Out with the Sigma 100-400mm this afternoon under moderate overcast skies most of the time, the setup did a good bit better than yesterday.
Still, what I’m seeing reminded me of my first copy of the DN DN OS L-mount (as opposed to the SLR version which I also owned some time ago), in that AF was accurate at close and medium differences, but very inconsistent at long to infinity distance. Given how close it gets to critical accuracy when it misses, and that the near misses comprise both front-focus and back-focus, I’d be inclined to lay the blame on the SL2.
The good news is that the lens is capable of very good imagery when AF is accurate, though middling contrast saps fine details on some subject matter (see below). Images were encouraging enough that I’ll add the TC-1411 in the hope that a larger subject (in relative terms) will aid AF accuracy. Curiously, the Eye/Face/Body Detection setting enabled accurate AF whenever the subject was large enough to fill the focus square.
Spot focus on the SL2 seems to be same as it is on the current Canon bodies: definitely NOT a single point within a crosshair. In fact, the only camera I’ve ever owned that appeared to have a genuine single point for AF was the venerable Nikon D700. It also had the most consistently accurate AF of any camera I’ve ever used.
The trick with the others is to figure out how/where the “point” is applied and work from there. Maybe time to hit up my friend from the Holy Land of Wetzlar who worked in product management for the original SL and see if he can offer some insight. For now, it’s spray and pray.