Because a meter aims for an average brightness rather than for the picture you intended. Point it at a bright sky or a white wall and it darkens everything else to compensate, which is arithmetic working exactly as designed.
A meter measures light and proposes a middling result. It has no way of knowing what the subject is.
What the meter is actually doing
Averaging, over whatever area the mode covers, towards a brightness in the middle of the range. A scene dominated by something bright — sky, snow, a white wall — reads as too bright, so the camera shortens the exposure, and the subject in front of it goes dark.
The two ways out
Exposure compensation, which tells the camera to depart from its own proposal by a stated amount. It is the direct instrument and it is published on every body as a dial or a menu item.
Metering mode, which changes the area measured. Useful, and it makes the result depend on aim — which is why compensation is the more predictable of the two.
Where this bites hardest in this catalogue
At the long end of a variable zoom. A lens quoted f/3.5-6.3 is two and a half stops dimmer at full extension, so the shutter has already fallen further than most people expect before any compensation is applied. The second figure of the pair is the one that decides what remains possible.
What no figure describes
How a given body’s meter behaves in a particular scene. It is not published in a comparable form, which is why this site reports the published aperture pair and the sync limits rather than ranking meters.
Where this page and the maker’s own manual disagree, the manual decides. Every focal length, aperture, guide number, stop of stabilisation, load rating and sealing claim on this page is reported from the manufacturer’s own documentation, with its unit written as published, and named as such — none of it is measured here. An engraved figure is reported as the maker engraves it, and is never restated in another convention, because the two are not the same number. And the second statement repeated rather than assumed: a zoom quoted f/3.5-6.3 has two maximum apertures, not one. The first belongs to the short end of the range and the second to the long end, and the value moves between them as the barrel extends. Both ends are written here every time, because the second is the one that decides what the lens can still do as the light goes.
What people ask before they commit
Is spot metering the answer?
It is one answer, and it moves the problem rather than removing it: a spot reading is correct for the spot, so it depends entirely on where you put it. The more reliable habit is to read what the meter chose, decide whether the subject should be brighter or darker than average, and apply exposure compensation deliberately.
Does shooting raw fix it afterwards?
It gives room to lift a dark frame, and the room is not unlimited: detail that was never recorded cannot be recovered, and lifting shadows raises noise with them. Raw widens the margin for correction; it does not make the exposure decision for you.
Last reviewed 12 September 2026