Fundamentals June 15, 2026 4 min read

Circle of Confusion Explained: The Real Basis of Depth of Field

Every depth of field figure, on this site's calculator and everywhere else, rests on a number that almost nobody mentions: how big a blurred point of light can get before we agree to call it sharp. That number is the circle of confusion, and once you know where it comes from, the figures on a calculator stop looking like facts and start looking like a convention.

What it is

A lens focused at one distance renders a point at exactly that distance as a point. A point nearer or further lands before or after the sensor, so it appears as a small disc instead. That disc is the circle of confusion. Only one plane is ever perfectly sharp. Depth of field is the zone where the disc is small enough that, at a normal viewing size, you cannot tell it from a point. So it describes acceptable blur, not an objective boundary.

Where the standard value comes from

The usual rule is to take the sensor's diagonal and divide by 1500. On full frame, 43.3 mm divided by 1500 is 0.029 mm, which is often rounded to 0.03. The idea behind it is an 8x10 inch print, enlarged about 7.5 times from a full-frame image and viewed from 25 cm. On that print, a blur disc of 0.22 mm is the limit, which works out at about 3 arcminutes of viewing angle. Other formats follow from the same rule:

Format	Sensor diagonal	Circle of confusion
Full frame	43.3 mm	0.029 mm
APS-C	about 28 mm	about 0.019 mm
Micro Four Thirds	21.6 mm	about 0.014 mm
Medium format (44 x 33 mm)	55 mm	about 0.037 mm

Tools differ slightly, some use diagonal over 1730 or 1442, so a calculator may give figures a little different from another. That is the convention showing.

The standard is generous

Three arcminutes is a lenient test. A sharp eye can resolve about one arcminute, which on the same print corresponds to roughly 0.01 mm on a full-frame sensor, about a third of the standard. The effect on depth of field is large. With a 50 mm lens at f/2.8 and the subject 3 m away, the standard gives a zone 58 cm deep. With the one-arcminute value it is only 19 cm. For a landscape lens it is bigger: the hyperfocal distance of a 24 mm lens at f/11 is 1.84 m with the standard value, which puts the near limit at 0.9 m. At 0.01 mm it is 5.4 m, with a near limit of 2.7 m.

Why the same photo looks sharp small and soft large

Print or display the picture larger, or view it at 100 percent on a monitor, and every blur disc grows with it. A disc that was invisible on a small print becomes obvious. Pixels make the same point. If you decide that sharp means no blur larger than two pixels on a 45 MP full-frame sensor (4.4 micrometre pixels), the circle of confusion is about 0.009 mm. A 24 mm lens at f/8 then has a hyperfocal distance of 8.2 m, compared with 2.5 m by the standard. For a large print or close inspection, treat the near and far limits from any calculator as optimistic, and shoot with more margin.

Sensor size and the circle of confusion

A smaller sensor uses a smaller circle of confusion, because the image has to be enlarged more to reach the same size. That works against depth of field, not for it. Depth of field is proportional to the f-number times the circle of confusion times the distance squared, divided by the focal length squared. For the same framing from the same spot, an APS-C camera uses a lens 1.5 times shorter and a circle of confusion 1.5 times smaller. The shorter lens contributes a factor of 2.25 and the smaller circle takes back 1.5, so the net result is 1.5 times more depth of field. That is the crop factor, and it is why smaller sensors look deeper at the same f-number and framing. The full frame versus crop sensor post goes through it with examples.

Hyperfocal distance

The hyperfocal distance is the focal length squared divided by the f-number times the circle of confusion, plus the focal length. Halve the circle of confusion and you double the hyperfocal distance. That is why a stricter standard pushes the near limit out so far in the examples above, and why the hyperfocal distance method works best when you are generous with margin.

What to do with this

Use the calculator's figures as a good default for web use and ordinary prints. For large prints or critical work, assume the actual limits are tighter. Focus on the nearest thing that matters, not on the hyperfocal point, and stop down a little more than the figure says, keeping an eye on diffraction. And remember that sharpness fades gradually. There is no line where an image becomes soft, only a convention about where we stop caring.