Magnification is the only number that matters close up
Magnification is the ratio of the image on the sensor to the object in front of the lens. At 1:1, a bee that is 12 mm long is projected 12 mm long onto the sensor, which fills a third of the width of a full frame. Everything else in close-up work follows from that ratio: how much subject fits, how much extension the lens needs, how much light is lost, and how thin the depth of field is.
Under the thin-lens model, reaching magnification m requires the lens to sit a distance f(1+m) from the sensor, which is f×m further out than its infinity position. That extra f×m is the extension, and it can come from the lens focusing mount, from tubes, or from a bellows — the optics do not care which. A 100 mm lens needs 100 mm of extension for 1:1 and 25 mm for 0.25x, which is why a 25 mm tube does much more for a short lens than for a long one. On a 50 mm lens that same tube buys half life size; on a 200 mm lens it buys an eighth.
The subject sits f(1+m)/m from the front principal plane. At 1:1 on a 100 mm lens that is 200 mm, and at 2x it is 150 mm. Note the direction: more magnification means less room, always, for a given lens. The only way to get magnification and working room together is a longer lens, and that is the entire design brief of a 150 or 180 mm macro.
Why depth of field ignores focal length here
The close-up depth of field expression is roughly 2Nc(1 + m/P)/m², where N is the set aperture, c is the circle of confusion, m is the magnification and P is the pupil magnification. There is no f in it. That is not an approximation or a simplification — at a fixed magnification, the blur circle produced by a point at a given distance from the plane of focus depends on the aperture and the magnification, and the focal length cancels.
Put concretely, at 1:1 and f/8 on full frame with the conventional circle of confusion, the total depth is 0.96 mm, and it is 0.96 mm on a 50 mm lens, a 100 mm lens and a 200 mm lens alike. This is the single most useful thing to know about macro work, because it kills the instinct to reach for a different lens when the depth is not enough. The lever is the aperture, and after that the lever is stacking.
| Magnification | Subject width on full frame | Depth at f/8 | Effective aperture at f/8 |
|---|---|---|---|
| 0.25x | 144 mm | about 9.6 mm | f/10 |
| 0.5x | 72 mm | about 2.9 mm | f/12 |
| 1x | 36 mm | about 0.96 mm | f/16 |
| 2x | 18 mm | about 0.36 mm | f/24 |
The depth falls roughly with the square of the magnification, which is why the jump from half life size to life size feels so much more brutal than the jump from a quarter to a half. Those figures use the 0.03 mm convention; anyone judging a macro frame at full size on a monitor should halve it, and halve every depth in the table with it.
The bellows factor, and when it is already paid
Extending a lens spreads the same cone of light over a larger circle, so the illumination at the sensor drops. The effective aperture becomes N(1 + m/P), and the light lost is two stops at 1:1 with a symmetric lens. This is not a lens defect and it is not avoidable; it is the same geometry that makes the image bigger.
Whether you have to do anything about it depends on how you meter. Any camera metering through the lens is already seeing the dimmed image and has already accounted for it, so the exposure it suggests is correct and the only surprise is that f/8 behaves like f/16 for diffraction purposes. Working with a handheld meter, a flash in manual, or a guide number, the correction is yours to apply, and two stops at 1:1 is a large one to forget.
The diffraction consequence is the part that catches people. Set f/16 at 1:1 and the effective aperture is f/32, which is deep into the range where diffraction, not focus, sets the resolution. The frame will have more of the subject acceptably sharp and less detail everywhere. Where the subject holds still, several frames at f/8 blended together beat one frame at f/22 comfortably.
Pupil magnification, and the honest limits of all this
The P term in the formulas is the ratio of the exit pupil to the entrance pupil. It is close to 1 for a symmetric design, above 1 for many telephoto layouts and below 1 for retrofocus wide angles, and it changes both the light loss and the depth. Nobody can guess it from the outside, and the only trustworthy source is the lens data. Leaving it at 1 is the right default; the error it introduces is usually smaller than the error in your circle of confusion choice.
Two other limits are worth naming. Modern macro lenses focus internally, which shortens their effective focal length as they focus closer, so the published minimum working distance at 1:1 is often well under what f(1+m)/m suggests — the manufacturer figure is the one to trust. And at magnifications much above 2x, ordinary lenses used in the ordinary orientation start performing poorly enough that reversing them or using a dedicated objective is the usual route, at which point the thin-lens arithmetic here is a rough guide at best.
Beyond the close-up range, the standard formulas take over: near limit, far limit and hyperfocal distance from focal length, aperture and distance, which is what the depth of field calculator does. The crossover is roughly where the subject distance stops being a small multiple of the focal length.
Questions people ask
How much extension do I need for 1:1 macro?
Extension equal to the focal length. A 50 mm lens needs 50 mm of extension beyond its infinity position, a 100 mm lens needs 100 mm. That is why a fixed set of tubes behaves so differently on different lenses: 36 mm of tube gives about 0.72x on a 50 mm lens and about 0.18x on a 200 mm lens. Some of the extension may already be in the lens focusing mount, and a dedicated macro lens that reaches 1:1 on its own has the whole amount built in.
Does a longer macro lens give more depth of field?
No. At the same magnification and the same f-number the depth of field is the same, because the close-up depth expression contains the magnification and the aperture and does not contain the focal length. What a longer lens gives is working distance: at 1:1 a 100 mm lens puts the subject about 200 mm from the front principal plane while a 200 mm lens puts it about 400 mm away. That extra room is what keeps you from shading the subject, frightening an insect, or being unable to fit a light in.
Why is my macro shot darker than the meter suggested?
Because extending the lens spreads the light over a bigger circle. At 1:1 the effective aperture is two stops smaller than the number on the barrel, so a lens set to f/8 behaves like f/16. A camera metering through the lens already sees this and compensates automatically. A handheld meter, a manually set flash, or a guide number calculation does not, and the correction has to be applied by hand. The calculator reports both the set and the effective aperture so the gap is visible.
What magnification do I need to fill the frame with a coin?
Divide the sensor width by the subject width. A quarter is about 24 mm across; on a full frame sensor 36 mm wide that is a magnification of 36 divided by 24, so 1.5x, and on an APS-C sensor 23.5 mm wide it is close to 1:1. This is the third input mode on the calculator — give it the subject width in inches and it returns the magnification, the extension and everything that follows. It is also the reason a smaller sensor reaches a given subject size at lower magnification, and therefore with more depth of field and less light loss.
Should I use extension tubes, a close-up lens or a real macro lens?
Tubes cost light and nothing else optically, since there is no glass in them, but they remove infinity focus and the amount of magnification depends entirely on the lens they are behind. A close-up dioptre screws on the front, costs no light and keeps autofocus, but it adds an element in front of a lens that was not designed for it, and the corner performance shows. A dedicated macro lens is corrected for close work at close distances and holds up flat-field, which is what matters when the subject is a stamp or a circuit board. For occasional use tubes are the cheapest honest answer; for repeated work the dedicated lens earns itself back in frames that do not need apologising for.