Why focus stack?

Why Focus Stack, why not simply stop down the aperture? 

You soon realise this is not an option when you start to shoot at extreme magnification - at 10:1 and more, even stopping down to F22 won't give you enough DOF for anything beyond a perfectly flat subject - three dimensional subjects like insect faces are way too deep for one shot to get everything in focus. Also microscope objectives used to shoot at this level of magnification typically do not even have aperture control, they are fixed 'wide open' (and at apertures of around f1.5 or less you can imagine the depth of field of each shot is wafer thin)

The main reason though is that at real macro magnification (Anything much beyond 1:1) stopping down a lens will soon start to cause diffraction. The more magnification, the sooner this starts to degrade the resolution of your photographs.

Here's an example, even resized down for web, the diffraction blur is obvious in this shot of a tortoise beetle, shot at F11 and around 3-4:1 magnification. I'm not talking about the lower part of the subject where the rear of the beetle is going out of focus, that is simply demonstrating that F11 does not give enough DOF to get the whole subject in focus, I'm talking about the general resolution of the "in focus" section of the image.

 

If the loss of resolution isn't obvious in that shot alone, compare the detail to the following stack of the same bug (admittedly the other side of it), a composite of 47 images shot at f4.0

The reason diffraction starts to become a problem is that the aperture you set on the lens actually appears to become smaller from the camera's point of view as you add extension between the camera and the lens. This is known as the effective aperture.
The formula to calculate the effective aperture is:

Effective Aperture = Marked Aperture x (Focal Length + Total Extension) / Focal Length

As an example, with a 20mm bellows lens and 200mm of bellows extension for 10:1 magnification, the lens is set to f8.0 then:

Effective Aperture = 8 x (20 + 200) / 20 = f88

An effective aperture of f88 is very small, and is enough to cause diffraction on any modern camera (any camera with smaller than medium format sensor).

If the aperture on this lens can open up to f2.0 then the numbers are quite different:-

Effective Aperture = 2 x (20 + 200) / 20 = f22.

An aperture of f22 is still just into diffraction territory on smaller sensor DSLRs like my Olympus, and APS sized sensors, but only to a small degree. This lens would be quite usable for stacking on a DSLR at a set aperture f2.0, but set to f8.0 will produce a seriously degraded image. (At this extension/magnification).


(* This is a slight re-write of an old article, I am currently reorganizing my articles into a simple series of blog posts)