   |
Clay --
Are you familiar with the incident metering method in general? In any given light condition the meter's white dome averages the illumination and provides sufficient exposure to render a white object as 'toned white' (roughly zone VIII). At the same time it will render a black object as 'near black' (roughly zone II-III). Since any 3-D subject must necessarily include some areas of shade, those shade areas constitute a second, less intense light condition, so to include dark detail for full-scale rendering the 'shadow reading' should be taken there.
In a back-lit situation the meter position will depend on your interpretation of the scene. If you point the cell toward the camera (away from the light source) it will be in shade so it will provide usable exposure for both black and white shaded objects (assuming appropriate development) and that means that the light source itself will be grossly overexposed, and the shadow side of the subject will be rendered, on average, in relatively light gray.
On the other extreme, if you step out of the shade and point the cell at the light source (away from the camera) the meter will attempt to render a white object in that direct light as white. Consequently, the subject (in shade) will be underexposed and appear as a near-silhouette.
Splitting the difference — pointing the dome sideways so that half is in full light and half in shade — will render the subject dark, but detailed, while providing a sort of halo effect around it with some possible vestiges of texture in the less extreme highlights.
Now, obviously this is an over-simplification, and it may or may not apply completely to the back-lit situation you have in mind, but the trend should be clear; as you turn the meter toward or away from the light source, you're effectively shifting the interpretation of the subject from unnaturally light and low contrast, with totally burned out background, to silhouette rendering with bright, but possibly textured background. Intermediate meter angles will produce intermediate results.
Does this make sense? It's a lot easier to demonstrate than it is to describe!
|
|
|
|
|
|
|
