I mentioned last week that the Nikon Z cameras are pretty much all "middle." What exactly would bring one or more Z System cameras into the "upper" realm?
I now know some of what Nikon thinks the answer is (based upon information passed to me in confidence), but I'm not going to cheat. I've been contemplating that thought for awhile now and had already come up with my own list. Here it is:
- Body build. This takes a number of forms (pardon the pun). While there's nothing particularly wrong with the current Z6/Z7 II build, most of us have a few common requests. The first is simple: more direct controls and more customization. A D850, for example, has three buttons the Z7 II doesn't, and uses a dial for Release Mode instead of a button+dial. This is a bit of a no-brainer to accomplish: use the D850 dial cluster (four buttons plus dial) instead of a Mode dial, and make the Release Mode button into Fn3. I'd also vote for voice annotation and Ethernet from the D6. Some will ask for an integrated vertical grip body with the bigger EN-EL18 style battery, though I'm okay without that (as long as a vertical grip option is available). In terms of things that Nikon should copy from others, the swipe-able thumb stick of the Canon 1DX Mark III comes to mind, as does the hot shoe electrical connections that Olympus and Sony use to connect things like microphones without a cable. The controversial thing will be Rear LCD. In the most robust cameras (D6) it is fixed. In the video-oriented cameras it is fully articulating. In the daring design world (Pentax, Fujifilm) it is multi-way tilting. I'd just say that something a bit beyond the current limited tilt would be welcome. Finally, anything to improve weatherproofing that can be done, should be. Note that all of this is probably low-hanging fruit. Most everything already exists in Nikon's available engineering toolbox, both in the hardware and firmware sense.
- Missing/Incomplete Abilities. An example of an incomplete ability is focus shift shooting. That function needs to be rethought, as what the photographer wants to set is near point, far point, and acceptable focus blur. The camera then should calculate number of shots and perform them. But the other complex methods—multiple exposure, HDR, for instance—all have issues, too, as Nikon has continuously fiddled with these but never seemingly completed the set all of us would want (can we not create a NEF and keep NEFs anymore?). Conspicuous among the missing abilities is pixel shift shooting. Two forms of this exist in competitive products: the four shift "full RGB" set, and the fractional shift "more resolution" set. Nikon has neither and needs both now to be high end. There's more, though. Olympus' Live Composite function would be highly welcome. And why can't we stack astro shots in camera? Some engineer ought to be able to calculate how to offset shots correctly in a stack, as the earth doesn't move that erratically. The software to do astro stacking is licensable. This is EXPEED territory. That means that if the chip designers aren't coordinating with the photography-oriented product managers, the possible becomes impossible.
- We're tired of limited JPEG and NEF options. Nikon even gets into simplification here, leaving out things like Uncompressed NEF on some cameras now. Basically, I have only three useful options in current cameras: 8-bit JPEG and its compression, 12-bit Lossless NEF for action, and 14-bit Lossless NEF for base ISO landscapes, et.al. Why? (And why was TIFF removed from the Z6 II?) Two formats I'd love to see now: greater-than-8-bit HEIC, and 12-bit TIFF. One thing I note is that Nikon engineering keeps covering the same landscape, so to speak. It used to be that they were daring and explored other opportunities. Now? They seem to think they're done with having to add another image format. Technology will run over anyone that sits.
- Time for some Artificiality. If there's a data processing bit where Nikon seems to be behind and a high-end camera could use a boost, it's that broadly designated field of artificial intelligence (or machine learning, et.al.). Face and Eye recognition are rather crude processing algorithms that date quite far back now. I remember doing a form of face recognition on the original QuickCam (originally for exposure/color, later for focus). Olympus is into trains and planes. Canon is into wildlife. Nikon is into...well, faces and eyes still. Human, body, animal, action, near/far, trains/planes/automobiles, I can think of a ton of things I want "recognized." Artificially. With override. Once chips start building in AI/ML, this becomes another software problem, though often a trainable one. So train the engineers so they can train the system, Nikon.
- The View. There's nothing wrong with Nikon's EVF. But it's not state-of-the-art, and the high end is going to demand that. The good news is that the parts are available from their current supplier (Sony Semiconductor). The bad news is that we need faster refresh. I'm not just talking about 120 Hz versus 60 Hz. I'm talking about those autofocus sensors not keeping up with what camera has already done, the verification of tracking focus, and not having to put up with the slideshow effect at high frame rates. I'm hoping that the reason for the issues we do see in the initial Z's is mostly that the different groups didn't integrate enough. That the sensor, focus, and viewfinder groups were all set off on tasks, but never quite met enough to optimize between them. I'm hoping that not having enough bandwidth internally isn't the issue, because that's a big engineering task to fix.
- Flash. Nobody has done mirrorless flash justice. Here's a wild one for you: why can't you show Live View in the viewfinder, but with simulated flash adjustment as well? Oh, right, you'd have to have some CPU/GPU horsepower with the right smarts built in (see italic statement in previous bullet). But things really have fallen apart with flash overall. Flash control got buried in the menus, and then only with three current flashes. Flash-provided focus assist light is the wrong color and unusable. Radio wireless requires an expensive (and often out of stock) stub on the side of the camera that also blocks the HMDI slot. The flash options are a strangely worded set of Custom Settings, and some scattered other settings, like Auto ISO for flash. What happened to Bluetooth? Did the flash engineers never get introduced to it? Why don't we have any short duration LEDs, ones that could light up when you half press the shutter release, and shut down after the shot. And yes, the flashes themselves are ancient designs now. Unfortunately, this is global rethink territory. Nikon has gotten worse and worse at accessories, and flash is an accessory. If Nikon can't do this themselves, they need to find an ally who can do it for them (much like they did with the original Speedlights).
- Cloudification. SnapBridge was Nikon's response to my original smartphone connection idea. It's as if some of the ideas I presented to Nikon management took, but no one quite knew what they were for or how to flesh them out. The problem, of course, is that this is a software problem, and Nikon isn't good at software. And it certainly isn't good at keeping up with other company's software. Still, I'm more convinced than ever that our cameras need to communicate images, and do so without excess effort, with annotations/ratings/etc. that we control, and to places we want them to go. SnapBridge needs an innovative champion who is in touch with the constantly changing Silicon Valley (and increasingly Chinese) world of social media. Can we tag images with SnapBridge? Yes, if the tags you want are #NIKON and #SNAPBRIDGE. Can we get images to anyplace we want? Not really, though Apple's Share menu does a lot of heavy lifting if you're into manually sending things. Camera to Nikon Image Space to Anywhere should be simple and easy. This is a software problem that then needs some tools in the menu/playback system. The menu/playback side Nikon can do. The software side? So far, that's proven to be a real stretch for them, despite using outside parties to do some of the coding.
Hmm. You might have noticed I didn't mention image sensor specs ;~). I actually think that's a pretty controversial area where some companies are starting to stretch themselves into lots of expense for marginal returns. So let me step through a few reasons why I say that:
- Photons have revealed themselves. Most of the noise in your recent camera images is now the randomness of photons, not bad sensors. There's probably an 8 to 10 stop middle range where modern sensors are near perfect. At the bottom (deep shadows) they're still perfect, but in so being, they are recording the randomness of photons, not read noise. At the very top (brightest highlights) perhaps there's a bit of PRNU noise present, but most of you aren't going to notice it or worry about it. Personally, I'm not looking for more sensor capability in terms of signal to noise ratio. Maybe, just maybe, I'd take more dynamic range through the ability to capture more data above our current sensor saturations. Despite a lot of research on cascading electron wells and other ways of beating the highlight constraint, we don't see very close to anyone beating it without causing bigger issues.
- Pixels aren't revealing themselves. An Apple 5K display is 14.8mp. At over 210 dpi. I don't need more than that on screens. On paper? 24mp nets me a 28" wide print at the same dpi. Both look pretty darned good to me. So how many of you need more than that? Oh, right, you crop the heck out of your images. Get a lens. Not that I wouldn't take a well-produced extra bunch of megapixels: more sampling is always good, all else equal. The Sony A7R Mark IV is marginal for me at that (though the pixel shift can improve that). So it's not as simple as "just add more pixels."
- Still photography is not video. How many frames per second do you need? 8? 10? 12? Get much higher than that and you're in video land and you might just be better off taking video and extracting photos. Much of the pursuit of bandwidth in image sensors has been on the video side. If we're going to continue to pursue bandwidth—e.g. by going to stacked sensors—I'd like to see that bandwidth used on the still side. As in what Canon has done with every photosite being a focus photosite. Grab all that data, run it through the wringer, give us new abilities (depth maps!).
Oh, we'll get a few sensor tweaks. How about triple gain? ;~) But I'm not sure that it's at the sensor that Nikon has yet to produce a top end camera. The Z6 II and Z7 II are tough to beat in that respect. The image sensor isn't where I'm looking hardest for advancement. Just the opposite, in fact. The image sensor is the one place where mild and modest advancements are just fine by me.
All that said, here's the real reason why everyone dismisses the Z cameras as top-end systems right now: there's no "only Nikon..." By this, I mean something that would appear in marketing and reviews along the lines of "only the Nikon Z# has XYZ; no other mirrorless camera offers that." Currently, "only Canon offers 8K video," and "only Sony offers 60mp full frame, and more with pixel shift active." Canon also is the only one offering focusing at every photosite.
So what exactly is Nikon's bragging rights at the high end? Just making a very nice Z7 II?
While it might be easy to dismiss this last bit, it's important to messaging. The public perception—entirely wrong, in my opinion—is that Nikon isn't keeping up and is "behind" the competition. The YouTubers distorted that by just dismissing Nikon's autofocus capabilities without really learning how to master them. So we end up with two things: (1) the Z5 through Z7 II are really what would be the middle of a "complete" lineup; and (2) the perception is that Nikon has fallen behind.
Funny thing is, the Sony A7 Mark III is the worst full frame 20-24mp camera of the bunch right now, but #2 seems to keep people from acknowledging that. It's a middle camera that's now long-in-the-tooth compared to the competition, but you don't hear that expressed often.
