Dpreview published their Nikon D850 autofocus test results

Nikon D850 MultiCAM AF sensor module

Dpreview published their Nikon D850 autofocus test results. Their conclusion:

"Appropriately for a highly capable camera, the D850 comes with a highly capable autofocus system. In single-point Single AF and Continuous AF, as well as in its Dynamic area settings, accuracy, hit rate and precision all rank pretty high. However, the D850's massive resolution shows up some inherent issues in terms of how DSLR autofocus systems work that might prevent you from getting all 'keepers' all the time, especially if your lens requires calibration. This shows up most often in wide-aperture work, such as some styles of portraiture, where autofocus precision is of paramount importance.

In terms of subject tracking, we continue to be perplexed by our experiences with 3D Tracking - but we must admit that, during burst shooting, the D850 pulls away handily when compared to its predecessor, the D810, as well as the Canon 5Ds R and Sony a7R II (though the Sony tracks better in Eye AF mode as long as you're not shooting bursts). However, despite Nikon's claims of D5 levels of performance, we collectively find the D850 to be somewhat lagging behind that lofty benchmark."

Related post:

Nikon D850 camera autofocus (AF) coverage


Check D850 availability: B&H | Adorama | Amazon | BuyDig | BestBuy | Cameta Camera | Focus Camera | eBay | WEX | Jessops

Follow D850 on Facebook: Nikon D850 Page | Nikon D850 Group

This entry was posted in Nikon D850. Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • NJP

    I would be surprised if it was equally capable as the D5 with autofocus. That would make the people who spent the money on a D% very upset.

    • harvey

      you got it and as well, basically terminate all future D5 sales.

      • TurtleCat

        Not really, different tools for different jobs.

        • harvey

          ok so you get a camera with the same AF speed, does 9fps instead of 12 and it has 47mp instead of 20 but can give you an APS crop of about 20mp – downside higher noise at high ISOs? Best part: you can buy two for the same price as the one. Or one body and the 70-200e. ps. we won’t even talk about the 2 stop DR difference.

          • Vince Vinnyp

            I ran D5s alongside D810s. I will still run D5s alongside the D850. The D810 already was the goto camera for most situations and the D850 will be preferred even more often, but there will still be situations where the D5 is a much better choice.

          • TurtleCat

            Ok, so you have one tool optimized for high speed and better ISO and a robust body that can withstand most anything and you have one tool which is a generalist. That 2 stop DR is not through the entire ISO range so drop that useless comparison. Different tools. Both are great cameras but the D850 does not replace the D5 just as the D800 didn’t replace the D4.

            • harvey

              the D5 occupies a very small performance niche and yet some will buy it simply because of its high price status – why folks bought the D4 to shoot weddings(also the concept of having a more expensive camera than Uncle Bob). The 2 stop DR is not useless to the people who need it. The D800 never came close to the D4 in many aspects. Its claim was hi-res. The D850 covers many more bases as is. Giving it the same AF capabilities as the D5 would only allow it to claim more of the D5 footprint.

            • TurtleCat

              And many will buy the D850 for its high resolution status, what difference does that make? How many people are using a D5 for landscape? Sheesh. The D800 was introduced with many of the same focusing and metering capabilities as the D4 but slightly less. Similar to now. In any case your original point was that the D850 would eliminate D5 sales. You haven’t proved it.

            • harvey

              I said a D850 with the same AF as the D5 – read more, talk less. Besides, there is the example of the D3/D700.

            • TurtleCat

              And the D3 still sold, even when the D700 came out. Of course the D3s was on its heels. Your words were “you got it and as well, basically terminate all future D5 sales.” and you did not have any disclaimer or conditions.

            • harvey

              did you read the OP’s comment? The thing that my response was to? As for the D3/D700, when the D3s came out, why didn’t Nikon come out with a D700s or the D3s ‘lite’? They knew. It would have to wait until the D600/D750 for another lower priced FF model.

            • TurtleCat

              I did and his comment was pretty broad. And you said “and as well…” In any case, I’m done with this pointless back and forth.

          • Duncan Dimanche

            the higher dynamic range is for the very lower ISO and the Higher ISO on the D5 is still way better than the D850…. but yeah all the other points are valid 😉

    • Your missing the point here , it was sold stating it’s the same as the D5 , if these tests didn’t come about people would be still angry and sell thier D5s , Nikon didn’t expect this fuss, and expected D5 owners to jump to a D850

      • Mike

        No. They say it’s based on the D5 AF.

      • Vince Vinnyp

        This is not new, The D810 shares the AF with the D4 but they don’t perform the same. The D5 has the same AF as the D500 but again some performance differences. The D5 now even has 2 additional AF modes to the D850 (& the D500).

        • BdV

          I can still understand why people would be disappointed. Why shouldn’t any of the differences between D850 and D500 AF performance be in favor of the higher end camera? Similar AF modes can’t be expected to reach at least identical performance as a previously released cheaper model?

          • Vince Vinnyp

            I am not sure there are any differences to the D500 this is only about the D5

      • Robert Falconer

        It was sold saying it uses the same AF *system* as the D5, but Nikon never claimed that it had identical performance to the D5 (or D500 for that matter) in all shooting scenarios. This is where it’s up to the consumer to do their due diligence. All manufacturers will try to put the best spin on their products that they can…it’s always been that way.

        Caveat emptor (aka do your research carefully before buying).

      • António

        Imagine you’re a car builder and you use the same engine (exactly the same, with no modifications) to build two different cars: one to get the most reduced weight as possible and the other with heavy bullet proof iron and glasses, ending up being 1.5 tones heavier than the other.

        Will you expect them to equally fast going from 0 to 100 Km per hour?

        Doesn’t the sensors “size” and other possible differences in the electronics be reasons for the two cameras not showing exactly the same behavior?

    • ITN

      No, it wouldn’t. D5 has its own feature set, including ruggedness, ergonomic advantages, built in Ethernet support, ftp uploads to the Internet, convenient access to wifi (with a tiny module added to the side, compared to the huge one required for D850), better high ISO image quality, 12-14fps, greater burst depth etc. The D850 is a camera for different purposes and many D5 users will buy D850’s as well. Everyone who uses AF wants the best AF they can get. There is no point at all in making it less. D5 sales at this point are completely negligible compared to D850 sales anyway; the majority of those who were ever going to buy a D5 new got it at least one year ago.

      There are a number of possibilities one is simply bugs in the code, then there is the influence of the different battery, optimization for high resolution rather than speed, bottlenecks in the processing pathway etc. etc. none of these involve deliberate crippling of their star camera (yes, which is the D850, not the D5).

      • harvey

        not everybody can buy every body. Sometimes you have to buy that which will fit the most of your prospective needs.

    • Azimuth1

      it’s not about money but tech. What can be very accurate for a 20MP image, it _might_ be not that accurate for 46MP. It’s a technological limit and the article points it out very clearly, at least to me.

  • Huh, I thought the D850 had the best Eye tracking ( like D5 ). Weird the older Sony A7RII tracks better…. Nailing the eye focus with f/1.4 at 85mm is a third of my work.

    • Thom Hogan

      Every autofocus test is as much a test of the shooter and their understanding of the way the system works as it is the system itself.

      I’d say that for someone casually using the system, the Sony Eye Detection probably works better at getting eye focus than any mode they might try on a Nikon DSLR because the Nikon will require them to understand what that mode is trying to do while the Sony will do it for them (within some limits).

      I don’t really know someone I’d call a true professional who’s having issues with the D500, D850, or D5 in terms of focus, though. We all had a few struggles early on, as the responses and options are different, and the actual AF sensor sizes/placements are different. But I think we all just adjusted through practice.

      • Johnny in Philly

        I tried face-priority, AF-S on 85 @1.4, and it got it right *mostly* – but would love to learn more. Any pointers to understand the D850’s various focus modes?

      • “Every autofocus test is as much a test of the shooter and their understanding of the way the system works as it is the system itself.” Great point. There’s a reason I have several tweaks to my menu setups to get my keeper rate high.

        “Sony will do it for them”, another solid point. Sometimes people just want to be lazy and have it done for them, I’m guilty of that, especially with event shooting when many changes are being done to maximize AF settings.

        I do prefer to use AF-C, single point, AF-On only, short lock on; personally yields me the best results. But, always requires moving the AF point to suit the composition and model eye placement. And, at times, I run out of AF point area, because the pattern isn’t to the near borders. So, full sensor eye AF could be quite handy in some situations.

      • ninpou_kobanashi

        There’s also visual confirmation that it’s tracking on the mirrorless side. With DSLRs, you have to check after the fact.

        That being said, my A7RII has problems with the faces I tend to look at (cosplay, hats, props, etc.)

        • Thom Hogan

          I suspect you’re conflating DOF with focus. Both EVF and OVF systems show where the focus is during shooting. The EVF system can (for most systems) show that at shooting aperture, the OVF systems show it at max aperture.

          • El Aura

            Re OVF system: Or with fast primes at an ‘intermediate’ aperture (about f/2.8).

          • ninpou_kobanashi

            Hi Thom,
            actually, I wasn’t talking about DOF, but maybe I’m misunderstanding what’s happening.

            On my Sony, the faces are marked by a square that looks different from the AF points. When using eye focus, the square becomes much smaller and is drawn around the eye.

            On my Nikon when shooting AF-S Auto or AF-C 3D, when pointing at a face, I tend to get a lot of squares that light up on a face but I don’t really know which point it’s going to use (i.e. will it really pull out the eye?).

            That’s what I meant.

            • Thom Hogan

              Ah, okay.

              Yes, the DSLR feedback on exactly where the focus is being put is loose. But it’s unclear exactly what the camera is telling you (and by the way, the Sony A’s have the same problem in the all-auto modes without eye detect enabled).

      • Proto

        So, what different style of focusing habit is needed for D850 compared to D800 or D810?

        • Thom Hogan

          Still working on that. But the quickness of the Dynamic AF modes changes the way most of us use it on the D5/D500, and that would be true of the D850, too.

  • ZZ

    Wha da you mean it doesn’t AF as good as D5!!?? ….. the outrage!!

    • Todd

      Well I mean it’s got the identical AF system to the D5.. so people are wondering why it doesn’t perform the same….

      One can only assume the massive resolution has something to do with that

      • j j

        It is the same *hardware*. There are layer upon layer upon layer of software piled on top to drive it all.

        Anyone who thinks a $3000 camera will function to the level of a $6000 camera has got rocks in their head.

      • While the AF system is the same, the “engine” is different. In this case, that means the battery. The larger battery on the D5 can drive the focus system faster than the battery on the D850. I noticed this in using a D3X and the old AF (not AF-S) 80-400 lens. I had a good hit ratio with that combo, but nowhere near as good a hit ratio with that lens on smaller battery cameras.

    • br0xibear

      Ah, you missed the asterisk, lol.
      *”No matter how fast the action, the camera does not compromise. The broadwidth of the autofocus coverage and its impressive 3D tracking mode offer unparalleled accuracy when tracking subjects around the frame.”

      *(apart from the D5 which is better)

  • Nikkor300f4VR

    Drink!

  • bobgrant

    With all of the little shortcomings, let’s not lose focus ourselves. The AF is a bit shy of the D5? Sony can track better at times? Mirrorless is more compact of course….but the reality is that the Nikon D850 is the best all around DSLR on the market. It may miss some tiny specific requirement for a minimum of advanced shooters, but they will be the exceptions. This is the camera everyone wanted before they even knew a camera could be this capable. Nikon hit it out of the park.

    • Benno Hessler

      !!!! wtf…

    • On the money, except that “more compact” bit. Lighter, maybe. More compact? Doesn’t look that way to me when comparing same for same with lens/battery (mirrorless uses them up faster so you need more or grip).

      But yeah, this is all of what I’ve been asking for.

      If they included AF with the mirror up comparable to the native mirrorless cameras, they’d have a win on all fronts.

    • fanboy fagz

      the camera can cater to every photographer and have minimum compromises doing it.

      • The right way to describe D850.

  • lucian

    I bought my D5 because was not good in colours, contrast and white balance ! I am afraid to buy another new Nikon ! I don’t know whats happened with Nikon but in privat I can send some photos taken with D5 and you will see it ! So my D4 is better in photos ! And with biggest ISO to D5 is a problem too: not work as we waiting for ! see my compares with D4 !

    • Benno Hessler

      !!!!! wtf..

    • TL Robinson

      Huh? I think you’re using your camera wrong.

      • lucian

        No, no I only use profi camera Nikon over the years : from D2, D2H, D2X, D2xs, D3, D3s , D4 , D4s to D5 ! compare yourself one D4 vs D5 and you will see ..

        • Vince Vinnyp

          I have compared! I had all the cameras you have listed as well. The D3s output in low light was marginally better than the D4 and even D4s so I kept one for some occasions. The D5 output in the real world at high ISO finally means I only need the D5s now for high ISO. It is a different animal, far from perfect, but play to it’s strengths and it gets shots that the others, plus the D850 never could.

          • TL Robinson

            Yeah I’m not sure about this guy (Lucien)…He even states the D5 is a problem at “biggest iso” – there’s a ton of evidence both real world use and reviews that say the exact opposite. I held off on the D850 to see where it sat on the spectrum and glad I did. The D5 is now my next purchase. I stand by my statement he is using his camera wrong. Just because he only uses “profi” camera doesn’t mean he’s proficient at using them…..

            • Michiel953

              And here I was thinking that shelling out to get a profi camera would make me proficient.

              Oh well.

            • Lucien

              I said: send me one email address to send you some pictures ! I am not the olny ! see in ebay, more peoples try to sell used D5 at half price ! and buy back D4s

            • Lucien

              No, no, you mistake ! you will see ! no problem…

            • Lucien

              hey men, look at here what he write in ebay , he want to sell D5 because : ” Achtung: Der Autofokus der Kamera ist leicht dezentriert. Dies bedeutet
              der AF sitzt nicht mittig sondern ganz leicht versetzt. Die Funktion des AF ist dadurch jedoch nicht eingeschränkt.”
              So you have not right, D5 is not good ! D4, D4s is best ! go in ebay and see !

          • Lucien

            If you write your email adress I will send one real iso test with my D4 vs D5. Right, D3s I still have today in my old collection and the result in low light is a little bit better than D4 and my D5 ! I am not the only person which I had problem with D5, you can read in dpdreviews D5, others…

            • TL Robinson

              Just send my your D5 and I’m pretty sure I can make it sing in low light…. 😉 I don’t hand out my email address to random people on the ‘net sorry….

            • Lucien

              ok, buy one D5 then you will convince yourself! Just remember this discussion!

          • Lucien

            hey men, look at here what he write in ebay , he want to sell D5 because : ” Achtung: Der Autofokus der Kamera ist leicht dezentriert. Dies bedeutet
            der AF sitzt nicht mittig sondern ganz leicht versetzt. Die Funktion des AF ist dadurch jedoch nicht eingeschränkt.”
            So you have not right, D5 is not good ! D4, D4s is best ! go in ebay and see !

        • CommonPleas

          Yeah, no. I am no pro, but I thought the same with my D700 until I figured out that my settings needed adjustment. Worked. Colour, contrast, w/b, all came up to what I expected.

          And it depended on the colour space selected and printer I was using, too.

          It just took time to figure out the camera. In other words, it’s me, not the camera.

    • Gary Howarth

      Lucifer, it must be over heating issue ?

    • outkasted

      I agree Lucien the D5 just did not do it for me after seeing the results between the D4-D4s,D5 image wise I was not too excited although I have been waiting on the D5 with baited breath.. Hi ISO is impressive indeed but sacrificing on low iso DR is truly a lost opportunity to get it right. I would have honestly thought that D5 would be stop better at hi ISo natively but truly give really great results at low iso’s

  • Well, it’s being tested with the stock battery, not the EN-EL18. I’m not at all surprised. Seems like they played it more conservative for it’s use of the smaller battery, after the D500 battery fiasco.

    Wake me when someone tests it with the right battery…

    • Vince Vinnyp

      I have tested it with the grip and EL18a battery at 9Fps and it’s still not quite as good, it’s not a big deal and pretty close. This is all relative, we are comparing one 3D tracking mode against, almost certainly, the best AF camera currently available and the D850 is almost the same. It’s equal in all other modes.

      • How’s it compare to the D500?

        • Vince Vinnyp

          I don’t own a D500 sorry, from what I understand though it’s about the same, as there was a slight difference between the D5 and D500.

  • Matt Comerford

    I want my D850 :'(

  • Jim Huang

    I wonder why though, since they have the same AF module and processors.

    • Different software/internal-configuration – pixel density is different on D5 and D850. Circle of confusion is smaller in D850, probably Nikon tuned the config so the AF is more precise, sacrificing speed in the process.

  • Eric Calabros

    There is a fact that technically they can’t compare the AF with another body. Because the other body doesn’t see the subject at the same time or from the same angle of view. So they see different things with different movements.

    • Captain Megaton

      Statistical averaging.

  • Mistral75

    The tracking capability of the Nikon D850 is indeed quite poor :).

    Tamron SP 150-600mm f/5-6.3 Di VC USD (1st gen.) 500mm f/8 1/1,600s, ISO 8,000

    https://uploads.disquscdn.com/images/e8dddd7f0ad731fa7a39ff9d9431e772b82a71e6af1b6aebbb5c645787a28105.jpg

    Source: https://www.chassimages.com/forum/index.php/topic,274835.msg6542324.html#msg6542324

    • Allan

      Very nice. Thanks for including the EXIF data.

      { How do you say sarcastic in French? 🙂 }

      • Mistral75

        Sarcastique. Same pronunciation as the English word: sahr-cas-tik, save for the tonal accent on ‘tik’.

      • T.I.M

        TIM (for short).

        • Allan

          Hi, T.I.M.

    • wow!

    • Aldo

      You can go to the moon with 1960’s technology =] … nice shot

    • Roger S

      Fantastic shot. I went to your site and learned that the subject is traveling at close to his maximum speed = 300 km/hr! Imagine what the D850 could capture if it had decent tracking capability :’)

      • Proto

        at that speed, did the bird crash into the ground or was the photog in a tall building – taking the photo from well above ground?

        btw, speed of human free fall = 200 km/hr.

        • Roger S

          I think you raise interesting points, so out of curiosity I did a bit of research.

          The speed of 300 km/hr was quoted by the poster in the source cited for the photo, and is correct according to various sources I checked. The peregrine falcon dives steeply (as in this photo) at speeds over 320 km/hr and have in fact been recorded diving at speeds higher than that. Google peregrine falcons for info on how falcons are able to avoid damaging themselves in the process of diving at speed. I’m not a BIF shooter so I don’t know how to catch a falcon in the middle of a steep dive. I am sure that others do know, however (as this photo demonstrates), and the latter strikes me as a remarkable accomplishment.

          Terminal velocity for a human being usually assumes a random or spread-eagle position and takes into account the considerable wind resistance involved. A human being falling in “bullet” position, curled up in a ball, however, can apparently reach a terminal velocity over 330 km/hr — even faster than an average falcon diving.

      • Naacryl

        you don’t need AF for these birds, manual focus works just just fine

        you can just manual focus on peregrine when they hover and circle their prey, wait a few seconds and they’ll start their dive

        with f/8.0 you just follow the dive and you take a shot, not that complicated to photograph birds of prey

        peregrine are in most cities nowadays, just have to look out of your appartment and you can shoot them, thousands of diving peregrine pics online

        I don’t mean to attack the picture, but this is about camera hardware, peregrine don’t require “tracking”, they’re not hard to take a pic of, small songbirds in the sky are way harder to take a picture from

    • Ben Brayev

      coulve done the same thing with a d90+70-300.

      small birds that fly towards you and further away from you erratically is a complete different story.. tracking falcons and crows is easy..

      • Tuna

        You’ve got to be kidding. Crows indeed are easy, flying in a relatively straight line at 15-2o mph. Easy to pan and fill the frame with a bird 50 feet away flying slowly in a straight line. Its much more difficult to take a picture of a peregrine or gyrfalcon in a full stoop taken fairly close to the bird (within 100-150 feet). Look at the Peregrine’s nostril in the picture. You can clearly see the nostril cone. To get that level of detail, it means it was taken with the bird fairly close. Panning on a bird moving that close and that fast is extremely hard. The bird is in a full stoop meaning it is likely moving somewhere between 70-250 mph. They don’t dive in a straight line, they adjust there direction in response to their prey’s evasive maneuvers.
        And it was not flying toward the shooter, but diving vertically from the frame of reference. That would be hard for any autofocus. Whoever took this picture is more than good, as is the autofocus system in the camera. I’m extremely impressed by both the photographer and the camera.

        • Naacryl

          he’s right

          crows and birds of prey are the easiest things to track, you can just use manual focus

          that’s a peregrine and they just dive straight down and their level flight is usually slow and not erratic

          you have tons of time to track a peregrine before its stoop, usually they fly in pairs and they hover around extremely slowly

          some harder birds are birds like a wren in the sky, cameras that can track wrens,…now that would impress me

    • Gabriel Border

      Nice shot, looks like my peregrine.

    • Tuna

      Nice Peregrine butt. Nice to know d850 autofocus can keep up with fastest animal on earth! Too bad no human can pan that fast. Think this picture shows what autofocus can do.

      • Tuna

        Oops, didn’t open whole image. Wish I could pan like you. Seriously, from experience this answers any questions about how good the autofocus works on the D850. Very few cameras are capable of following a bird that can move erratically at over 200 mph. Given how tight the wings are folded, that bird is moving!!! What were your autofocus settings?

    • ISO 8000?!?!
      Looks at D810 on desk forlornly…
      That looks like ISO 4000-5000 on this three year old body, that doubling of performance was no joke.

    • ModifiedJason

      Stunning.

    • thundrrd

      This is not only a testament of the D850 focusing tracking capabilities, but also of the Tamron SP 150-600mm lens being an OK lens for shooting Birds (not literally shooting birds).

    • TheMeckMan

      That’s horrible. The 54th feather on the leading edge is slightly out of focus…

    • Naacryl

      birds of prey are the easiest things to track

      try a smaller songbird

  • Adam Brown

    Interesting….. the “problems” with getting perfect accuracy in the portrait examples. This is where mirrorless can shine — and maybe eventually a mirrorless version of the D850.
    You really don’t get back/front focus issues when using AF-S in mirrorless cameras. You really can nail perfect accuracy nearly 100% of the time.

    • Robert Falconer

      I think it’s safe to say that mirrorless AF will eventually overtake DSLR autofocus in all categories. It’s just a matter of time.

      • Cynog

        The one thing difficult to overcome is that when one’s eyes are accustomed to a bright, sunny day and you look through the electronic viewfinder, you can’t see a thing for several seconds. That’s the main reason I prefer OVFs. Also, the dynamic range of the EVF is nowhere close to that of the human eye, so shadow detail is lacking. Maybe these can be overcome in time, and perhaps mirrorless will take over in all categories of photography, but I’m old enough and have sufficient DSLR cameras to not worry about it.

  • sickheadache

    This AF testing should be a true non issue…when using Manual Lenses. OR even Manual on the Camera. See what I did there.

  • Russell Ferris

    I swear by the split-image focusing screen in my D800, but this AF module does look like quite the improvement.

  • This is the real world.

    Do you remember last years washing powder, well this years washes whiter, as it has for fifty years, ……..

    In the real world, no one notices, we do not live in a lab

  • Originaru

    i wonder if they excluded the resolution difference which makes more or less noticeable little errors that can be apssable a lower rez.

  • Back to top