Multi-pattern metering, also widely referred to as matrix metering or evaluative metering depending on the camera manufacturer, is an advanced through the lens exposure metering system that divides the image frame into multiple discrete zones or segments and takes simultaneous brightness readings from each zone independently, before feeding all of these readings into a sophisticated computational algorithm that analyses the overall pattern of light and shade across the frame and calculates an exposure value designed to produce a well balanced result across the full tonal range of the scene.
The fundamental advantage of multi-pattern metering over simpler metering approaches such as centre weighted or spot metering lies in its ability to interpret the lighting situation across the entire frame rather than basing its exposure calculation on the brightness of a single area. By comparing the relative brightness values of multiple zones simultaneously, the metering algorithm can identify common lighting scenarios - such as a bright sky above a darker foreground, a backlit subject surrounded by bright background illumination, or a predominantly dark scene with bright highlights - and apply compensatory adjustments to the calculated exposure that prevent the kind of systematic exposure errors that simpler metering systems are prone to in these situations.
Backlit subjects represent one of the classic challenges for exposure metering, and multi-pattern metering handles them particularly effectively. When a subject is lit from behind, a simple centre weighted meter reading the overall brightness of the scene will be strongly influenced by the bright background light and calculate an exposure that renders the background correctly but leaves the subject underexposed and silhouetted. A multi-pattern metering system, by contrast, can recognise the characteristic pattern of a bright surround with a darker central subject, identify this as a probable backlit situation, and apply an exposure increase to ensure the subject is adequately exposed, even at the cost of some highlight overexposure in the brighter background areas.
The sophistication and accuracy of multi-pattern metering systems varies considerably between different camera models and generations. The most basic implementations divide the frame into just two or three metering zones, providing a modest improvement over simple centre weighted metering in handling non-average lighting situations. More advanced systems use eight, sixteen, or even dozens of metering segments, and the most sophisticated implementations found in current professional cameras may incorporate hundreds of metering zones coupled to an extensive database of analysed photographic scenes against which the current reading is compared to identify the most probable lighting scenario and select the most appropriate exposure response.
Many advanced multi-pattern metering systems also integrate with the camera's autofocus system, using information about which focus point is active and therefore where the intended subject is located within the frame to weight the exposure calculation towards the tonal values in the vicinity of the focused subject. This coupling of metering and focusing information allows the camera to prioritise the correct exposure of the actual subject rather than the overall scene average, significantly improving exposure accuracy in situations where the subject occupies a relatively small portion of the frame or is positioned against a background of very different brightness.