Hey, what happened to Day 4 and 5? Well, in order: COVID testing, flight, customs, flight, COVID testing, customs, dinner with friends.
Today it was another (short) flight on a Cessna Commander to be dropped off at the Chitabe Airstrip, where it was time to jump in a Land Cruiser and see what the Z9 can actually do. Within minutes, I was at a Kudu kill site that had just happened and photographing the leopard that did the deed. Oh, and since it’s the rainy season here, it started to rain.
I hadn’t updated my FTZ firmware yet, so I started with the 70-200mm f/2.8 and a 2x teleconverter mounted to it on the Z9. I’m fast gaining more respect for that combo (the 70-200mm f/2.8 and 1.4x are excellent).
First image sequence? Turn around and photograph the Commander taking off.
Didn’t have time to think about things, so the shutter speed is 1/4000 here and that freezes the propellers (normally you’d want something down around 1/250 so as to get some blur to them). As it turns out, this is a reasonably good test of rolling shutter effect, if any. As you can see, none: the propellers look correct. As Nikon has said, it appears that you can treat the Z9’s electronic shutter pretty much the same as the mechancial shutters you’ve been using. Obviously, when final cameras appear, I’ll drill down into the details. However, in the initial quick-and-dirty tests like this photo turned out to be, I’m not seeing anything I’m concerned about with the all-electronic shutter.
Next up, here are the first true wildlife results from my pre-production Z9:
Yes, I saw that all of you raised your hand wanting to ask questions. How many of you want to know about focus? Thought so.
I decided to do my testing backwards from my usual regimen. Instead of starting with Single-point AF and AF-S, I started at the other end of the progression I usually recommend people work through: I began with Auto-area AF and AF-C with all subjects detection enabled. That plane image? Auto-area AF and AF-C with all subjects detection enabled. The leopard photos? Same thing.
The reason why I started this way is I can almost guarantee that this is the way you’ll initially begin using your Z9. After all, Nikon has marketed the heck out of the focus detection system, and it’s one of the reasons you probably bought the camera. Can it really handle all the heavy lifting?
Within the first few hours of safari my conclusion is that the focus is everything that Nikon claims it is, and it’s nothing like what you’re expecting. That’s going to take a lot of explaining, which I’ll try to do over the next two days.
Let’s start with the Cessna.
Nikon claims the focus system will pick up a plane and follow it. Yes, it did. I intentionally started the sequence with the plane out of the frame and letting it fly through my composition. I could bore you with the entire sequence, but it would show that the Z9 quickly found the object and followed it. You’ll see that in the viewfinder (foreshadowing: much more on that later). Score one for marketing. Oh, wait, engineering ;~).
The leopard was somewhat of a different story. Nikon makes the claim that the camera recognizes body, head, and eye, often in that order, of animals. I believe that’s driven by a great deal of shape detection in the Machine Learning behind the focus system. Why? Because shapes don’t account for spots, which are, of course, more shapes. In other words, shapes on a shape.
Much of the time the Z9 did what you’d want it to: recognize a leopard, find the eyes, and put focus there.
This particular leopard was nervous. It knew it couldn’t hold a kudu kill very long because it couldn’t drag that size animal up a tree. Thus, it ate, then went checking around for other predators, repeat. That led to a great deal of posing in a great number of positions, situations, and lighting.
Sometimes the spots threw the focus system off course. On this particular animal, I saw the eye detection jump to the ears and back a number of times. Something about the ear pattern in particular light—mostly shaded light—had the camera seeing an “eye” up on the ear. More often than not, the focus was put on the eye, but since we were moving with the cat in and out of light (and off and on rain), I could see the camera back off and struggle at times to do the “perfect thing.” (Foreshadowing: there was another animal where this was worse.)
Now you might think that this is a bad thing and I’m about to go all Angry Photographer on Nikon.
Nope. For a number of reasons.
First, the thing I immediately noticed is that I got a strong sense of what the camera was doing and how well it was doing it. Good light and normal animal, the focus tends to stay more locked on the smallest thing it detected (again: body, head, eye). If the camera was a slight bit indecisive, as in the eye/ear flopping, that was a signal for me to step in and do something to help it. I’m not going to get into what that might be yet, but the Z9 is armed with features that might help. We’ll get to those as we roll through my camera use.
Another think I noted is that if the focus was locked onto an eye and the animal walked behind brush that might obscure it, the default Tracking Lock-On setting was enough to hold focus briefly as the animal continued moving. However, if it stopped behind that trig, the focus system would eventually want to move. The “stickiness” of the Z9 was more than I’m used to on the Z’s, and I’ll need to spend more time figuring out how to tune that to best results.
As you’ll note from some of the details below, the camera isn’t perfect. You need to be prepared with another tactic to get focus, which I’ll get to tomorrow. Fortunately, Nikon has given us a number of ways to override or change the focus system.
Animals: ostrich, leopard, impala, eagle, kudu, warthog, giraffe
Ostrich: mostly detected body at distances they would allow; when facing and closer, did detect head.
Leopard: mostly detected eye, even at long distances, but sometimes would see something erroneous as eye temporarily. Locked onto body or head easily.
Impala: detected eye, even through thick bush.
Eagle: detected eye even at great distance.
Kudu: detected face, then eye.
Warthog: sometimes detected eye, though the camera sometimes got confused by the tusk, for some reason.
Giraffe: had a difficult time with the giraffe unless the face was towards the camera.
