graflex flash over it on a bracket for fill and another identical flash on a rolling stand; the gear he used at wedding receptions. He seemed to have followed Kodak's advice too . It worked just as well with speedlights as it did with shop reflectors — the flash head of the Graflex was even smaller than shop lights, more similar to the size of my DIY diffusers.
Zucker used a technique similar to the one I learned in the Kodak book. He taught me how to put flattering light on a face by placing the OCF at 6ft. 45° from the subject's nose and 45° higher than the eye line to create a "short" lighting pattern, then by stepping back to 8ft the difference in the key and fill distances created a 3:1 ratio, which matched the entire range from grooms black suit to bride's dress perfectly to the range of a color print.
I didn't fully understand the cause and effect back then, but I could see with my eyes that simple techniques suggested by Kodak and Zucker created very natural looking and flattering 3D modeling on faces with very small light sources. That early experience with small direct sources taught me the impression of softness comes mostly from the clues created by the tone of the shadows, particularly the nose shadow, and those clues can be controlled with the fill source. Creating a "softer" look for a portrait was simply a matter of increasing fill intensity.
A problem with small sources is the specular highlights they create on damp or oily skin. So for me the more compelling reason to increase the size of the key light when centered fill is used is to control the character of the highlights, not wrap the key light around the face as much as possible. Bigger sources create bigger catchlights and more gradual Zone 10 specular > Zone 9 solid white transitions (and a bigger footprint that generates more spill fill).
The shots below are an unplanned comparison between my 580ex speedlites with DIY diffusers and my Alien Bees with larger modifiers using a similar centered fill lighting strategy. It wasn't a planned portrait session. The kids were house guests who had just arrived from a day sightseeing in hot weather. I grabbed the camera and speedlights to shoot the boy then decided to set-up the studio lights for a group and other individual shots.
In the shot of the girl despite more fill and lighter shadows the overall lighting appears "harder" because the highlight clues on her face are more specular. Conventional wisdom would predict the opposite, that the smaller DIY diffusers used for the boy would create more specularity in the highlights than the 20" soft box (24" x 32" with 20" circle mask) used for the girl. But he had washed his face on arriving, she wearing make-up didn't. Oily skin trumped the larger modifiers.
I know the DIY diffusers are too small to create any "wrap" effect which is why I use them with two light. The shadow gradient front>back on the face is the result of the centered fill falling off front > back. The larger key source used for the girl wrapped more, but overall the difference in the overall quality of the lighting isn't that different on the front of the face. I didn't reshoot the boy's solo shot with the studio lights because the speedlite shot had turned out well. The shadows under his chin is the result of the fill being up on the bracket vs. around tip of nose level for the girl with the studio lights and shooting from closer than I normally would which increased the angle of the fill.
My starting baseline for lighting is what looks "seen by eye" normal. What I learned from my time with Zucker shooting color negative film with flash was that lighting a face with a 3:1 ratio with key light 45°H/45°V from the nose fit the scene range to the print range with detail everywhere when combined with centered fill and the overall result looked just like the person did in real life when I took the shot. That's not creative lighting, that's more along the lines of just recording what was seen accurately despite the physical limitations of the process such as the shorter than eye dynamic range of the recording medium and the fact the image is 2D but must create the illusion of 3D with the contrast gradients.
Why is a 3:1 ratio normal? A 3:1 ratio fit the range of a subject from black suit to white shirt to a color negative/print perfectly. Today's digital cameras record about same dynamic range and I find the same 3:1 set-up for the light record a the same full range of tone, which technically is half of what makes a photo seem real (the other half is normal looking WB).
Why put key light at 45° to the side and 45° degrees above the eyes? Most of the clues to 3D shape in a 2D photo come from the shape and angle of the shadow and specular highlight clues. A key light at 45/45 perfectly models the shape of the nose by putting half of it in shadow, with the ridge shadow falling along the base and tip shadow falling over the nostril. It also puts the specular highlights in the same places on the face natural downward light does. For a full face pose centering the light on the nose in a butterfly pattern hides the nose shadow from view and makes it seem to disappear, which puts even more emphasis on the eyes and mouth. It also makes a full face view look more symmetrical than side lighting.
Why is centered fill normal looking? First because it mimics skylight fill as seen from the POV of the camera. Like natural skylight it makes the nose shadow closest to the fill source the lightest one on the face and creates the same front > back shadow transitions seen in natural open shade and indoor room lighting. By comparison artificial fill placed in other positions casts shadows which will become unfilled voids in the lighting pattern and other than front > back gradients.
The "unseen" variable in the cause and effect is the spill fill. All light sources, even direct ones, create it to some degree. Joe Zeltsman, who taught lighting to Zucker, bounced his studio fill backwards off the back wall of his shooting gallery to create a more natural wrap-around character to his fill. I take it into account when using my speedlights and when using a dish as my studio fill source. I want my fill to spill and wrap, but my key light to go only where I aim it so use a SB for key light to control it's spill.
More often than not when a big modifier is use lightening of the shadow side if a subject comes not from the wrap, but from the fact the large footprint bounces as much light off the ceiling and walls as it puts on the subject directly. That is certainly the case with umbrellas and shallow dishes, and even with large softboxes. Next time you shoot put a wide angle lens on the camera, move to the back of the room and trigger the key light lights to see the amount of spill — it may surprise you