Plus Three Reasons Why I Could be Convinced to Move On
Update: since this article got multiple mentions elsewhere and discussions about it have ensued on various fora and comment sections, I need to remind everyone to read the title carefully. This article is about me and my choices for my photography. It’s not “you should buy a DSLR because…” or “you should avoid mirrorless because…”. It’s why I’ve made the choice I’ve made at the moment, simple as that. You are not the same as me, so please don’t try to justify your buying decision by saying I’m wrong.
People ask me all the time why I haven’t just moved to a mirrorless system. Sure, there’s some inertia in my choice to stick with Nikon DSLRs for most of my work given that I have a full F-mount lens set, but there are some genuine other reasons for my sticking to Nikon DSLRs, too. Let’s take a look at them:
Much of what I shoot is what I’d call “timely.” Sports, events, wildlife. Miss by even a millisecond and I don’t have the peak moment and the photo is less engaging than it could be. And no, frame rate doesn’t begin to make up for this if you’re at the level of shooting I am.
EVFs always have a lag. The very best lag I’ve seen in one is the one in the Samsung NX1, which runs at about 1/250 behind reality. Samsung did something that is mostly done in the video system: they “genlocked” the EVF to the image sensor. I won’t get into the technical details of what that means, but in essence the signal is about as direct as you can get from sensor to EVF.
The problem is that to make that 1/250 lag faster you need to run the sensor faster. Which means more technology in and around the sensor, and the sensor will probably run hotter, too. Samsung and a few others are running EVFs at about the max speed you can with current technologies. The good news is that will change with time as new abilities and technologies emerge. But it will change somewhat slowly, I think.
And that 1/250 lag is lag. That 1/250 visual lag sits on top of your recognition lag (response to the scene in front of you) which sits on top of the shutter lag. So you get lag+lag+lag = real lag. I don’t want lag. Indeed, I have to make sure that I’m on top of my game and my brain is processing fast when I’m shooting a number of subjects. I mostly choose cameras—such as the Nikon D5 or the D500—that have minimal shutter lags and no viewfinder lag.
But the initial EVF lag isn’t the only lag in the mirrorless systems. In DSLRs we talk about “viewfinder blackout,” which is the time that the viewfinder is dark while the mirror is flipped up. The D5 has an incredibly short viewfinder blackout, meaning that when I’m panning with action there is very little time I’m not looking at my subject in real time. The D500 is also quite good at this.
With mirrorless systems, some have this funny “slide show” kind of effect when you’re shooting continuously. The latest Fujifilm’s and Sony’s have minimized this, but they also need time to clear the sensor data before turning it back on. So there’s still a bit of disjointedness to what you’re watching. And again, if you keep running mirrorless systems continuously, you’re building heat at the sensor. I’d prefer to keep that to a minimum when I’m shooting in low light.
So, the DSLR’s complicated optical viewfinder just keeps me closer to what’s happening in front of me, both in timing and continuousness.
I already mentioned this one up front. Indeed, it’s the primary reason for most DSLR users to resist moving to mirrorless: they already have built a lens set, and the cost of doing that wasn’t minimal.
It used to be that Nikkors retained their value quite well. But the overabundance of lenses that occurred with time and the abandonment of the F-mount by some has severely reduced used prices for Nikkors, so that adds another inertia: if you spent US$4000 on lenses and can now only get US$2000 for them and now need to spend US$4000 on lenses on a new platform to match those… Right, that new platform better have US$2000 worth of advantage to it right out of the box.
But that’s considering that you can even match the lens set you used to have. With the F-mount there are only about six lenses I’d want that don’t exist in the current offerings (and three of those can be approximated by dipping into the used Nikkor market). Oh dear, you want to know which six. Well, quick and dirty:
- A really wide PC-E (e.g. 16-18mm)
- Wide DX primes (20mm and 24mm equivalent) (buzz, buzz)
- A longer Micro-Nikkor, preferably a zoom
- Compact telephotos with slower apertures (400mm and 500mm f/5.6 PF)
m4/3 probably has the broadest selection of lens choices, with Fujifilm X being reasonably broad in the wide-to-normal range and deficient in the telephoto range. Sony is mostly in the 24-200mm range still (just announced a third ~50mm as I write this), though the recent 70-300mm gave us a first peak outwards, and the third-party Loxia 21mm gets us further into the wide end with quality. Still, Not a single mirrorless platform matches Canon or Nikon full frame offerings in breadth and depth, and if Canikon weren’t brain dead, they’d make sure that was true for APS/DX, too (buzz, buzz).
I press lens choice pretty hard compared to most. I shoot landscapes (thus the need for very wide and PC-E). I shoot wildlife and sports (thus the need for telephoto). And yes, I shoot events and other things in between where the 24-200mm zoom/prime range most mounts can fill adequately lives.
3. Battery Life
Mirrorless cameras are not only running the sensor continuously, but they’re running either the EVF or LCD continuously, too. Remember there’s a lot of circuitry behind both, and that starts chewing battery. It doesn’t help that many of the mirrorless camera batteries have lower Watt hour ratings than most DSLR batteries, either.
It also doesn’t help that most modern cameras tend to “crash hard” when batteries go low. By that I mean you don’t get a lot of warning before, poof, the camera is inoperable (but apparently has enough power to tell you that it’s out of power ;~). This makes me tend to change batteries the first chance I get after I notice the battery indicator has dropped to low. (And don’t get me started on two and three-segment battery status indicators; power is something that should present nuanced and meaningful information, not a full, less full, empty value.)
So if I’m only getting 300-400 shots per full charge with a mirrorless camera, I’m tending to actually only get 200-300 shots in practice because I’m changing the battery early to avoid not having power at a time when I need it for critical action. Last year I covered a soccer match with a Nikon D4 and a Sony A7rII. The D4 barely made a dent in its battery in two+ hours of shooting. I went through three batteries on the Sony, and I was running them down further than I normally do. In Africa, I’ve actually made it more than a week at a time on one battery with my big DSLRs, but end up running through multiple mirrorless batteries every day.
4. Focus Speed and Accuracy
In their latest high-end DSLRs Canon and Nikon have once again shown that the old DSLR-type phase detect system coupled with good in-lens motors just is unmatched in speed and accuracy when a subject isn’t stationary. I don’t know how else to state that. Unmatched in speed and accuracy when a subject isn’t stationary.
This isn’t to say that mirrorless cameras are slow in autofocus or inaccurate. Fujifilm, Olympus, Panasonic, and Sony have made great strides forward in mirrorless focus performance. For static subjects, they may be as fast as and more accurate than DSLRs now.
But frankly, other than landscapes and an occasional cityscape, my subjects are rarely static.
Both DSLR and mirrorless phase detect systems have liabilities that have to do with geometry and physics. A DSLR produces data to the focus system that has a great deal of discrimination to it: the DSLR knows very precisely the distance the subject was from the focal plane. Where the DSLR gets into trouble is that it relies on the lens to absolutely get to that same precise focus point in one move.
On-sensor phase detect has far less precision to the distance measurement due to the very small gap between the microlenses that produce the phase shift and the data collection mechanism. But they tend to make up for that slight imprecision by doing a contrast detect verification step immediately following the lens moving focus to where the phase detect system indicated. This adds a little extra time to getting focus to the right spot. For static subjects, you won’t notice. For moving subjects, you will.
What happens with moving subjects is twofold: (1) the contrast detect verification sometimes has to play catchup, and sometimes can’t; (2) the autofocus systems in most mirrorless cameras seem to let the camera take a photo if they think the focus plane is within the depth of field. What I see—especially with the current Sony implementation—is that in continuous shooting of a moving subject you get some images dead on, some images a bit off. On the very latest DSLRs, I rarely see images of moving subjects “a bit off.”
Again, I shoot a lot of moving subjects, some of them quite fast and irregular in motion. No mirrorless camera currently comes close to matching my current shooting pair, the Nikon D5 and D500.
5. Controls & Ergonomics
As much as I complain about small ergonomic issues with Nikon’s recent designs—in particular, a lot of what I call “moving the cheese”—I tend to be nit-picking and perfectionist in those complaints. I can fully control my D500 without taking my eye from the viewfinder when shooting, the hand position is good for long periods of time, and I have enough information presented in the viewfinder to make good decisions.
I can’t say that about any mirrorless camera except maybe the Fujifilm X-T1. Even there I had problems with a few things (a couple of which look like they’ll be addressed with the upcoming X-T2). Olympus’ terminology and menus and conditional options are a mess. They are the same exact mess as they were on the original Pen E-P1, which means that no one at Olympus cares about ergonomics. Sony’s menus have gone through change after change after change from the original NEX to the current Alphas, but now are a sprawling, disordered mess. They, too, have some confusing terminology, though not nearly as much as Olympus.
But it goes beyond just menus and terminology. Some mirrorless cameras just seem to have controls strewn wherever the designer found convenient. Others have such tiny indented buttons that I can’t find them by feel, let alone use them with even thin gloves. Some cameras don’t have reasonable hand grips. Others have uncomfortable hand positions when used for any length of time. A few have all of the above.
Yes, I know about Fujifilm’s retro style controls. I also found my original X-T1’s indented Direction pad unusable for the most part. The mirrorless camera maker that I generally don’t have a lot of issues with across a wide range of their product is Panasonic. Other than their propensity to use cryptic and odd abbreviations in their menus, it actually seems like most of their cameras were designed with someone shooting for long periods across multiple situations in mind.
The DSLRs have the advantage here in that both Canon and Nikon are building on decades of refinement of the same basic form and shape, and over a decade worth of menu refinement (plus Nikon started out with a reasonable structure to start with). But even the DSLRs could use some good ergonomic advice; meanwhile, the mirrorless makers tend to need more than just some additional advice, they need to start locking in on long-term benefits to their designs. I can tell you this: I don’t want to be using a Sony A7VVII in a dozen years that has the same ergonomics as the current A7.
6. Buffer Performance
The mirrorless cameras were getting better at buffer issues, at least until the D500 rolled into town. 10 fps and a 200 frame buffer basically add up to “never worry about continuous shooting ever.” I’m not a shutter masher, but even with my D810 and D7200 there were times when I found myself waiting for the camera. Not so with the D4, D4s, D5, and D500. Just doesn’t happen.
Until the X-T1 came around, Fujifilm had a real buffer issue: they simply put slow write mechanisms into their cameras and wrote huge uncompressed raw files to the card. Slowly. Many of their still-on-sale models still have this problem, though it looks like Fujifilm is going to address that in all the next generation updates.
Indeed, it’s the dirty-little-secret across the camera industry: they make all these claims about technology and performance, and then they wimp out in manufacturing and buy parts that most of the tech industry wouldn’t touch because of their slow-to-current-standards performance. Even in the DSLRs you’ll find cameras that can’t write to a card faster than 50Mbps (and don’t get me started on Wi-Fi parts). But that’s the norm in the mirrorless world, it seems.
Almost all the mirrorless cameras use SD cards, but virtually all aren’t even close to capable of what a UHS-II SD card can do. Couple that with putting less physical memory into the camera, and you have “shoots at maximum frame rate right until the smallish buffer fills, then takes forever to clear.”
And Sony? OMG. Continuous shooting on the A7rII can be exceedingly frustrating if you’re on and off the shutter release using back button focus (AF-ON). If I had a dime for every time I’ve seen the “Writing to card. Unable to Operate” message, I’d be rich.
This reason for preferring my DSLR is probably going to surprise you a bit. I like Nikon JPEGs and NEFs. The JPEGs don’t tend to have those complicated hue twists that Canon and Fujifilm and Olympus use. Indeed, if you set a Nikon DSLR to a Picture Control of Neutral you will get about as neutral an image as you can get out of any camera. It’s devoid of color shifts, extra contrast, hue changes, extra saturation, and even edge artifacts due to sharpening. Clean, 8-bit JPEGs. That’s what I want from a camera and I get from a Nikon DSLR.
Not so much with the mirrorless cameras. Fujifilm and Olympus both hue shift and add saturation. Olympus tends to add contrast in every setting, while Fujifilm at least has many settings that temper that down to almost neutral. Panasonic’s JPEG engine still seems less refined than many of the others, producing less pleasing color, more noise, and more artifacts than I see from the other companies.
Sony is probably the closest to the Nikon JPEGs I like, and the latest BIONZ processing has finally tempered the overaggressive noise reduction and produces clean JPEGs. Still, Nikon Neutral is Neutral (especially with the new Keep White automatic white balance when shot in sunlight). And clean. And remarkably free from artifacts even if you decide to crank the compression all the way up to Basic Size Priority (which can cut an already compressed JPEG size down by 7/8ths).
But I find the Nikon NEFs pretty well behaved, too. One thing to note about Nikon DSLRs is that Nikon has long labeled ISO values with numbers when the sensor/ADC is controlling the gain to produce the final digital numbers (DNs) in the raw file. ISO values not labeled with numbers (LO, HI), use various other techniques to generate the DNs including multiplication of data, repositioning of data, and noise reduction in the data. Other camera makers don’t seem to do this consistently, if at all, so you have to be careful to figure out where they start “cooking” the data in raw files.
Likewise, on the upper end cameras where RAW shooters prevail, Nikon offers a choice of 12-bit versus 14-bit, and various compression options. At base and low ISO values, 14-bit can produce slightly better results. But at ISO 400 and up 12-bit is generally all you need, because the extra two bits in the DNs aren’t useful or meaningful data.
Virtually no mirrorless camera maker has gotten to that level of sophistication in their raw files yet. In fact, it’s only been recently that some have offered any compression (Fujifilm) or have offered something other than lossy compression (Sony).
With my Nikon NEFs I tend to make about a half dozen adjustments to Adobe’s converters to get the color/tonality right. With my Olympus, I make 18. Developing those adjustments for every new camera is a lot easier in the Canon/Nikon world than it has been with some of the mirrorless cameras. I’m still not 100% happy with any raw converter—even after adjustments—on Fujifilm’s X-Trans sensor designs. Close, but not 100%.
Most of the mirrorless system issues I note above will tend to go away with time, as technology “solves” some of the problems, and digital cameras evolve even more than they have.
But DSLRs have been benefiting from technology moving forward, too, so it’s not as if mirrorless is trying to catch up to a stationary target. The D5/D500 proves that the target is still moving in big strides forward in some areas, and the DSLR/mirrorless problem has become like one of those algebra problems you had in grade school: “If train A leaves the station headed west at 75mph at 1pm and train B leaves the station headed west at 100mp at 3pm, when does train B pass train A?”
That said, there are things that the mirrorless systems provide that I like, which is why I use them for some things. Let’s next address the things that tend to make me consider moving from a DSLR to a mirrorless.
A. Size and Weight
This isn’t a given advantage to mirrorless. The Panasonic GH4 is an awful big camera for its sensor size, for example. Still, in general the mirrorless camera producers have tended to make smaller and lighter cameras than DSLRs.
That was (and still is) intentional.
DSLR makers could make smaller and lighter DSLRs, but they don’t tend to. Thus, as the camera makers that couldn’t knock Canon and Nikon off the pedestal in DSLRs moved to mirrorless to try to establish a new category they could dominate, they looked at things that they could do differently that would clearly distinguish those cameras from DSLRs. Number one on the list: just make their new cameras smaller and lighter, but seemingly as flexible (e.g. interchangeable lenses, system accessories).
Over time, this “smaller/lighter” mantra in the mirrorless realm has led to some very tricky plays. When Sony built the full frame sensor A7 series, they didn’t make the camera body all that much lighter and smaller, but they “cheated” by clipping off the hand grip and producing f/4 zooms that were obviously smaller than the f/2.8 zooms people were using on their DSLRs. In particular, a wide angle f/4 zoom for a mirrorless system can take advantage of the short flange distance and be substantively smaller than what a DSLR would produce with the same specs. Other corners were cut, as well.
Given that cameras now have built-in lens corrections, many of the mirrorless lens designs simply don’t control for linear distortion or vignetting: they let the camera fix those. I’m seeing some pretty big distortions and extreme vignetting in many cases with mirrorless lenses. 6-10% linear distortion and 2+ stops of vignetting. Obviously, if you design lenses this way, you can make them simpler, lighter, and smaller.
That’s not without consequence. If you push the corners of an image up by two stops in post processing, you tend to produce visible noise, especially if you were at a higher ISO value or using a smaller sensor to start with. And distortion correction writ large like that makes for some unusual optical effects in the corners, too.
Still, there’s no arguing that you can get a ridiculously small mirrorless system (Olympus OM-D E10 for example) or even a small, usable, full frame mirrorless system that packs small and light (Sony A7 series with some, but not all, of the lenses Sony produced).
For backcountry hiking and for travel photography I find small and light to be a big win over my bulky and heavy DSLRs. And I can generally work my way around any of the tricks and consequences that the mirrorless makers snuck into their systems to make them so small.
B. Accuracy of Static Subject Focus
DSLRs aren’t slouches at this, for sure, but some of the mirrorless cameras just truly nail the focus plane in ways that the DSLRs would love to, but don’t. It’s that last contrast-detect step that most mirrorless cameras do that nets them the more precise focus position (assuming the lens you’re using is capable; a few of the lower cost lenses aren’t quite as good at this precise positioning).
For landscape photography and travel photography I appreciate that. But there’s an even better aspect to this on many of the mirrorless cameras: touch focus. The mirrorless cameras with touch screens allow you to simply tap the thing on the rear LCD that you want to be precisely in focus. (Technically, a D500 and a few other DSLRs can now do this too, but the focus system they use in Live View tends to be slow and not as good.)
Many of the mirrorless cameras can be set to a truly quiet mode, where the physical shutter is disabled and an electronic shutter used instead. This makes for completely silent shooting.
I wrote when the Nikon 1 came out that I couldn’t figure out why every pro golf photographer wasn’t using a V1. Truly silent shooting at 20 fps lets you shoot an entire swing on a golf course without some caddy coming over and bashing your camera into the ground.
But there are plenty of mirrorless cameras that now do the same thing, and plenty of situations besides golf where shooting totally quietly is a huge advantage that DSLRs can’t match. Theatre and music performances. Street photography. Even some parts of weddings.
Like many things mirrorless, silent shooting doesn’t come without a downside, though. When you run the sensor in electronic shutter mode you tend to get slightly more noise in your image than when you run with the mechanical shutter. This is especially true if you’re doing shot after shot. Still, the visual impact is minimal, and worth getting the shot you couldn’t get otherwise.
So it isn’t a shut-out for the DSLRs. As with many types of tools, the one you really want to use for any situation depends upon exactly what you’re trying to do.
For me, the majority of the time the thing I’m trying to deal with is moving, and I have to react quickly and precisely in my handling of the camera to get the image I want. DSLR.
Some of the time I’m working more deliberately, or ranging far and wide on foot and don’t want the extra weight. Mirrorless.
Everyone who reads this article is going to have a slightly different take than me. That’s because you don’t shoot like me, you don’t shoot the same things as me, and you don’t have the same requirements as I do. Still, I wanted to put this article out there, because a number of you continue to ask why I stick with Nikon DSLRs. The short answer is: because they work for what I do and other things don’t work as well.
This is just another reason why you have to analyze want versus need. I need my cameras to do certain things in specific situations. Mirrorless simply hasn’t managed to get to where I need it to. It may never get there (in my remaining lifetime). At times I want some of the things that mirrorless provide, but that’s not the same as need.
So I continue to shoot with Nikon DSLRs.