>There are a number of things that can go wrong, determining
>what if any requires stringent testing. And I would caution
>you against quickly jumping on the back-focus band wagon. It
>could be the issue, but then again….

Well, I did avoid putting it at the top of the list But seriously, while I'm as tempted as any other relative neophyte to blame my equipment, I realize it's much more likely the fault lies elsewhere.

>First, Phase Detection AF works with acceptable accuracy, not
>exact precision. The whole capture looks relatively sharp in
>the main areas (especially if you use creative sharpening) and
>it’s only when pixel-peeping that you see the slight miss
>focus. If precise focus is required then live view or manual
>focus using zoomed live view would be a preferred method I
>would think. Although the results you received are not
>tolerable to you, it’s my opinion that it’s within tolerable
>acceptance of the Phase Detection AF system.

That is really interesting. I haven't really used Live View much for a combination of reasons, good and bad: I prefer the optical VF, especially on sunny days; I find focusing with the VF takes less time; it's my impression that "real photographers" prefer focusing with the optical VF over Live View. Now I have certainly used Live View to focus (manually and zoomed in) when shooting a stationary subject and using a tripod, but I've never really used it for more spontaneous people or action shots.

>It’s also been my experience that when investigating focus
>accuracy shorter focal lengths are prone to a wide range of
>focus variation when achieving focus lock. Shorter focal
>lengths have greater optical power which can be difficult for
>the Phase Detection AF system. When testing AF a while ago I
>noticed that when using shorter focal lengths that an exact
>repeatable focus distance was not possible, and depending
>which direction you started in you could achieve focus lock at
>varying distances even though the camera is tripod mounted and
>the subject is inert. So again, if exact focus is critical,
>manual focus would be the preferred method.

Stupid question perhaps, but if you (or any other experienced photographer) were taking a shot such as this -- handheld, subject in movement -- might you actually focus manually? I am interested in what photographers actually do in such situations. I have next to no experience (and consequently no skill) in focusing manually on moving subjects. And if you did focus this shot manually, would you use the VF or Live View?

>Also the AF illumination will only occur if the camera deems
>the ambient light insufficient, and my not occur with all
>focus points. You can test this by using the flash AF assist
>in a darkened room and using AF-S single point, move the focus
>point to different positions while half pressing the shutter.
>As Brian stated you should see a cross type pattern projected
>from the AF assist.

As I just posted above, it is possible I disabled AF-assist in-camera and forgot to re-enable it. I'll check that tonight.

>In short, it’s really difficult to draw any conclusions from
>your example.
>
>Hope this helps.

Helps loads. Thanks!