Originally Posted by michaelwj
This was a seperate rant unrelated to Mingís article...
It doesnít exactly double, but itís close enough (thereís a small angle approximation in there). Regarding the different linear changes when focal length changes, the f in f-stop is normalising for focal length. We need to un normalise it for depth of field. So, stopping down to f/8 on a full frame lens is the same linear aperture size as stopping down to f/4 on m43 for the same angle of view. Depth of field is related to the actual aperture size.
If I can try to express myself more clearly after a few coffees on a Saturday morning...
For a given field of view and depth of field (f/4 12mm m43; f/8 24mm FF) the airy disc covers the same % of the recording medium. If we have the same number of pixels then diffraction becomes visible at the same depth of field and angle of view regardless of the format. The argument that a smaller format allows for greater depth of field is false (except when you run out of stops at the largest f number, which is a lens problem, not a format problem)
Adding more pixels however allows you to sample the airy disc at a higher spatial frequency, and in the end looks better.