September 25th, 2011

First some practical rule I use: I call it the 10/10 rule:

At f/1, 10mm the hyperfocal distance is 10m.

To get to real situations:

  1. Divide 10m by the f-number you want to use (example f/5.6 gives ~1.8m)
  2. Compute the square of your focal length and divide by 100 (example 28mm * 28mm ~ 800, hence 8 )
  3. Multiply the two results (example 1.8m * 8 = 9.6m)

This rule is reasonably exact, you can use it as well to find the right aperture (the near-field is roughly half of the hyperfocal): Say, you have a 50mm lens and need everything from 10m to infinity in focus.

  1. Multiply the min focus distance by 2 (example: the hyperfocal is then 20m).
  2. Divide the lensfactor as computed in (2) above by the hyperfocal distance (example 50*50/100=25, 25/20 = 1.25m)
  3. Multiply by 10 to get the required aperture (example 1.25*10 ~ f/13)

You can shorten this by dividing the hyperfocal distance by 10 and execute only step 2, this gives numbers that are easier to compute.

This approximation is designed for “normal” APS-C cameras with a pixel-pitch of ~5µm. To be fair this makes sense down to f/8 - for smaller apertures the diffraction will get you unreasonably long hyperfocal distances. I recommend to fix the algorithm by setting the start distance of 10m to 8, 6 or 4m for f/11, f/16 or f/22.

What is hyperfocal distance?

Glad you ask, without much ado, hyperfocal distance is the focus distance from there even objects at infinity will still appear sharp. It is computed as

H = f*f/N/c + f

where f is the focal length, N the f-number and c the diameter of the circle of confusion (CoC)

Read the rest of this entry »