Sensor size explained: what it changes, and what it does not

Sensor size changes the field of view, the depth of field and the light the whole frame collects — all three, by the same factor. Most explanations get at least one of them wrong.

The rear of a lens, showing its mount

The actual sizes

A sensor size is a physical dimension, not a grade. Here are the ones this catalogue contains, measured against the 36 by 24 mm full frame reference.

medium format44 × 33 mmfull frame36 × 24 mmAPS-C23.5 × 15.6 mmMicro Four Thirds17.3 × 13 mm1-inch13.2 × 8.8 mmdrawn to scale
Nested to scale, from 44 by 33 mm medium format down to 13.2 by 8.8 mm. Crop factor is the ratio of the diagonals; the light a frame collects follows the ratio of the areas, which is its square.
FormatSensor sizeCrop factorLight vs full frame
Medium format44 x 33 mm0.79xAbout three-quarters of a stop more
Full frame36 x 24 mm1.0xThe reference
APS-CAbout 23.5 x 15.6 mm1.5x, or 1.6x at CanonAbout 1.25 stops less
Micro Four Thirds17.3 x 13 mm2.0xAbout 2 stops less
One inch13.2 x 8.8 mm2.7xAbout 3 stops less
Sensor sizes, crop factors and light collected across the whole frame

Two things there are worth pinning down. The crop factor is the ratio of the diagonals: full frame is 43.3 mm corner to corner, Micro Four Thirds 21.6, hence exactly 2.0. The light column is the ratio of the areas, which is that figure squared — halve the diagonal and you quarter the area, two stops.

Note also that the name one inch describes nothing: that sensor is 15.9 mm across the diagonal. It is a leftover from the outside diameter of video camera tubes. Never do arithmetic with it.

The counts here: 23 full frame, 22 APS-C, 7 Micro Four Thirds, 7 small-sensor and 2 medium format. The cheapest of each is €1,500, 600, 700, 450 and 5,500 respectively.

The crop factor applies to three things, not one

Almost everyone gets the first of these right and at least one of the other two wrong. Multiply by the crop factor and you convert all three at once.

  1. Field of view. A 50 mm lens on a 1.5x APS-C body frames like a 75 mm on full frame. Its focal length has not changed; the smaller sensor simply records a smaller part of the image circle, so the view is narrower.
  2. Depth of field. At matched framing from the same distance, that 50 mm at f/2 on APS-C gives the depth of field of a 75 mm at f/3 on full frame. The f-number multiplies by the crop factor exactly as the focal length does.
  3. Total light across the frame, which is what sets noise. f/2 at a given shutter speed puts the same brightness on every sensor — which is why the meter reading transfers unchanged — but the smaller one catches it over a smaller area. Fewer photons in total, so more noise once both are enlarged to the same size.

Written out: APS-C at 50 mm, f/2, 1/100 s, ISO 400 gives the same framing, depth of field and total light as full frame at 75 mm, f/3, 1/100 s, ISO 900. Those two files should look alike, noise included. That is the whole of equivalence.

What does not scale is the exposure. f/1.8 is f/1.8 on any format. This is the source of nearly all the confusion: people observe correctly that the exposure is the same and conclude wrongly that the picture is. Brightness per square millimetre is equal; total light is not, and depth of field and noise follow the total.

Two other things do not scale. Diffraction bites at a wider aperture on a smaller sensor, because the same blur circle is enlarged more to reach the same print: if the practical limit is f/16 on full frame, it is around f/11 on APS-C and f/8 on Micro Four Thirds. And a smaller sensor does not make a lens sharper — it uses the middle of the image circle, usually its best part.

The same photograph, three ways

Take one picture: a head-and-shoulders portrait, the subject filling the frame identically, the same light, the same shutter speed. Here is what each format needs to make it.

FormatFocal lengthAperture for the same lookThe consequence
Full frame85 mmf/1.8The heaviest lens of the three, and the dearest body
APS-CAbout 56 mmf/1.2Glass that fast exists, but it is not cheap
Micro Four ThirdsAbout 42 mmf/0.9No such lens here — the catalogue bottoms out at f/1.2
One portrait, matched framing and matched rendering

That is the argument in one table. It is not that a smaller sensor cannot take the picture; it is that matching the rendering costs a faster lens, and fast lenses get large and expensive quickly. APS-C with an f/1.2 lens matches full frame at f/1.8. Micro Four Thirds with the same f/1.2 matches f/2.4, and nothing takes it further.

Run it the other way and the smaller sensor wins outright. To fill the frame with a distant bird, full frame needs 600 mm where Micro Four Thirds needs 300 — smaller, lighter, cheaper, focusing closer. Reach here runs to 914 mm full-frame-equivalent, and not all of it is on full frame lenses.

The honest caveat, by the same arithmetic: that 300 mm f/4 is a 600 mm f/8 equivalent, so you have traded light for weight and money. In good daylight most wildlife photographers would take that trade. At dusk they would not.

The cost is the system, not the body

The body price gap is the smallest part of the difference. Full frame starts at €1,500 here against 600 for APS-C and 700 for Micro Four Thirds — €800 to €900, once. The lenses are where it compounds, every time you buy one.

A full frame lens must project an image circle 43.3 mm across; an APS-C lens covers about 28, a Micro Four Thirds lens 21.6. More glass means more mass and more money for the same field of view at the same equivalent aperture. Lenses here run 116 g to 2,115 g and €220 to €3,100, and both the weight and the money sit at the full frame end.

So do the arithmetic on the whole kit. List the three lenses you would actually end up owning in each system, add their weights and their prices, and compare those two numbers. That settles more purchases than any sensor review, and it is the question how to choose a camera puts third of five.

Check the depth of the system too: Sony E has 15 lenses behind 12 bodies, Canon RF 13 behind 8, Nikon Z 11 behind 9, Fujifilm X 10 behind 7, Micro Four Thirds 9 behind 7, L-mount 7 behind 4. The camera finder filters on sensor size, but the mount is what you marry.

When the smaller sensor is the better buy

Three cases, and none of them is a consolation prize.

You are reach-limited

Wildlife and sport where you cannot get closer. The crop factor works for you: an APS-C body makes a 400 mm frame like a 600, and Micro Four Thirds like an 800, for no extra weight or money. In good light that is a straight win, which is why many bird photographers choose APS-C deliberately. The APS-C wildlife answers collect the bodies worth it.

The precise version, since this is where the internet errs in the other direction: what you gain is pixels on the subject, not magnification. If a full frame body has the same pixel density, cropping its file gives the identical picture.

You are weight-limited

Travel, hiking, anything done with a child on the other arm. Bodies here span 154 g to 1,415 g and the lens spread is wider. A kit that comes with you produces more photographs than a full frame kit that stays in the hotel. The APS-C travel answers start from that constraint.

You are budget-limited

The cheapest APS-C body here is €600 against 1,500 for full frame, and every lens for it costs less. Under about €1,500 in total, APS-C with two good lenses serves you better than full frame with one kit zoom. If budget binds, read do you need a camera at all first.

The medium format question

Two cameras here are medium format, at 5,500 and €8,500, on a sensor measuring 44 by 33 mm. Notice how modest that step is: the crop factor is 0.79, so the area advantage over full frame is about 1.7 times — three-quarters of a stop. One notch, not a different world.

The notch buys resolution and tonal gradation — the catalogue tops out at 102 MP — and a slightly shallower rendering at the same field of view. It costs autofocus speed, burst rate, weight and a great deal of money. The two bodies sit on Fujifilm GFX and Hasselblad XCD, one each, which tells you how narrow the lens choice is.

The rule is blunt. Medium format is for photographers who work slowly, print large and are paid for the result. If you are asking whether you need it, you do not — and the honest measure is that the gap between APS-C and full frame, about 1.25 stops, is larger than the one between full frame and 44 by 33.

Questions readers ask

Does a crop sensor make my lens longer?
No. A 50 mm lens is 50 mm on every body. The smaller sensor records a smaller portion of the image the lens projects, so the field of view is narrower — the same result as cropping the full frame picture. You gain detail on the subject only if the smaller sensor has a higher pixel density than the larger one.
Is full frame twice as good as APS-C?
It collects about 1.25 stops more light at matched framing and gives about 1.2 stops shallower depth of field at the same f-number and framing. That is a real difference after dark and a negligible one at noon. Set against it, the cheapest full frame body here is €1,500 and the cheapest APS-C is 600, and the lenses follow the same pattern.
Why do people say f/1.8 is f/1.8 on any sensor?
Because for exposure it is: the brightness falling on each square millimetre is identical, so a meter reading carries across unchanged. It stops being true the moment you compare depth of field or noise, which depend on the total light gathered over the whole frame. Both statements are correct; they answer different questions.
Should I buy Micro Four Thirds?
If weight or reach is your binding constraint, yes. It is about two stops behind full frame in light collected, and the lenses covering its smaller image circle are correspondingly smaller and cheaper. There are 7 bodies and 9 lenses for it here, from €700. If you shoot indoors after dark and want very shallow portraits, it is the wrong tool and no lens fixes it.