cgardner
Offline
• • • • •
Upload & Sell: Off
|
 p.2 #2 · Feedback on basic flash technique | |
The traditional approach for learning lighting is rote duplication of classic patterns like butterfly, Loop, Rembrandt, etc. Been there, done that.
But what I came to realize upon learning how to create them is they all aren't ideal for accomplishing the real goal in a conventional portrait which is flattering the subject. That's the most important lesson I gained from my apprenticeship assisting Monte Zucker. He was a master at analyzing faces, spotting all the irregularities and knowing which combination of camera angle and lighting pattern would flatter them the most.
Most heads look wider and larger full face than if turned to the camera obliquely. When viewed full face with a centered key light overhead in a "butterfly" pattern any asymmetry on the face will be obvious. If none is seen, as on a mannikin, then the full face view will be flattering but usually not the most flattering.
A trait that photographic models have which most aveage faces dont' have is natural symmerty, well defined cheekbones and the shape of skull that when viewed full face the face looks slender, even when the lens compresses the planes of the skull in the photo. What makes the model's face look slim and an average face look wide in a full face view is what is behind the eyes and front "mask" of the face. When a person has a small cranium very little is seen past the eyes. If the person has a big crainum it is seen behind the eyes in the full face view and when the lens compresses the planes the overall impression is a wider face.
The modeling in a centered lighting pattern is more subtle than in a sideways 45V/45H key light pattern. If the cheekbones are not prominent the lighting wiil not model their 3D shape much making the face appear flat dimensionally. That's why the mostly flat centered butterfly pattern while flattering on a slim model or mannikin face isn't always flattering on average faces. The face, even if symmetrical will wind up looking very wide and flat.
What the oblique view does is reveal the 3D shape of the cheekbones and brow via the camera angle and contrast of the far side of the face with the tone of the background. But an oblique view with centered lighting will look wider than the full face view. It also looks very wide when the key light is placed on the "broad" side facing the camera on a darker background . But when the key light is moved to the opposite "short" side so it only highlights the front of the face it will look much slimmer due to a perceptual trait of focusing attention on what contrast the most with the background.
The net effect perceptually of a well placed "short" lighting pattern is that the face seems slimmer and very symmetrical. The overall balance will shift as the tone of the shadows is changed with the lighting ration on a dark background. The more the shadow side of the face matches the background the more the viewer will focus on the more strongly contrasting highlighted parts.
In terms of placement of the pattern what creates the illusion of symmetry is that on the near side of the face when the key light highlights reach just to the corner of the near eye the line formed from eye, over cheekbone, to chin will look nearly identical to the profile which was created on the far side with the oblique camera angle. The brain reacting to what is highlighted sees the same "mask" of highlights on the oblique pose that is seen on slender models face with centered light. The net effect of combining the asymmetrical camera angle and assymetrical lighting pattern 90 degrees from the lens axis is an ILLUSION of 3D shape and symmetry, even when a face isn't naturally symmetrical. That combination will also flatter a symmetical face. The only type of face it doesn't flatter is a very long narrow one.
I did this exercise to illustrate this and to figure out how to take a decent looking self-potrait:
http://super.nova.org/XP/FaceSymmetry/CEG_Mirror.jpg
I started with a full face view then mirrored the right and left sides in Photoshop to more easily see the asymmetry. I'd been looking that face full in the mirror for 50+ years at that point and had never realized how lopsided it was. Which of the two mirrored versions looks more flattering? That provides the clue which oblique view will be the most flattering in a "short" lighting pattern. Here's the final result:
http://super.nova.org/XP/FaceSymmetry/CEG_oblique.jpg
Here's the set-up:
http://super.nova.org/XP/FaceSymmetry/CEG_Oblique_Setup.jpg
Because of the asymmetrical shape of my narrow face and chin, and the fact one ear is lower and sticks out more than the other I look better in an oblique view. But to get balance in the oblique view I needed to adjust from the "classic" oblique view that would work on a more symmetrical face with a "strong" chin and also let the far side ear hang out in a way that looks balanced as in a full face pose where both ears look the same.
Because I have a nose with tip higher than base the nostrils I needed to get the camera above my face to hide the nostrils more than for a nose where the tip droops down and hides them in an eye level view. Look at the faces of actors and models and you will not see many upturned noses. That's because it's easier to flatter the nose when in naturally hides the nostrils. Since I was shooting on a tripod that put the camera about chin level I had to lower my face relative to the lens, which is why I'm sitting on the stool. When I shot other adults I'm usually standing on the same stool to get the camera above their eyes for the same reason: the topside view of the nose is more flattering than the view up the nose.
The camera wound up about two feet higher than eye line and about 12 ft away. Shooting distance changes facial perspective, how big nose looks in relation to ears and balance of near / far eye in oblique views. I start from 8ft, which works for most faces, then compare the face through the lens from closer and further. I was about as far back as my space made possible.
The light placement was predicated on where the face wound up pointing when the view was flattering to get the "mask" pattern of highlights on the face. When not shooting self portraits the workflow is different. I first pose the face to the light, then find the most flattering camera angle.
I have enough experience to look at a face and spot that it is symmetrical and know which view full, right, left will be most flattering. But I usually start anyway with a centered lighting pattern and full face views to confirm this, have the subject take a break, then move the key light to the right or left side depending which direction they face obliquely is the more flattering side; their "best side". Unlike models a lot of actors have asymmetrical faces and know their "best side" for oblique views and how to light and "cheat" the face to the camera in full face views to make it seem more balanced to the camera.
Precise alignment is needed with the oblique/short combination to achieve the illusion of symmetry. If the key light is too far forward it will start to wrap past the near eye, increasing the area that is highlighted and making the face seem wider. It will also often go through the < notch of the near side eye socket and hit the ear, making it pop out and distract on the otherwise shaded side of the face.
The goal is creating a balanced look between both highlighted sides of the face. Due to variation in the shape of faces there's no absolute "rules" for how to angle the face to the camera. But the view is more flattering when the ear isn't seen hanging out past the profiled cheek and the camera isn't moved so far to the shadow side that the eyeball, lashes and eyebrow stick out. You need to evaluate how the far side of the chin looks in the oblique angle. When a subject has a narrow or receding chin it will often disappear from camera view when the modeling near the eye socket looks ideal. Sometime when the ideal balance between the look of the eyes and chin is achieved you'll still see the far side ear (or earring) hanging out and creating a distraction.
What I was trained to do was to first align the face to the light to create the symmetrical mask pattern and flattering 3D modeling of the nose with how the shadow falls over it, find the angle which profiles the eye socket without the eye projecting out in space, then look at the chin to see how it is balanced on the near and far sides on the highlighted front part of the face. Then you move the camera up/down, left/right a few inches to judge, by comparison if there is some other camera angle which is more flattering.
This was very intuitive because Zucker used window lights for portraits. There weren't 100 different lighting options to try (and confuse). You took the subject, put them near the window and turned the face towards it. When the nose got about 45 degree to the window it started spilling past the front of the face and hitting the shadow side ear so you stopped turning it and as simple as that achieved the ideal symmetrical "mask" pattern of highlights on the face. Then by aligning the face up/down you change how the nose shadow fell over the nose trying to get it to fall entirely over the nose, not hang out. Then with the face posed to the light you walk around the face to the shadow side and move the camera relative to the face to find the most flattering angle.
While one of the most flattering combinations the oblique / short combination does not project the ideal overall body language for something like a business headshot. But one of the qualities of that 45/45 "short" (when seen obliquely) pattern on the face is that it also looks very natural and flattering when the camera moves to full face or profile. I did this exercise to illustrate why:
http://super.nova.org/XP/snowhead2008/ShortFar.jpg
Starting in the morning but knowing where the sun would be in the sky at mid-day, 45 degrees above the horizon in the SW part of the sky, I carved the snow head facing west so the sun would be 45/45 from the nose and eye line at mid-day.
If the camera is placed so the sun is at behind the camera lighting isn't totally "flat" because it's coming from a 45 degree downward angle:
http://super.nova.org/XP/snowhead2008/SnowBroad.jpg
Moving around so the camera is full face and key light is 45/45 to the face produces this modeling:
http://super.nova.org/XP/snowhead2008/SnowFullFace.jpg
In the oblique view the face is 45 degrees to the camera 90 degrees from the source:
http://super.nova.org/XP/snowhead2008/SnowShort.jpg
Moving to the side the profile is seen and the key light is 135 degrees from the camera axis:
http://super.nova.org/XP/snowhead2008/SnowProfile.jpg
It's not a perfect profile because I'm not that good of a sculptor 
It's the same lighting pattern on the face, seen from four different camera angles. Which make the face appear symmetrical? The oblique and profile views because of the way what is highlighted fools the brain. The full face view is modeled in a flattering way but doesn't look as symmetrical.
Missing from that comparison is the centered lighting pattern because I couldn't turn the head to face the sun. I remedied that the following year by aligning the head due south planning for centered lighting with the winter solstice:
http://super.nova.org/XP/Lighting_Snowheads/Snowhead2009/SnowHead2009_Solstice.jpg
In the morning when the sun was in the SE sky it produces oblique/short modeling from the left:
http://super.nova.org/XP/Lighting_Snowheads/Snowhead2009/SnowHead2009_1.jpg
In the afternoon the sun in the SW sky created the same modeling when viewed for the right:
http://super.nova.org/XP/Lighting_Snowheads/Snowhead2009/SnowHead2009_Afternoon.jpg
When a face is naturally symmetrical all three views will look symmetrical and both oblique views will be identical if the key light is also moved to create the "short" pattern. But when a face is not naturally symmetrical a center lit full face view will NOT look symmetrical in the photo and the right and left oblique views with similar lighting WILL NOT look the same. One will be more flattering than the other.
So when you get a face that is not symmetrical you shouldn't just arbitrarily turn the face one way or the other, you should compare both views and pick the most flattering side. Then by placing the light on the same side the nose is pointing for the "short" pattern. You to make the pattern fit the face not every face to the same rote pattern.
If the face is not symmetrical and a full face pose is required you take advantage of the optical illusion that makes the highlighted side of a 45/45 pattern seem larger than the shadow side and highlight the side of the face which is physically smaller so it will appear more balanced in the photo.
None of this rocket science or even very difficult if you start by understanding the goal of the exercise is flattering the subject and accept that using strategies which make a face appear symmetrical and slim will achieve that goal, and also understand how natural light typically models faces as illustrated with the snow heads. Putting a key light 45/45 from the nose isn't "rule" based, it based on observing how faces appear in natural lighting and recognizing those angles and patterns form the perceptual baseline for what seems "normal".
Once you master the 45/45 key light angle on a variation of real faces in full, oblique, and profile views and try all the other lighting option like Loop and Rembrandt it won't be from the criteria of matching the pattern you saw in a book (usually on a symmetrical faced model) it will be from the criteria of "Does this really flatter the subject?"
Many people I've photographed who had never before seen their face in a "short" lit oblique view or 45/45 full face view with the light placed to make the face look as symmetrical as possible comment on how much better they look. Their baseline for comparison? The same full face flat lit view in the bathroom mirror as in my mirroring experiment above.
Lighting and portraits do not always need to be flattering but light in the eyes and smiling mouth is the norm for a flattering portrait and patterns that put light in the eyes and a ratio that keeps the shadows normal or lighter than normal and make the face look slim and symmetrical will be more flattering. When a portrait isn't flattering by those criteria I call it a "character study" not a "conventional" portrait.
A portrait with 4:1 ratio Rembrandt lighting will not be particularly flattering for the wife, but will be ideal for a character study of an old man because it makes anyone in it look old and moody or wise. Loop lighting? It is seen in a lot of classic paintings because painters used high windows and skylights as the key source. But is it as flattering as the centered-on-nose pattern which models the face but also can completely hide the nose shadow with the nose when the camera looks down? YMMV but I don't think so.
|
