Today was my last full day in Chitabe, and it turned out to be eventful. Much of the day we spent trying to keep up with a pack of 15 wild dogs. We found them early on resting (and occasionally playing). But the variety of poses, lighting situations, and motion allowed me to assess how the Z9 was doing pretty thoroughly. In a word, excellent.
I know that a lot of you expect—partly from the marketing statements of both Sony and Nikon—that autofocus will hit "perfect" with the right camera. That's no closer to the truth today than it ever was. I've been spending these first few days using the Z9 much like most of you will: in Auto-area AF with subject detection set to Animals. It's the simplest way to get "really good" focus performance. Okay, maybe "really excellent."
This freedom-from-controlling-focus enables you to pay more attention to the composition and your camera handling. But I'll caution you to think you're going to go out with any camera, set focus to some "all auto" mode, and get 100% keepers. It just doesn't work that way.
What many of you are going to find is that you get more keepers than ever with a Z9 set in one of the auto focus modes where the camera does the heavy lifting. But you can do better, and I'll try to describe some of the ways you can do that coming up. What I'd say to most new-to-Z9 owners is, yes, play with the Auto-area AF mode and subject detection, and concentrate on getting your camera handling and composition nailed as you do. I've already noted one prominent YouTuber—who's not a wildlife photographer by any stretch of the imagination—trying to show what the camera is doing when he was too far from the subject (or not enough lens), and had poor camera handling (small subject careening all over the frame). The Z9 does indeed do quite okay in that scenario, but it can do better. Far better, as you'll probably see at some point in this blog (foreshadowing: I got one shot I've tried to get many times in the past, but failed at, with every camera I've used).
I should probably point another thing out that's being described incorrectly on the Internet. You'll see comments such as "follows subjects to the edge of the frame and back."
Not exactly. You'll note four red "corner" markers in Auto-area AF. Within the area defined by them, the camera is near flawless (assuming it correctly detected a subject). As you move outside that area—which is about one Single point box worth of area on each side—things get a little more complicated. Depending upon what is being tracked by the focus system, you may or may not lose detection briefly as something gets to the frame edge. Again, note my comment on improving your camera handling, above. Your subject really shouldn't be getting out into this no-man's focus land if you're holding the camera well and following your subject well.
In the above photo, note that the front wild dog's eyes are above that no-man's focus land: the camera had no problem at all as long as I held to that. When the front dog was partially cut off and the eyes went "out of the box", the camera has to rethink what it's doing for focus. Sometimes it found the next dog, sometimes it struggled a bit to figure out what it wanted to do in this situation.
3D-tracking AF does a little better at moving all the way to the edge for focus, but be careful about letting the area being tracked leave the frame. If it does, it may or may not recognize it when it comes back in frame, seemingly dependent partially upon how long it was out of frame.
Finally: As I watch people respond to the blog, I'm noticing people quoting things I didn't say. For instance "Nikon’s eye-AF is a little less reliable than Sony or Canon’s". The problem with spontaneously blogging in the wild is that I don't always have the time to get my wording so crystal clear that it can't be mis-interpreted. I'll have much more to say about this in my Zoom Presentation on Monday, and subsequently in my review and book for the Z9. But let me be perfectly clear here: neither Sony nor Nikon's all-automatic focus modes are without issues. Again, I'll speak to that on Monday, but I actually prefer the Z9 "lies" to the A1 "lies" in the focus system. I also find the Z9 easier to "correct" (override) in a number of situations. That said, both Sony and Nikon are using Machine Learning (ML) algorithms, so I expect both to improve over time as the ML gets tuned.
New Animals: wild dogs
Wild dogs: for the most part, no issues. On a few of the 15 dogs we were tracking, the focus system sometimes had a problem with a particular spot pattern.
Note that both the wild dogs tend to have dark eyes on a dark section of face. One thing I've noticed with the Sony A1 is that dark eye on dark head (bird, other animal) sometimes confuses it. That doesn't seem to be the case with the Z9.
