Points in Focus Photography

Image: Nikon USA Press Release

Thoughts on Nikon’s D850

0

Nikon finally announced their latest high-end/pro-level DSLR, the D850. Honestly, there’s a lot to talk about with it both as an upgrade over the old D810, as part of Nikon’s current lineup of cameras, and with respect to Nikon’s future prospects in the industry.

Recent times have not been good to Nikon. Camera sales have been down and of the big the camera makers, Nikon is arguably the most vulnerable to declines in camera sales. Moreover, Sony has been aggressively trying to move into the market space, and most of their gains have been coming at Nikon’s expense.

Compounding matters, Nikon’s D800 series cameras can certainly be perceived as being extremely long in the tooth at this point.

Nikon’s first real attempt to get things moving again, so to speak, was the D5 and D500. Cameras that very much were both simultaneously impressive and disappointing.

In many ways, the D5, and to a very much lesser extent the D500, looked to me like a panicked response by terrified management that saw market share eroding and could only think of how to inflate numbers to make something that looked, in their eyes at least, competitive.

Fortunately for Nikon users, the initial flailing response that the D5 and D500 seemed to embody seems to have past. Nikon seems to have remembered what matters; not putting bigger and bigger numbers on a box to delude gullible photographers into buying the latest thing, but providing a solid tool that photographers can use to produce the images that want to make.

Looking at the announced specs, and some very preliminary performance estimates that have surfaced, the D850 looks to be exactly the camera Nikon needs right now. So lets dive in to looking at some of the announced specs, and what some preliminary testing implies.

Sensor

It’s reasonable to say that for most photographers the sensor is the most important part of the camera. Unfortunately, far to many take this some incoherent extreme seemingly hell bent to judge cameras based on their sensor’s metrics (usually dynamic range) alone, and then proclaim in some nonsensical absolutes terms that everything else is garbage.

From where I sit, I find it hard to become so caught up by such numbers. Many of the most enduring images in the history of photography had been made on film — frequently with worse resolving power, more coarse grain, and poorer color rendition than modern digital cameras — and with what glass that wouldn’t even pass muster as consumer grade crap by modern objective standards. Yet the images still stand and still evoke a response. It’s not the camera, but the photographer that matters.

That said, a better sensors is always better to have. And while a better sensor won’t make a bad photographer into a good photographer, it can allow a good photographer to press the envelope even further. So while great photos can be made with nearly any reasonably modern DSLR, that’s not a good reason not to push sensor development forward.

So the pressing question then becomes, does the D850 do that?

After all, one of the biggest disappointments in the D5’s sensor was the loss of dynamic range at los ISOs. While I don’t know the exact reasons for that loss, the most reasonable thought is that it was a by-product of the frame rate the D5 operated at. And with the D850 increasing the frame rate over the D810, by nearly 50% too, it certainly seems like a good question to ask if it’s happening again.

Out of curiosity, I turned to an increasingly favorite site of mine for sensor performance data: photonstophotos.net. And sure enough, Mr. Claff has already found and posted some estimated D850 performance numbers (unfortunately he doesn’t note where the data/estimates came from but presumably they’re raw images that have made it out into the wild somewhere).

Nikon D850 Estimated Dynamic range versus D810 measured dynamic ragne.
Chart from Photons to Photos.

Looking at the above chart (click here to go to photonstophotos.net where you can add cameras to compare the D850 with), ISO 100 photographic dynamic range (PDR) is essentially identical to the D810’s at essentially 11 stops[1].

At a minimum this strongly implies that the D850 doesn’t suffer from the same low ISO DR limitations than the D5 suffers from. Either the sensor tech uses here is sufficiently advanced compared to the D5 to address the noise/DR issues, or frame rates simply don’t create an issue. Truthfully though, if the D850 can turn out performance like this while still being able to read out 120 FPS video, the D5 should have had none of the the limitations it has

However, looking at just the ISO 100 data point alone doesn’t tell the whole story. Looking at the estimated performance as a whole, the D850 tracks extremely closely, a little better even, to the D810 in terms of PDR.

Overall, based on this early data, the image quality of the D850 should be pretty stellar. It sees a 26% increase in resolution over it’s predecessor, and does that with what appears to be equal DR, and likely nearly equal noise performance. Combined those two points mean that the resulting images will be better than the D810’s.

One final point, the ISO range is 64-25,600 native, and 32-102,400 expanded. Quite simply, this is sensible. In fact, the ISO range is the reason I think Nikon has gotten past the panic-induced-flailing.

The simple reality is that there are finite limits to just how much ISO is really useful, necessary, and even more importantly, there are finite limits to just how poor of image quality photographers are willing to accept. While I won’t say that ISO 102,400 for certain is that threshold, it is the threshold currently where current sensor tech can still produce an image with 4-5 stops of dynamic range and enough reasonable detail to have an understandable image.

In any event, with the given ISO range, Nikon seems to be back to prioritizing delivering top notch image quality instead of big numbers on the box.

The Pitfalls of 45.7 MP

That said, while the sensor looks technically sound and generally an improvement over the D810. The real concern for me isn’t technical, or at least it’s not in the camera per say.

Twice now I’ve made significant increases in resolution in my time with photography. The first was an insane 120% increase going from a 10.1 MP 1D mark III to a 22.3 MP 5D mark III. The second was a more reasonable 36% increase going from the 5D mark III, to the 30.1 MP 5D mark IV.

Both times I’ve done this, I’ve run into the same set of problems: poor glass, less than stellar technique, and increased storage and processing requirements.

First, the higher resolution sensor found more and more faults in my glass. Lenses I thought were awesome on my prior camera started showing more faults on the newer higher resolving one. Though the D800 and D810 already are relatively high resolving cameras, certainly. The D850 still presents an appreciable increase in resolution, and those demands fall squarely on the glass owners have. A lens that looks good on the D810, simply may not be good enough to hack it on the D850.

Actually, if you’re fortunate enough to have a D500, the pixel pitch is almost identical between the two cameras. Any lens you currently own that fails to meet your sharpness requirements on the D500 (and bearing in mind in the case of a full frame lens, the D500 won’t let you see the corners), almost certainly won’t meet your sharpness requirements on the D850.

Secondly, the increase in resolution demands a likewise increase in good technique and stability. Even if you’re an existing D810 shooter, and even if you’re comfortable that you’ve adjusted your technique to meet the needs of the D810, you’re almost certainly going to need to refine and improve your technique further with the D850.

Shooting action at 7 or 9 FPS? Expect not to have pixel sharp images, even at moderately high (say 1/1000-1/2000) shutter speeds.

Finally, the 3rd point is storage and processing. In my experience, many people, especially those with some background in technology and computers, are fond of pointing out how storage is cheep. And compared to 10 or 20 years ago, storage is cheep. However, in practice when factoring in backups and redundancy, that never seems to actually be the case. Never mind the increased need for cards in the field.

The other big problem is processing, bigger images are that much more computationally intensive to process. While computers have been ticking out incremental performance improvements for some time now, not only are the images getting bigger, but the software is adding more complexities to them. Processing these images, especially on a laptop in the field, is going to be just much more of a problem compared to the D810.

None fo these are insurmountable problems. In fact, many can be addressed with improvements to technique, but they are still something to consider. All the image quality improvements the D850 offers might not matter if you can’t effectively process and store the images.

Storage

I can best sum up my point on storage with this: more XQD.

Dual XQD/SD Card slots.
Image courtesy of Nikon USA

The D850 Nikon provides users with dual card slots of differing types. One is a UHS-2 capable SD card slot, and the other is the increasingly niche XQD slot. Unlike the D5, they cannot be changed.

Nikon’s continued use of XQD fascinates me. Technically speaking one could argue it’s superior to the alternatives (namely CFast and UHS-II SD). It’s a PCI-express based storage format, which promises high speeds and low latency. However, support and adoption simply isn’t very broad. Moreover, broadly speaking, availability, especially when the practical performance differences approach zero, trumps technical purity every time.

On the manufacturing side, only two companies have made XQD cards; Sony and Lexar. Of those two, Lexar is exiting the market, leaving only Sony. Moreover, none of the other major flash manufactures are building XQD cards (SanDisk, Delkin Devices, Kingston, PNY) or have announced any intention to take up the vacancy.

The lack of broad manufacturing support means there are no alternate sources for cards. And now that Lexar is leaving the market, users simply will have to hope that Sony is interested in continuing to support the format going forward (and not raise the prices significantly as well).

Admittedly, pricing right now for XQD cards is reasonable and on part with Compact Flash and generally cheaper than CFast II. Moreover, digital media is generally slow to wear out and not something that needs to be replaced frequently. However, at least half of the storage situation does give me pause; and the other half, SD, I’m still not a huge fan of as a professional storage media.

Frame Rate

One of the long standing complains by many D700 owners when the D800 was announced was the major jump in resolution at the cost of a lot of frame rate. The D700 was a general purpose camera, equally at home shooting sports and landscapes, but perhaps not ideal at either.

The D850 very much brings the general purpose role to the D800 line, with a maximum frame rate of 7 FPS (or 9 with the battery grip and the D5’s battery) and the resolution to be used for demanding material to boot.

Of course, 7-9 FPS comes at a cost; you’re going to blow through storage at an obscene rate.

Estimating based on the premise that D850 files will be about 26% larger than D810 files. We’re looking at a 14-bit lossless compressed RAW being in the 60MB range. Loss-y compressed RAW files, which you probably should be shooting anyway for most things with this camera, will be in the range of 46MB.

At 7 FPS you’ll be generating 420 MB/s of lossless compressed 14-bit data, and a 64GB card will be good for a mere 150 seconds of continuous shooting. At 9 FPS, you’re looking at 540 MB/s of lossless compressed RAW data, and a 64GB card will give you a mere 111 seconds of continuous shooting.

Backing the 7-9 FPS drive rate up is a claimed 51 frame buffer. Tough the actual buffer will be entirely dependant on the write speed of your media.

Camera makers, and Nikon is no exception, like to talk about buffers in terms of the number of frames you’ll get out of them; so 51 frames or 28 frames. I find this to be a largely useless metric when it comes to continuous shooting, especially when framerates change. Instead I like to look at the buffer in terms of how long I can hold the shutter down and shoot.

For the D810, the 28 frame buffer yielded 5.6 seconds at 5 FPS (the max framerate, regardless of power source, that you could shoot the full frame resolution at). The D850’s larger buffer combined with it’s max 9 FPS frame rate, yields a nearly identical 5.7 seconds of continuous shooting. At 7 FPS, the D850 yields a nice 7.25 or so seconds of continuous shooting. At least using simple math, I would expect the actual performance to be slightly different.

The performance of the D850 as a high speed shooting camera is certainly solid. 6.6 seconds of continuous shooting at 9 FPS at 45MP is nothing to shake a stick at, and certainly beats the 3-4 seconds my 30 MP 5D mark IV does at 7 FPS.

Autofocus

High FPS without being able to focus and track what you’re focusing on isn’t all that useful, so you need a good AF system. As critical as I’ve been of Nikon or the seemingly market driven point out behind the Multi-Cam 20K AF system, it still delivers a very solid performance.

That said, since the D850 reuses the D5 and D500’s AF system, all the same caveats apply.

Additionally, like the D5, the D850 includes the built-in auto AF micro adjust functionality. A technology that I think is going to prove far more important on the D850 than it does on the D5 given the demands for focus placement on the camera by the much higher resolution sensor.

As far as Live View AF goes, the situation looks like a bit of a mixed bag.

On the plus side, there’s a pinpoint AF mode that gives macro users an AF box 1/4 the size of the normal one to more precisely focus with. As well as manual focus peaking. Both of these modes are certainly going to be useful additions to the photographers toolbox.

On the other hand, Nikon still doesn’t have an good on-sensor phase AF technology — well that’s not entirely true, they do have embedded phase AF points that they use on the Nikon 1 cameras. This will continue to limit the performance of the live view AF, including when shooting video. Though given the 120 FPS video capabilities, contrast base AF may be pretty quick, though without phase information it will still suffer from the characteristic contrast based hunt at the start. This is one place where both Sony (embeded phase points) and Canon (dual pixel AF acting like one huge phase AF sensor) have a practical lead on Nikon.

Video

Thom Hogan has argued that Nikon needs to stop middling and embrace video or drop it. I don’t see any way Nikon can do the latter and hope to survive. Which ultimately means Nikon needs to stop playing at it.

Fortunately, Nikon seems to finally have stopped playing at it, at least for the most part.

The D850 ticks all the normal video boxes

  • 4K and 1080 progressive recording
  • Full frame and DX/super-35 crop modes
  • All the usual frame rates (24, 25/30, 50/60)
  • High Frame Rate recording at 100/120 FPS in 1080p mode
  • 4K 4:2:2 HDMI output

It even ticks some really nice to have points, like electronic stabilization, highlight weighted metering, and a flat picture control.

However, there are still some points where Nikon is not really seeing the serious or at least cinema side of things. The flat picture control is not a log gamma curve, and so far as I can tell it’s not as capable as one. Frame rates are NTCS (or PAL) standard, meaning 30 FPS is actually 29.97 FPS, and there is no true 24 FPS mode, at least not one mentioned in the specs. 4K is still the consumer UHD standard (3840×2160) not the wider DCI standard (4096×2160)[2]. Finally, while we only have the initial press documents and the product brochure to go on, they still refuse to provide appreciable documentation for bitrates, frame types, and chroma subsampling, for internally recorded video.

Put simply, the video capabilities fo the D850 look significantly better than any previous Nikon camera — and I’m not even going to bother with the 4K and 8K time-lapse modes. Nikon is definitely doing the video thing significantly better than they have in the past. I would even go so far as to say that the D850 has a better set of video features than Canon’s 5D mark IV, well at least as long as you don’t need AF while shooting video.

UI Changes

Historically Nikon has been very conservative at making changes to their camera’s user interface. While I can appreciate how this can make sense form the perspective of continuity between camera models. Sitting here on the sidelines (as a Canon guy) watching Nikon plod along with infuriatingly bad UI designs for the sake of apparent consistency is maddening.

There’s a predictive model of human movement, Fitts’s law, that describes the relationship between the distance a control to be manipulated is from the user’s cursor or finger and the size it needs to be to rapidly be able to interact with it. Fitts’s law must frequently comes up in computer UI design, but it’s been shown to apply to all kinds of user interactions situations.

One consequence of Fitts’s law is that we can look objectively at the organization of controls on a camera and objectively determine if they’re well placed or not.

For example, the shutter release falls directly under the index finger of the right hand when the camera is held normally. It’s also not an unreasonable assertion to state that the shutter release is the most used and most important control on the camera. As such being fairly large and immediately under a finger, the shutter button is an example of a control who’s placement matches it’s importance.

Given it’s proximity to the shutter release, and therefore a finger, Fitts’s law tell us that the area on the top plate of the camera by the top LCD is valuable real-estate for oft changed settings. So what has Nikon historically put there besides the shutter release? The on/off switch, the exposure mode selector, and a exposure compensation button.

Things started to change for the better when Nikon added a reprogrammable movie record button to the right shoulder of the D4. Most serious Nikon shooters, quickly reprogrammed this to be the ISO button and found a serious boost in usability.

The D5, and D500 solidified this move by replacing the old exposure mode selection button with the ISO button proper, and the D850 carries on with this positive design move.

The prime real estate on the top right shoulder now is made up by a reprogrammable movie record button, the ISO button, and the exposure compensation button. A much better overall situation.

D850 top plate detail.
Image courtesy of Nikon USA.

Nikon additionally has continued pretty much all of the the positive UI design points that they’ve had in the past. All the major buttons are backlit, so they can be easily found in the dark — something I do wish Canon would do.

While I still don’t find Nikon’s UI to align with my way of thinking, Nikon has made significant strides in providing users with a much more efficient platform work on in recent years, and the D850 carries on that tradition.

Other Stuff

There are a couple of other points that are worth mentioning but not devoting a huge discussion to. To start with Nikon has redesigned their viewfinder to support 100% coverage at 0.75x magnification. This is the highest magnification viewfinder to date on a Nikon DSLR, and rivals the 0.76x magnification of the Canon EOS 1D mark II. Unfortunately, the D850, like Canon’s 5D series, lacks the extended nose relief built into the D5 and Canon 1Dx series, so you’ll still be squishing your face up against the back of the camera like always.

One other side effect of the new higher magnification viewfinder is the loss of the built in popup flash the D810 had.

The popup flash on a camera like the D850 is in some ways kind of silly. It’s low power, and small, which makes it almost useless for flash applications. However, the one place it really shined is in controlling other Nikon flashes in an off camera configuration. The tiny built in flash, was all you needed to get 3 zone lighting working in very creative ways, and you didn’t need a big heavy flash on your camera to do this (or really high end Nikon flashes for that matter).

Nikon, like Canon, has move the focus of their flash control to a radio frequency based system now. However, slightly older models (such as the SB-910) do not support RF operation. Moreover, even the current top of the line model, the SB-5000, still supports optical control. Point being, even though the D850 has RF capabilities (Bluetooth 4.1 and 802.11b/g Wi-Fi), none of those support controlling Nikon’s flashes. So it’s back to having a flash in the hot shoe to do that[3].

Speaking of Wi-Fi, Nikon also doesn’t appear to be include the fastest 2.4 GHz standard 802.11N, instead it maxes out at 802.11G. Potentially 802.11n would offer more bandwidth to transfer images faster, which certainly would be useful with the large raw images the D850 will produce. However, that comes with a cost in power consumption that may not be worth it. In a perfect world, a D850 on 802.11g (54Mbps) could transfer a 60MB raw file in about 10 seconds. In practice, I would expect transfers to take twice as long.

That said, while the utility of hanging 802.11n support can be argued, the utility of having Wi-Fi available for tethering is an amazingly useful thing. Admittedly, I had been on the fence about it, seeing some use but not seeing it as a serious feature. The reality is, for anybody who shoots tethered in a studio, not having a wire to worry about taping down or tripping over, Wi-Fi is an amazing thing to have.

Finally, there’s the tilting touch sensitive screen. Both things that are immensely useful in a modern DLSR, and both are things that have been a long time coming for high end cameras. Tilting screens are especially useful for landscape photographers who want to shoot with a low camera angles and still want to be able to see what the camera is doing.

Grumping a Tiny Bit

Some people would accuse me of being biased against Nikon because I’m a Canon guy. The truth is, I’ve seen enough marketing and cameras to be jaded about the whole marketing show. Plus I have a background in engineering, which, if it taught me anything, taught me that everything is a compromise and comes with trade offs.

What I’m trying to say is that there’s no end all be all perfect product, and never will be. So there’s always something to complain about.

Let me be blunt, Nikon has done a lot of really good stuff in the D850. It’s not so compelling that I’d be willing to switch platform for it — remember, I’m a firm believer in the reality that the photographer is far more important than the camera.

That said, there are still some points to fault them on, though honestly not that many. Okay for me there’s really only 1 point.

To there’s no integral GPS.

The D850 is poised to be a fantastic general purpose camera, and part of that is going to be landscape, wildlife, and travel work. All of which certainly can be done without a GPS, or while using a accessory, like Nikon’s GP-1A. Though having to use an external accessory is an unnecessary hassle and a mediocre experience at best.

Having tried to figure out a decent way to get a GPS talking to my camera over here in Canon land — prior to getting my 5D mark IV which has a GPS — there’s simply no comparison to having an internal GPS unit[4]. Nikon continuing to eschew built in GPS support on the D850 is disappointing to say the least.

Final Thoughts

Admittedly this entire piece is based entirely on publicly know informational and preliminary data that’s been made available on various sites around the net. However, the short of what I’m seeing is that Nikon has as solid and eminently capable camera on their hands here.

Is it too late to stem the tide of ebbing market share? That’s a more difficult question.

Let met start with what I hope; I hope it does. While my Nikon shooting friends occasionally accuse me of hating Nikon because I’m frequently critical of them, I really do dislike Sony. So anything that stems the tide of anything Sony flowing into the market is a great thing as far as I’m concerned (and yes, I know Nikon is deeply entangled with Sony for sensors and that’s especially now XQD cards).

What I hope, though, isn’t a good answer to the question. The outstanding question in Nikon’s loss of market share ultimately comes down to this: Was Nikon’s problem the age (apparent or real) of their products relative to Sony’s or is there a fundamental shift in high end user’s desires away from larger DSLRs towards smaller mirrorless ILCs?

The D810, while certainly a better camera than the D800 in most, if not all respects, didn’t have so much of a obvious difference that it easily appeared to be a big step forward to someone who wasn’t invested in every technical detail. I know many D800 owners who felt no need to upgrade to the D810 and that certainly further held back Nikon’s sales. In many ways, it’s certainly understandable to feel that Nikon hasn’t upgraded their D800 series cameras since the D800’s release in 2012.

Simultaneously, Sony’s Alpha line of full frame mirrorless cameras appeared on the scene later (2013) and were somewhat rapidly iterated on with mark 2 versions released in 2014 and 2015. This put them in a much more up to date looking position compared to the D810 at least form a paper specs perspective.

A the same time, Sony’s mirrorless cameras offered buyers a smaller camera, which some certainly considered a major advantage.

That said, I think the D850 is going to sell like hot cakes, and it’s definitely the camera Nikon needed. It’s a major overhaul to the D800 line in a way that didn’t happen with the D810. It puts Nikon back in a good place being competitive with Canon and Sony for resolution. And appears to maintain the D810’s dynamic range and overall image quality position, arguably at the top of the pack.

Put bluntly, this is the kind of camera I expect form Nikon and the kind of camera they should be delivering if they want people to continue to stay with the platform.


  1. At ISO 100 the estimated PDR of the D850 is 10.96 Ev, compared to the D810’s 11.06 Ev. ↩︎
  2. And while I would really like Canon to include a UHD format on their cameras instead of only using the DCI standard, it is at least possible with minimal post process copping, cut DCI 4K down to UDH. The inverse, however, is not possible. ↩︎
  3. Yes, I know Nikon has the WR-R10 RF controller that you can use on camera to control their RF flashes, that still doesn’t enable one to use their old SB-900/SB-910 flashes without a flash in the D850’s hot shoe. ↩︎
  4. And Canon’s GP-E2 has some serious advantages over Nikon’s GP-1A unit, namely the ability to communicate via the hot shoe so there’s no need to plug cords, that can get yanked out, into a no-longer weather sealed terminal on the side of the camera. ↩︎

Comments

There are no comments on this article yet. Why don't you start the discussion?

Our cookie and privacy policy. Dismiss