Exposure control comes from appreciating and then
recording the light in the scene as you want to render it. There are two key
values that determine your success in doing so: highlight and shadow.
Here is a recording that delivers a full range of tones. The highlights are the side of the shed wall and the tower. There is no need to "open" the shadows within the dark area of the shed as that seems natural. The details under the silos are "readable" with both tone and texture. This represents a range of brightness values that can be handled by sensors (and film) with no need for any special exposure techniques such as HDR or highlight biasing or compensation.
A principal highlight might be a freshly painted
picket fence in which you want to show the grain of the wood, or grains of sand
at the beach that you want to render in textural fashion. This highlight
needn't be white; it is just the brightest value you are recording with detail
or texture. For example, a bright yellow or orange can be the brightest part of
the image, and exposing it correctly means that it will not be washed out or
overexposed. (You can underexpose it slightly to add saturation, but you may underexpose darker parts of the scene when you do so.)
Spectral highlights are formed by light reflecting off a shiny surface and usually cannot be reproduced with detail; exposing for them "drives" the darker areas into deep underexposure. Thus, no detail other than the spectral highlight becomes visible. The setting sun off this beach created spectral highlights in the tidal area and in the ocean. This can be used for special effects, but no sensor or film can "hold" the light in the darker areas and they become lost through underexposure.
There are times when you "previsualize" a scene and want to use a bright highlight to make a bright area the main and usually only detail that is recorded. This "locks" the scene into one rendition. Here the bright formation was read by spot metering off the bright rock. This made the other areas much darker than they appeared in the original scene, an interpretive approach that is worth investigating.
These terms, (principal) highlight and (significant)
shadow are important to keep in mind when making exposures. While we do have a
wide recording range with digital cameras (less so on film), seeing the image
in terms of a tonal spread, and how that spread will record on film, is key to
making good images. And, applying HDR (multiple exposures that bracket plus and
minus and then combining them in software later) or working with highlight tone
priority algorithms on advanced camera (which in effect suppressing the
highlight value) certainly expands the “dynamic range,” or exposure latitude.
For example, say you are photographing a white car in
bright sunlight. You take a reading off the white hood and roof and get f/16;
you read the shadow cast by the car and get f/4--this presents no exposure
problem that cannot be resolved with minimal processing. However, let's say you also take a reading from the tires on the car in the
shadow area, and get a reading of f/2; this is clearly out of range. In this
situation do not expose for the brightest area, but back off a bit and expose
at f/11 or f/8.5 or, better yet, use highlight tone priority or HDR to get the
most out of the light values you record.
The question then becomes, is it important to record
detail in the tire tread in this photo, or can the tires record as pure black, with no visual information other than tone? If the answer is that the tires
may record as pure black, the significant shadow detail is the subject
information in the cast shadow of the car. If no, then other steps, already
mentioned, must be taken to record both the bright car and the tire tread.
The thing is—you can’t fight mother nature (in this
case the physics of the exposure system and the range the senor or film can
record). But keeping significant highlights and shadows in mind when you expose
will lead to much better results in both exposure and printing and help you see
with a photographic eye.