Where Are We? (Update IV or V, I've Lost Count)

For over a decade I've been writing the following about the currently-available digital cameras: "if you can't get good-looking prints at the maximum size a desktop printer can produce, it isn't the camera that's the problem." (The print size I'm referring to is basically 13x19", though we've gotten a few printers recently that could theoretically fit on a desktop and go beyond that.)

DenaliAK2004-08-13 0002

The 6mp D70. And yes, this image, with a lot of tender care in post processing, holds up quite well in a 19" print, despite seeming to only start with enough pixels for a 10" one. In fact, I processed it in a window on my monitor that was about 19" wide. 

As I've noted in some earlier articles, I've been using this rainy spring to go through my image files (all the way back to 1992). Those come from dozens of different cameras and brands and hundreds of lenses. Nothing I've seen so far says that the quoted remark at the top of this article isn't true.

So what is the problem? Why might you not get good-looking photos? And where do we really stand with today's products? Where might we go next?

What is the Problem?

So if the camera isn't the problem with getting bad looking images, what is? The shorthand answer is:

  • Bad setting decisions — poor exposure, poor JPEG settings, wrong white balance, wrong shutter speed (motion), wrong aperture (focus), and so on.
  • Bad shot discipline — poor handholding, support that isn't, poor timing.
  • Wrong lens — too much cropping, observable lens fault for the subject (e.g. linear distortion for architecture, vignetting for landscape, etc.).
  • Bad processing — too much or improper sharpening, wrong colors, too much saturation, improper white/black levels, incorrect noise reduction, wacky contrast decisions.

I've seen plenty of all four of these things in imagery from others when they begin complaining about their camera. 

Before you complain about your camera, though, make sure you're not guilty of any of the things I just called out. Buying a new camera doesn't solve any of them. Some think that more automation will solve their problem—e.g. set Auto on a new camera and bad setting decisions go away—but more often than not in the best case that just locks them into a very specific, and not optimal, look. It's faux progress, really.

I'm going to add one additional problem that's going to be a bit controversial: wrong camera for the job. 

There's a reason why sports photographers on the sidelines of college and pro games are for the most part using Canon 1Dx's, Nikon D5's, and Sony A9's. If they're not using one of those, then often they're using Canon 7Dm2's, Nikon D500/D850's, or Sony A7m3's.

In some types of photography, a few features and performance factors do come into play. Make sure you know what those are and that you're getting them if you value one type of photography over another.

Where do we Stand?

Before I get into specifics, let's go through those four problems again:

  • Bad setting decisions — more automation and specifically AI automation is being used, with more coming. If you believe that computers make better decisions than humans, you're probably jumping for joy. I don't think so, and automation tends to drive all decisions toward the mean, reducing creativity.
  • Bad shot discipline — image stabilization is what many think solves this. For those that are really bad at handholding, yes, it probably does solve most of their problem. Don't think this comes without a penalty. IS impacts bokeh, for instance. At certain shutter speeds, it can impact acuity. Again, things are being driven towards the mean, so be careful if you're someone that wants to stand out.
  • Wrong lens I'll get to the cropping note when I talk about pixels in a bit. But now that we're in a world of "lens correction algorithms", be careful that those aren't producing unwanted side effects.
  • Bad processing — I can't tell you how many people I've seen turn their cameras up to 11 (that would be something like using a Vivid picture control with extra sharpening and contrast). Because they can. My images look bland coming out of the camera. Because I shoot raw and remove all the Extra Sauce. But what I see more and more people doing is abandon bad after-the-fact raw processing for bad in-camera processing as the cameras get more sophisticated and offer more selections.

Here's today's state-of-the-art: 36mp or more for large format work or if you are heavy cropper, 24mp for low light and fast frame rate work. A realistic 11 stops or more of usable dynamic range at base ISO. All of Nikon's current full frame cameras except for the Df and D5 hit those marks, as do some past ones (e.g. D800, D810). All of Sony's A7/A9 current cameras and a few of their past ones hit those marks (except for the A7S models). The Canon R and 5DmIV hit those marks. 

(The D5, A7Sm2, and 1DXm2 are all what I'd call speciality cameras; see my comment in the previous section about right camera for the job. These cameras are the right camera for some jobs.)

APS-C (DX)—at least from Fujifilm, Nikon, and Sony—tends to fall just off today's state of the art for low light and fast frame rate work, but not far from it. You'd expect a stop of difference in DR due to sensor sizes, but in some cases it's a bit less than that. But we don't have any APS-C cameras that could be said to be suitable for large format work (you'd have to shoot panos or use sensor shift techniques). And you don't have a lot of room for cropping, so make sure you have the right lenses.

(What about Canon APS-C, you ask? Unfortunately, Canon has fallen behind the Sony-based crop sensors enough so that it shows. Apparently, buyers are starting to notice, too.)

Features and automation for top-end cameras is getting pretty well balanced across brands now. I've not picked up a recent camera that I can't get exposure, focus, and much more right with very little effort. You do have to study the camera and make sure you understand how to control it, though. UX (user experience, which includes user interface, grip, and more) is different from brand to brand, and even from model to model in some cases (I'm looking at you, Canon RF). 

Which brings me to another thing I've been writing for some time: given the relative parity among camera features and even performance these days, how you react to controlling the camera is probably the most important factor that most people don't pay enough attention to. 

Nikon, for instance, has been doggedly consistent (overall) with their button+dial interface dating back into the 1980's. You can pick up an N8008, F100, D100, D500, D850, or Z7 and find that the basic Nikon UX DNA is intact. Yes, a button or two may have moved—often for no good reason, which is why I call Nikon out when they do that—but it's pretty easy for an existing Nikon user to pick up any new Nikon and adjust quickly. Change to another brand? Nope, lots of new learning will be triggered, and muscle memory has to be redeveloped.

That's why I've often written that it makes little sense for a Nikon user to switch to another brand. I find a lot of those that switch end up switching back after time, and that's because they are most at home with the Nikon UX. (I could say the same thing about Canon cameras until recently, when the M and R series started messing with the Canon DNA. I'm starting to say the same thing about the Sony Alphas, though I really think the menus have been messed up and keep juggling enough change between models to confuse folk moving between Sony cameras.)

So, let me state a couple of strawman proposals about the current state-of-the-art:

  • For most folk, current 24mp is enough (APS-C or full frame), and most current cameras at any given sensor size and pixel count are within sight of parity at any given price point. 
  • For those producing large prints or needing to crop aggressively, most current cameras of 36mp or above are enough, and most current cameras at the full frame sensor size and high pixel counts are again within sight of parity, even at different price points (e.g. 45mp full frame versus 50mp medium format).

The corollary is that if you're below the above (e.g. <24mp or <42mp), then you're probably going to end up significantly below state-of-the-art. (Again, the specialized cameras like the Canon 1DXm2, Nikon D5, and Sony A9 I'd judge a bit differently. But these are not cameras you buy for all-around general use, IMHO.)

What about 20mp, you ask? After all, the top Olympus m4/3 cameras and the Nikon D7500 and D500 are at that mark. Well, here's the somewhat damning answer: you don't have a lot of crop flexibility, so you need to be using the right lens. If you're producing mostly for social media, you probably aren't bothered by the lack of pixels. If you want to print or display large, you probably will be. But again, it depends a bit on whether you composed right in the field with the right lens or not.

So, are you below or above what I just described? If below, sure, maybe it's time to think about upgrading. If you're at or above my strawman bar, I'd say you should spend your time looking to improve in places other than the camera.

Like lenses. While what I wrote about cameras is arguably true, you can muck that all up by putting the wrong lenses on the right camera (or the right lenses on the wrong camera). The 28-300mm f/3.4-5.6 Nikkor is not state-of-the-art on a Nikon D850. Simple as that. That lens is all about convenience, while that body is all about performance. That's a poor combination choice, in my opinion. Yet I keep finding people trying to make that choice.

Likewise, the 70-200mm f/2.8E lens—the best telephoto zoom in that range I've encountered to date—isn't exactly a good choice on an older D700 or D7000 body. At 12mp full frame or 16mp crop sensor—and both older technology sensors to boot—you're putting more lens on the camera body than really resolves well. The older 70-200mm's were quite adequate for those older bodies. 

To me, this is the trickiest part of "state-of-the-art" for photography: lenses. In the past decade I'd argue that we've seen more progress on lenses pushing performance upward than in image sensors. Some recent lenses I've tested are spectacular. True, we wouldn't see exactly how spectacular they are without more and better pixels, but the thing I keep noticing in my image files as I go through them is this: the best lenses show through, no matter what camera body I put them on. 

I'd argue that if you're going to make your images stand out from others—besides with your eye for composition and your technique—things have shifted some recently from "use better cameras" to "use better lenses." 

Where are We Going?

Here's a truism: there's no stopping the Japanese engineers when it comes to incremental iteration. Once you hire them—and in Japan there's still a strong sense of hire-for-life—they just keep making whatever it is they were working on better. It's what they do. And they do it incredibly well.

So more pixels, better pixels, more bandwidth, faster/better focus, deeper feature sets are a given. Lenses, particularly on the short mirrorless mounts, will continue to get better, too.

The problem I see for all of those things is that there will be fewer and fewer takers on the buying side, as the general image quality capability of today's cameras and lenses already far exceed what most people need. Pros certainly want to stay on top of what's possible in imaging, otherwise they can't charge good money for what they do. They need to use any small advantage they find to help them stand out from the crowd. But that's a small subset of the buyers of sophisticated cameras. 

The high-end enthusiast often can be triggered to buy something new based on FOMO (fear of missing out). More Pixels! Faster Frame Rates! Higher MTF! This was what propelled the High Fidelity snob aficionado back when CDs and then MP3s became the norm for most people, but the number of those folks who carried on at the high end was determined by disposable income, and eventually buying exhaustion. The higher the HiFi market reached, the fewer it reached ;~). The same thing is starting to happen with interchangeable lens cameras now.

In my shortest essay ever, I point out a truism: "There is no shortage of customers who want to take photos and share them." The camera makers, unfortunately, aren't making cameras for those folk. And I see no signs that the camera makers want to really try. 

So I have to ask: do you really want a 70mp full frame camera? Bigger, better, heavier lenses? More dynamic range than the median scene has in it? AI automation that makes every camera decision the same and thus most pictures look the same? Because that's where we're headed. 

text and images © 2019 Thom Hogan
portions Copyright 1999-2018 Thom Hogan-- All Rights Reserved
Follow us on Twitter: @bythom, hashtags #bythom, #dslrbodies