Center-weighted averaging is a bit more “old fashioned” in that it is how light was read, for the most part, before advanced microprocessors got into the mix. The light is read from all parts of the viewfinder, with 70% of the light reading coming from the center of the frame and the remaining 30% of the calculation from the edges of the frame. It is called “averaging” because it takes in all the various brightness levels and then averages them to what is called a “middle gray” exposure reading. This is the basis for much of how metering and translating light value works, so is worth some consideration.
When the metering system receives light from the scene it attempts to average the exposure values so that the bright areas record as bright and the shadows as dark, in essence arranging the light values along a scale of light and dark. Let's say you are working with a scene where there is a range of values, from the bright white in clouds to the deep shadow of a valley. If you read the clouds alone with a spot meter they might read f/11. The edge of the valley might read f/8; the shadow area reads f/5.6. An averaged exposure would be f/8. This “places” the brightest clouds as a highlight, the edge of the formation as the middle value and the deep shadow as quite dark. Thus, exposing at around f/8 places the brightness values as they appear in the scene.
Because the pupil of the camera (the lens opening) is fixed at the moment of exposure, there is no leeway for adjusting to various levels of brightness within the frame. This means that one exposure time has to handle all the lights and darks in a scene, and try to get detail from them all. This is, as you can imagine, a delicate situation. How it is handled is to arrive at an exposure that allows in just enough light to bring detail into the dark areas (the shadows) and not get overwhelmed by the bright areas (the highlights.) This is usually an average of the two intensities of light. The average reading sorts out the lights and darks accordingly so some records as brighter than the average and some darker than the average, as it should be.
Although the exposure system is quite sophisticated you too have to do your part. In essence, the information you “feed” the brain of the meter is the information it acts upon. In exposure that means making a reading by pointing the camera and sometimes locking exposure values to get it right.
For example, let’s say light values EV 11, EV 7 and EV 9 exist in the scene. A change in 1 EV (exposure value) is one stop difference. The average of these three readings is EV 9, which we’ll call f/11 at 1/125 second. That’s the reading the meter will recommend and set for you. EV 9 then becomes the middle gray of the light to dark brightness value, or tonal scale of that image. In essence, the meter has read and set up a range of tones that will be recorded. With EV 9 as the middle gray, EV 11 records as a brighter value and EV 7 as a darker value.
If, however, the reading was made incorrectly and the middle gray was set at EV 7 (which you would get if you just read the shadow areas) then the EV 11 (quite bright) value would record as very bright, and result in overexposure. Conversely, if the middle gray were EV 11 (created if you just read the highlight area) then the EV 7 reading would become much darker than it appears to your eye. All the values work in lock step, so making a bright value a middle gray makes all the dark values darker (and perhaps underexposed, where no detail is seen) and making a dark value middle gray can cause all the bright values to become quite overexposed.
A sunset scene is a classic shot for center-weighted metering. The intent is not to get detail in the ground but to use it as a form that offsets the sky and defines the horizon line. The simplest approach to sunset shots where you want to have a rich sky is to use center-weighted metering, aim the camera at the sky (not the sun!), lock exposure and shoot. Use this method and you’ll never miss a dazzling sunset again.
This might seem quite confusing in the abstract, but working with the camera and making readings exclusively from certain brightness values in the scene, and observing results, will quickly show you how this system works. In fact, you can even make use of this knowledge to create very expressive exposures.
The key to center-weighted reading is to "bias the exposure" towards the highlight. In other words, point the camera towards the highlight area (including other areas as well) and make the reading from that area. This is especially true if the highlight area sits at the corner or out of the center of your framing.
Photo and text copyright George Schaub 2010. In this scene the camera was set on center-weighted metering pattern, pointed towards the brighter area in the upper right, and then exposure was locked and the image reframed to the compositon you see here. Matrix or evaluative would have undoubtedly overexposed the highlights in this scene.
