Matrix metering (called evaluative metering by Canon, multi-pattern metering by Pentax, ESP metering by Olympus, and multi metering by Sony) is the default exposure metering mode on essentially every modern camera. The system divides the frame into zones, measures brightness across each zone independently, and then uses pattern-recognition algorithms to assign weighted importance based on what kind of scene it appears to be. The result is a single exposure value that handles most real-world lighting conditions correctly without user intervention.
The mode was introduced by Nikon in the 1983 FA, which used a five-segment sensor and was the first SLR to incorporate scene-recognition logic into exposure metering. Canon followed with evaluative metering in the EOS system in 1987. Modern implementations have orders of magnitude more zones: Nikon’s Matrix III in current bodies uses a 180,000-pixel RGB sensor for metering and scene analysis; Canon’s iFCL meter uses up to 384 zones combined with subject color and focus distance data. The increase in segmentation, paired with on-board databases of tens of thousands of reference scenes, lets the camera distinguish between a backlit portrait, a snowy landscape, a stage performance, and a sunset, applying different bias rules to each.
The system uses several inputs beyond raw brightness. The active autofocus point tells the meter where the subject likely is, so brightness near that point gets weighted more heavily. Color information helps identify likely scene types: large blue zones at the top suggest sky, large white zones suggest snow or beach, large warm zones at low elevation suggest sunset. The focus distance, available from the lens, tells the camera whether the scene is close-range (portrait) or far (landscape). All these inputs feed into the algorithm to bias exposure appropriately.
The reliability of matrix metering is its strength and its weakness. For typical scenes (daylight portraits, normal landscapes, indoor events with mixed but not extreme lighting), it produces correct exposure 90 to 95 percent of the time, freeing the photographer to focus on composition and timing. For atypical scenes (very high contrast, deliberate silhouettes, stage performances with single-source spotlighting), the algorithm can be fooled in ways that produce technically correct but creatively wrong exposures. A stage spotlight on a single performer against a black background will often be overexposed because the meter tries to bring the dark background up toward middle grey.
The correct response when matrix metering fails is either exposure compensation to bias the meter’s output, or switching to a more deterministic mode like center-weighted or spot metering. Many experienced photographers leave the camera on matrix metering as their default and dial in compensation when the scene calls for it, rather than switching modes mid-shoot. The matrix algorithm’s bias toward middle grey across the weighted average means that scenes dominated by white tones (snow, white walls) need positive compensation, and scenes dominated by black tones (stage backgrounds, dark interiors) need negative compensation, matching the same logic an experienced film photographer applied with a handheld incident meter.
Modern mirrorless cameras with electronic viewfinders have changed the metering experience because the EVF shows the metered exposure live before the shutter clicks. The photographer can see whether the camera’s matrix algorithm got it right and adjust compensation in real time, rather than chimping after the fact. Combined with eye detection AF that confirms subject placement, the matrix metering of a current mirrorless body is more accurate than ever, and the historical reasons for switching to spot or center-weighted have diminished. The mode remains the right default starting point for nearly all photography.