"Tracking AF is made up of two processes: first, "find (detect)" the subject, and then "focus on and continue to track (track)" the found subject." — description of Canon R1 focus system
Hmm. Two? Sounds like three to me:
- Detect
- Focus
- Follow (track)
Even in that simplification, moving subjects become a bit of a problem, in that between #1 and #2 and between #2 and #3 the subject may be moving to a different position.
This exposes one of the problems we currently have in describing the various autofocus systems and how they differ. We've had this problem pretty much since day one of autofocus. For instance, the Canon DSLRs in the oughts tended to detect quickly and begin moving focus but you could take photos prior to focus actually being achieved. During this same period, Nikon DSLRs tended to detect more slowly, but at the point where the first photo was taken, they were almost always in focus (current Nikon cameras have a Focus+Release attribute that can be set, too, which locks that behavior in).
One of the things I've been teaching for quite some time is that when a user takes more control of #1, then #2 and #3 tend to work better ;~). The amateur habit is to set Auto-area AF of some kind, and let the camera do all the work, starting with detecting. The pro habit is to tell the camera where to look for the subject they want focused and tracked, and that requires you control the focus position (and sometimes focus area size).
I see clear differences in cameras for #3, by the way. Current Nikon bodies tend to be spot on in tracking (up to the point where they aren't, which is usually user or subject caused). My experience with Sony bodies up through the A1—I've not had any significant experience yet with an A9 Mark III or just-announced A1 II—is that they're a little lax about focus plane during tracking: they can drift focus slightly forward and backward from the perfect point between images.
With relatively static subjects, all the brands currently do #2 quite well. The differences come in #1 (plus how the camera is set and handled), and sometimes with #3 if the subject is moving either slightly or dramatically fast and erratically (the in between all the systems handle about the same). Moreover, these things work different when the camera is set for video than they do for stills. Nikon these days tends to have the best-behaved video focus, while I find Canon and Sony to be twitchier.
"Meanwhile, I'm the same speed. I didn't actually need my computer to be faster." --Mike Johnston, The Online Photographer
This speaks to a point I keep making: not everyone needs to be at the bleeding edge of technology.
I've watched a lot of folk get excited about pre-release capture at 30 fps, for example. They claim that they need or will use this. However, other than some birders, I'm not seeing that in practice. I notice similar things with 8K video, Focus shift shooting, Pixel shift shooting, and a bunch of other recent features. These all seem to be "sometimes things" that don't get a lot of use by the person who insisted it be in their next camera. The latter two you can perform—with more complexity and time obviously—without having the feature built in. So people are often buying convenience for something they don't (often) do.
Mike is going to discover one small falsehood about his statement, though. For his Web surfing, writing, and a host of other things he does regularly, he's right, a faster computer doesn't make him faster. Where his new M4 Mac Mini will be faster in a way that could make him more productive would be in more sophisticated image handling. He did note how much faster he's ingesting images, for instance. He should also notice it in processing, as well.
So let me pontificate a bit about that last bit: there are slow thinkers and fast responders in image editing. The slow thinker doesn't really mind that it took Lightroom 10 seconds to build their pano because they need time to think about what it is they might want to do next. The fast responder wants that pano rendered stat because they want to start tinkering with it now. I'm amused when I see a slow thinker trying to figure out whether they need a Mac Studio M2 Max or just a Mac Mini M4. They don't need a Studio with the Max or Ultra chip, because they'll find themselves staring at the finished pano for quite some time before figuring out what to do next. On the flip side, I've seen fast responders just order the latest and greatest because they think it is always faster.
I just updated my Recommended Macs article again with some things that speak to that fast responder. For example, a Mac Studio M2 Ultra is indeed faster at most of the tough photographic processing tasks than even the Mac Mini M4 Pro maxed out. Why? Because there are more cores and the memory bandwidth used between them is faster. For Apple Silicon, the priority goes (1) better memory stats, (2) better M chip core stats. So if you're a fast responder, currently a maxed out Mac Studio is the fastest Apple desktop you can buy (it has both #1 and #2 ;~).
Still, it pays to note Mike's comment. Do you really need something to be faster? Most of you don't. I'm someone who's often ahead of the bleeding edge of tech (unreleased products in my possession), and I'm not sure I need everything that we're getting now. My portable is still a maxed out M1 Max 14", and I'm not seeing it sluggish in any way for things I do. In fact, these new Mac desktops are making me have to up my game as a fast responder, as sometimes I find myself still staring at the result thinking about what to do next.
"Sony Has the Problem Every Competitor Wants: Its Flagship Was Too Good to Follow" --Petapixel headline
The article goes on to spew "interesting" statements like a machine gun: "The first Sony A1 set a nearly impossible to beat standard," "...it's challenging to feel about it the same we did about its predecessor," and "...doesn't do anything surprising."
A lot of this article was essentially apologetic posturing. When the A1 first came out it was arguably the best all-around high-end camera you could buy. It had a number of small annoyances, but those could be dealt with by most users because of what the camera did deliver. Of course, the Z9 (and then the Z8) came out at lower prices effectively matching the A1, and the King of the Mountain contest suddenly started to become a congested Knights of the Plateau one.
This isn't the first time we've had this contest. It won't be the last.
Thing is, sometimes you get big steps forward in products, sometimes you get small ones. The A1 II progress is one of those small ones. It mostly sands off a number of rough edges (perhaps literally due to the change in body to the A9 Mark III one).
My thinking hasn't changed. If you're in the E-mount world, the ultimate camera to have is now the A1 II (it used to be the A1). I don't know a Sony pro who shouldn't have one of the two A1s. The problem, though, is that the A1 II isn't enough of a change to get an A1 user to upgrade, nor does the A1 II change anything in the Canon/Nikon/Sony pro camera dogfight.
But contrary to the implication of Petapixel's headline, other competitors have the same problem. The Z9 will be an extraordinarily difficult camera for Nikon to follow up. A Z9 II is likely to fall into the same category as the A1 II: a lot of small changes that make the best even better. Canon, at least, does have the opportunity to take their 24mp pro body into a higher pixel count, which would stop their pros from all buying R5 Mark II's to keep up with the pixel race.
For what it's worth, digicame-info, a Japanese "information" site, put the following headline interpretation on Petapixel's article: "The reason why Sony's A1 II is not getting the best reviews is because the A1 and A9 III had too big an impact." That's another way of saying "top cameras are already really good," but it seems to imply that every new camera needs to make huge impacts or else it's "not that interesting."
Progress is progress. Sometimes we take big steps forward, other times we inch forward. This does bring up a dilemma for the purchaser (as well as Sony), at least temporarily. Right now you can pick up a new A1 for US$5500, or wait for the newer A1 II at US$6500. I think the more important thing to consider right now is whether any of the improvements in the newer model justify paying US$1000 more at the moment. I'd tend to say no. So there's a temporary situation where you can get "nearly the same" for significantly less. Thus Sony's real problem is that pre-orders for the A1 II are likely to be low until the A1 supply is exhausted. (Update: I appear to be at least partially wrong. Sony is saying they'll have trouble filling pre-orders, and you may expect delays. I'm not sure whether that's more demand than expected, or whether it's just low supply.)
