cgardner
Registered: Nov 18, 2002
Total Posts: 8551
Country: United States
|
First your set-up isn't a good strategy. it is cross-light and there will be no fill for the shadows created by either light.
Instead of starting with light placement start with a goal or two, like detail everywhere in the photo and a flattering pattern on a face with no distractions.
The foundation of any lighting is fill. Its a technical necessity if you want shadow detail because the range of a digital sensor is too short to record highlights and shadows at the same time in a full range scene.
If starting in a dark room you'd first want a light over the camera. Why over the camera? That is the only spot the light will reach EVERYTHING the camera sees. Move the fill off axis and it will create shadows. Where fill creates shadows there will be NO DETAIL IN THE PHOTO. You'd raise the fill until the camera sensor can record detail in the darkest areas where you want detail.
Separation with a dark background can be controlled by using a separate light to illuminate the background (a black sheet in this example). Separation can also be achieved by just starting with a lighter background and letting the fill illuminate it: distance will control how light or dark it is rendered.
The illusion of 3D in a photo is created by contrasting highlight and shadow. The brain is tricked into thinking the lighter parts are higher and the darker ones lower. If you start with fill, to the level the shadows are revealed exposure is quite simple: keep adding the off axis KEY light until the "paper" white highlights (i.e. solid objects) are just below clipping and detail is retained in white fabric. 255 = pure light. If you want a fill range of tone captured by the camera the only thing that should be 255 in a photo, as captured by the camera, should be specular highlights. If you want to blow the highlights, its better to do it in Photoshop where it can be precisely controlled.
All that is needed really is the camera, its playback and your eyes and brain. Yes you will find a meter helpful, but understanding how a meter works and not understanding how lighting works will not get you photos which fit the range of the sensor and flatter the subject.
The role of rim light is to enhance the illusion of 3D. The most effective illusion of 3D is actually created with rim light and frontal fill...
The rim light defines shape, the fill reveals detail. Again I stress the need to keep the fill centered if you want to avoid harsh dark shadows. This is a shot taken with fill on a bracket and the off camera flash parked in the corner behind, an effective strategy for action shots.
When added to the three lights in the exercise above back light creates a stronger illusion of 3D:
Lighting a face adds a complication. We reach emotionally to eye contact. To do that we need to see the eyes. To see the eyes we need to put light in BOTH of them. Hence the use of a key light in front of the face on top of the fill. But there's also the sundial nose to contend with. The most effective strategies for modeling a face in a flattering way either hide the nose shadow or use it to model the shape of the nose in a natural way. There are no rules for light placement, just cause and effect. Natural light comes from above most of the time. So logically, when the artificial light creating the highlights comes from the contrast pattern will look "normal". If the key light gets lower than a face or falls below casting upwards shadows the face will look unnatural.
The important relationship for the key light is the angle to the nose, not the camera axis. To short light a face, the key light needs to wind up about 45 degrees from the nose, wherever the nose is pointing...
If using a window you would first angle the face to the light, then capture the three basic angles in short light by simply moving the camera around the face..
As for learning how to use a meter for exposure you first need to grasp the design. Reading the user manual will help also. The dome on the meter is design to be held in front of the face to average all the light hitting the face. Hold the meter in front of the face, taking care to stand behind and not shade or bounce light from your shirt into it, the POINT IT AT THE CAMERA.
Out of the box that will not produce a correct exposure most of the time because your camera ISO when you set it really isn't 100, its 90, 110, or 120. The meter on the other hand is ISO 100 because its calibrated at the factory to be. So you need to COMPENSATE the meter to the camera:
1) Set up a target with light and dark areas with texture. Take a reading with the meter (e.g. f/8) then shoot 7 frames from -1 stop to +1 stop (e.g., f/5.6 to f/11)
2) Open the RAW files an look at the highlghts. Find the one with the best highlight detail. Odds are it will not be the one shot at the meter reading.
3) Calculate the difference in f/stops.
4) Enter the difference into the hand held meter using the compensation feature. On a L-358 hold down both ISO buttons at the same time and turn the wheel.
Lets say the meter read f/8 but you find the f/9 file looks better in the textured highlights. A compensation of .3 stops would be entered into the meter. Now the meter will read f/9 not f/8 and will correctly expose the highlights next time you use it.
For ratios the meter dome (if Sekonic) is retracted. Place the meter where the light hits the area you want correctly exposed and POINT IT AT THE LIGHT. Key and fill would be measure from where the front of the face will be. The hairlight hits is closest to the top of the head so that's where the meter should be placed, with the reading set relative to the key light creating the highlights on the face.
The background should be metered with the meter on the background on the spot that appears behind the head. Imagine the subject is holding a piece of the background next to their face. The exposure meter reading (DOME UP - POINTED AT CAMERA) would reproduce it accurately (once your meter is calibrated). When you move it behind the subject if you want it the same tone it really is then you'd set it to the same meter reading. Or it can be made lighter or darker.
Nuking a white background is an easy way to get it even but it will make any thing white on the subject dark by comparison. The brain tunes out the background immediately if its smooth, and its better perceptually to have "paper" white or slightly darker so the white highlights will look brighter.
I set my foreground and rim light accent light first, keeping texture in the brightest non-specular highlights (rim light in this case)
The tone of the background is being created by the key and fill aimed at it in a centered pattern. Foreground set I turn on the background lights and bring up the background until its white, but not as white as the rim lit highlights to retain the illusion of 3D the rim lights create..
If I had nuked the background, the rim light and edge definition would disappear.
The most important lesson here is that while the meter is useful tool it is more important the the lighting look good perceptually and create the illusion of 3D.
Right now you are trying to build a house with a bad bblueprint without a real clue what a house should look like. Set the lights so they achieve the desired goals -- fit the scene to the sensor and flatter the face -- then meter to discover the numerical relationship between the lights. Then next time you will have a blueprint for building the lighting pattern and understand how it should look and why it works. 
Chuck
|
