Nikon designed the D850 sensor themselves


Imaging Resource published some interesting D850 facts they got directly from Japan:

While they didn't attribute it to the BSI sensor, Nikon told us that the D850 should produce the same image quality (both JPEG and RAW) at twice the ISOs as the D810, a full-stop improvement. That is, the D850 at its top "native" ISO of 25,600 should deliver the same image quality as the D810 did at ISO 12,800. If true, that's a pretty significant improvement; we can't wait to get our hands on a sample to check it out!

While Nikon contracts with a silicon foundry to actually manufacture the chips, Nikon confirmed that the D850's sensor is entirely their own design, vs. an off-the-shelf unit from a sensor manufacturer.

Nikon says dynamic range will be as good or better than that of the D810, despite the higher pixel count.

Pre-orders: B&H | Adorama | Amazon | BuyDig | WEX | Jessops

Nikon D850 Facebook Page | Nikon D850 Facebook Group

Nikon D850 directory

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

    OK the notification works perfectly this time!

    • akkual

      The video they lauched had a scene shot at ISO3200. Looked crisp, noiseless, and dynamic on my 30″ 2K monitor. Assuming they didn’t fake it.

  • Jeffrey G. Bank

    It happened to me too.

  • Dave Silo

    I really hope the DR is a little better than the 810. I will be completely happy if it is the same but I just hope its not worst. It actually scares me a little nikon made the sensor.

    • Why that fear?

      In fact they don’t “make” it, but they designed it.

      Anyway, I’m not afraid of that. Instead, Nikon’s sensors have been among the best performers since time ago.

      • Dave Silo

        if I understand correctly Sony has designed the past few sensors that gave great DR. If this is Nikons first attempt at the high MP large DR,I worry they won’t be able to get the same results. I don’t always trust what a company says, so I guess we will have to wait and see what a independent test shows.

        • Agree on the “wait and see”, but I trust Nikon in the sensor performance 🙂

        • Xam

          Incorrect. Nikon designed the d800, e & d810 Sony just fabricated the sensors.

          • Matt Grum

            No they didn’t, the D800 is a Sony IMX094, same as in the A7R as confirmed by teardowns of the cameras. Unless you’re claiming Nikon designed the A7R sensor!

      • Kosol Onwudinjor

        Most of Nikon cameras carry Sony’s sensors

        • Well, perhaps this one will carry a Sony one 🙂 Anyway I don’t think the maker is that important, given that the D7200’s one is Toshiba and is quite a good performer.

          Regarding design… I don’t know what’s going on in the kitchen (I’m far from that, a simple user…), but as I pointed I’m not afraid and trust in Nikon.

          See the D5’s sensor. A lot of people say isn’t a good sensor because of the (relatively) low DR at base ISO. Kind of “Nikon couldn’t made it better”. But I see it as low ISO DR isn’t as important in a action camera as in a landscape camera, and probably Nikon choose to sacrifice it for another parameter. In fact Nikon got a better (and not bad) low ISO DR in the 16 MP sensor of the D4/Df, and it wasn’t a Sony sensor.

          I think Nikon knows what it does in the sensor area. I won’t say the same in other fields…

          • I think I read that Sony bought the Toshiba sensor chip business a few years ago.

    • Kosol Onwudinjor

      The D810 has been quite impressive. I would consider it a great benchmark for image quality. If Nikon says the D850 will be as good or better than the D810, then that’s remarkable for a high megapixel camera.

  • Riooso

    For us Landscape guys this think has a Low Pass Filter… the very reason that I bought the D810 is that it did not have the LP Filter that the D800 did. With the filter comes some lack in sharpness. So I don’t see this unit working for me.

    • Dave Silo

      it doesn’t have a low pass filter

      • Dave

        And the d800e did not have the low pass filter.

        • Kosol Onwudinjor

          I thought there was a low pass filter cancellation on the D800E

          • Robert Garfinkle

            true

          • David

            That is correct. They filtered one axis and then the next component un-filtered if, if I’m not mistaken. The D810 removed that system altogether.

    • Albert

      What are you talking about? The D850 has no Low-Pass filter.

    • Evan Richardson

      Someone didn’t read the product information

      • Riooso

        I just read that one does not have to worry about moire patterns. The only way, up to now, was to put on a low pass filter!

        • David

          The more sampling you get, the less you have to worry about moire in theory.

    • Fly Moon

      You weren’t planning to buy it any way. If you were you’d have read the specs about the sensor. It was on top.

      • Riooso

        I did read the specs and they said no AA, but they did say that you did not need to worry about Moire which is usually accomplished by a Low Pass filter. Another review I read said it did have a Low Pass filter so we really will no know till it come out.

  • Nikkor300f4VR

    Just Wow! Can’t wait to see that!

  • well isn’t this sparkly 🙂

  • John C

    That BSI sensor comes from Samsung technology. The flip screen shows the Samsung NX1 connection also.

    • You sure? That Samsung making the sensor would be something really new and with implications in the future. I don’t see it as bad or good, but as different.

      • jsvfoto

        That would indeed be interesting. There was a rumor a while back that Nikon was buying Samsung sensor/mirrorless tech…it turned out to be a hoax, but who knows maybe there really was a meeting and this sensor is designed by Samsung.

        • no, this was 100% hoax/clickbait/fake news – just like the Sony buying Nikon and Fuji buying Nikon crap coming from the same place

  • Steve

    Nikon says dynamic range will be as good or better than that of the D810, despite the higher pixel count.

    Laugh, they don’t know or what? They designed it, seems like a dumb comment.

    • neonspark

      Nikon is a huge company. The guy that knows isn’t the guy that’s commenting.

      • Steve

        Still, you’d think they’d ask someone that actually knew what they where talking about, maybe someone from the sensor design team? Wow I am amazed at the level of fanboyism here sometimes. I guess it’s to be expected.

        • David

          It can also depend on the ISO you’re measuring. It may not be that it’s one stop better across the entire ISO range, for instance. A D600 is still better than a D5 if you’re shooting at base ISO.

          • Steve

            Different ISOs have different DR, wow this is all news to me? Should I use a tripod for landscapes as well and stop down to f/64? Would you recommend a 24mm 1.4G for macro? Thanks in advance.

            • In a “ISO less” type sensor (almost all in Nikon’s catalog except the D5 now), normally the higher the ISO the less the DR.

            • Steve

              I’m aware, but not the point.

            • David

              Haha if you’ve got a lens that stops down to f/64 then I don’t think you’re gonna be using a systme that lets you worry about high ISOs…and a 24mm 1.4G would be an okay macro lens on a Nikon 1 actually, especially with an extension tube.

            • David

              Also water is wet.

            • Pat Mann

              Only if you reverse it. Then you need a way to run the diaphragm – perhaps Nikon will do a reversing cable setup on their next bellows.

    • silmasan

      Because you’re not comparing a single value. The guy might be saying that to cover, for example, base ISO (“as good”) and high ISO (“better”).

    • MyrddinWilt

      The dynamic range achieved actually changes over the course of a production run. They will know what the upper bound on performance is but not how many devices will meet that performance.

      One of the reasons that Nikon came out with the D700 and Df1 was that the sensors produced come out on a curve. They put the best ones in the flagship cameras and the ones that didn’t quite make the cut go in the lower priced option.

      So no, they are not going to know exactly what the performance of the production sensors is going to be.

      • David

        CITATION NEEDED!

        …seriously, I’m really interested if you can provide an article that goes more into detail about that.

    • jtang128

      I agree what you said, Nikon designed Nikon D850 sensor is more brown and black color like mud and lack of dynamic range. it can’t be compare with D810 sensor.

      • jtang128

        I will pay the price $3300 if Nikon D850 have D810 sensor in the camera

  • Marco

    1 stop…if true, It is going to change my religious beliefs

    • Well, in fact the D810 sensor wasn’t all that good at high ISOs, IMHO.

      • Neopulse

        Better than the competition it was and much better than the medium format sensors that were costing in the $15,000+ range.

      • akkual

        It isn’t, if looked at 1:1. It’s higher score in dxomark is due to dxomark making their verdict on scaling down to 8 mpix print. D750 has in fact far better sensor at 1:1 than D810 had, but the scaling down to 8mpix evens it out due to oversampling theorem. Also, it’s one of the reasons, why D5 did so “bad”. D5 is exceptionally good at ISO 3200-> at 1:1, but being “poor” at low ISOs and being low res, it didn’t get any help from downscaling either. That is, if D850 sensor pulls same values at 1:1 that D810 did, it will have “better” DR in practice than D810 had and likely hit the top at dxomark.

  • neonspark

    I don’t mind if it uses Sony sensors. I’d be outraged if they used Canon’s for Sony is the better choice across all tiers.

    • MyrddinWilt

      Saying that Nikon designed it is missing the point. It is the production capabilities of Sony’s fab that determine minimum feature size and many of the other properties of the device.

      That said, Nikon own a stake in the ‘Sony’ fab. As is common in Japan, there are complex links between the businesses that operate semi-independently.

  • Eric Calabros

    One stop in raw is overestimate.

    • David

      It might not be. Thom Hogan said in his D500 review that the electron well of the D500 was twice as good as that of the D7200, which in theory should correspond to a one stop improvement (assuming the D500 sensor was the inspiration for this one). Pretty sure the D800/D810 sensor was more advanced than the 16/24mp DX sensors were but still you might see half a stop across the board and a stop at certain ISOs.

  • yes, every report I’ve seen confirmed that

  • I like the use of “suspects” 🙂

    In fact, if the sensor is really “ISO less”, that high ISOs (about) 1 stop better declared by the Nikon engineers, could imply DR at base ISO 1 stop better too.

    But perhaps I’m going too fast here…

  • XT

    I’ll be happy to see DXOMark results and compare them to D810.

    • mas921

      i grew less interested of DXOMark in recent times. i mean see their D500′ iso score…

  • For Nikon, I think this new sensor is “entirely their own design” in much the same way that Fiat designed the 124 Spider, not Mazda.

    The engine might be different, and things like the suspension and steering are custom tweaked, but the basic architecture and underpinnings (the heavy lifting) are all Mazda MX-5 Miata.

    In this sense I think D850 buyers are getting the best of both worlds– Sony’s unmatched sensor infrastructure and Nikon’s well oiled sensor tweaking and optimization.

    • Max

      I think you got it wrong. Nikon designed it and ordered the sensor to be built by Sony.

      • Isn’t Nikon’s design limited by what Sony (or any other fabricator) is able to manufacture? Canon’s best efforts lately still come up short compared to what Sony can manufacture. The production element isn’t as arbitrary as you make it sound.

        • paige4o4

          Nikon simply came up with a design that’s within sonys manufacuting capabilities.

          • The BSI tech alone is pivotal to achieving the additional speed and ISO performance at 46 MP. This is a Sony innovation. Nikon can check off that feature in their design, but it’s not as if Nikon had anything to do with its development. Not sure why this issue is so misunderstood. (It’s not like I’m knocking Nikon.)

            • MB

              BSI is not Sony invention, and Sony wasn’t even the first company to mass produce one …
              Actually BSI CMOS sensors are made very similar to regular ones but at the final stage the back is ground until photo diodes become “visible” …
              What Sony did is to refine this technology and make it cheaper to produce BSI sensors … but until recently they used to make only smaller sensors …
              First mass produced large format (APS-C) BSI sensor was actually made by Samsung as we all know …
              Today TSMC and others have the technology to do it but designing the sensor is a bit more than grinding the back of it 🙂

            • mas921

              true. OmniVision Technologies had it 2 years before sony (according to wikipedia though)

        • CaMeRa QuEsT

          Actually, Canon is doing quite a good job given that Dual Pixel almost doubles the amount of physical pixels on each chip (think about all the additional “walls” they have to add to isolate each pixel from the next), which also almost doubles the amount of data the chip needs to move out (think about all the additional wiring and components they have to add to isolate and process each pixel’s data), also their APS-C sensors are smaller than everybody else’s.

    • a14k

      You could also say all cars are the same because they have wheels, steering wheels…, right?

    • I would say ” ENTIRELY their own design” should clear any doubts about who designed the sensor.

    • Antonio

      Are you sure that “Sony’s unmatched sensor infrastructure” comes and depends entirely on their in-house resources and the only action left to all their sensor manufacturing customers is “tweaking and optimization”?

    • danceprotog

      Then explain why the Nikon D800 36 megapixels are so much better than the Sony a7r 36 megapixels? And the D750 24 megapixels are so much better than the Sony a7ii 24 megapixels?… If these are “Sony’s” designs, then why did Sony let Nikon use the 36 megapixel sensor before they did? Explain to me how Sony is so superior to Nikon that the same sensor used on their cameras seem to be of lesser overall quality than Nikon’s files?

      • David

        Don’t forget Pentax! The ASIC is just as important as the sensor.

    • mas921

      what makes you think its the same as the ND/124 case? there is a news piece somewhere or thats your opinion?

  • tell me the last time this website reported something that makes sense 🙂

  • Robin Prins

    For Nikon, I think this new sensor is “entirely their own design” in much the same way that Fiat designed the 124 Spider, not Mazda.The engine might be different, and things like the suspension and steering are custom tweaked, but the basic architecture and underpinnings (the heavy lifting) are all Mazda MX-5 Miata.In this sense I think D850 buyers are getting the best of both worlds– Sony’s unmatched sensor infrastructure and Nikon’s well oiled sensor tweaking and optimization.

    • PhilK

      Another duplicated comment (from Larry Templeton this time) – albeit without the whitespace.

      Hmm.

      • silmasan

        Right?? Pointless disqus bots in action… Also, the removed whitespaces are a proof of automated text parsing.

        • Right?? Pointless disqus bots in action… Also, the removed whitespaces are a proof of automated text parsing.

  • Chris Phillips

    I am sure that by the end of next week the latest , a hound dog by the name of Peter will flash out the sensor manufacturer easy. Do your thing Peter our pride is on the line here man!!

    • It probably won’t matter to anyone because it’s likely a company that doesn’t make cameras.

  • Markus

    Everytime Nikon designs the sensor by themself they have a winner. Like this cam a lot. I thought I’d go for a D810 but this changes all.

    • I agree.

    • Interesting thought. What other sensors did they design, do you know?

  • Originaru

    “Nikon says dynamic range will be as good or better than that of the D810, despite the higher pixel count.”
    As i understand at least on cmos, more megapixels = higher DR.
    I’m not sure why though, i have many theories, one relates to the proportion of green pixels to the others, higher MP so be more pixels to average the result, faster readout, better software algorithms, quicker processors (that process raw data and translate to “raw” files).
    But i’m clueless about this aspect.

    • ToastyFlake

      I think the general rule is that less megapixels on the same size sensor will mean larger photosites and thus higher dynamic range. So, the more megapixels you cram in, the less your dynamic range. Of course many other things factor in.

    • HD10

      It is possible that the comment referred to sane DR on a per pixel basis rather than the entire imaging sensor itself. If that is true, then the higher resolution D850 DR will be substatically improved from the D810.

      Personally, I am expecting equal or up to 1/3 stop at Base ISO compared to D810 and about 1/2 stop at ISO 800 and above. Anything higher means that Nikon made a major breakthrough in sensor design … which will.be most welcome when I get my D850.

  • BaZa

    I bet there’s Samsung technology on that sensor. BSI sensor and a “Electronic VR option available for Full HD, crops the frame slightly”, is the same thing as Samsung NX1. And Samsung put the “VR” option with just a firmware update. That sensor is/was a beast and from year 2014.

    Too bad the low bitrate for 1080p and the 8-bit external recording. Nikon should see what Fuji is doing with their firmware updates. Add new futures, be aggressive, test it out, let’s go Nikon.

    • well, I can bet it’s not from Samsung just based on the source of that rumor 🙂

    • Mehdi R

      Nikon is not Apple 🙂

      • mas921

        LMAO!

    • Jeffry De Meyer

      Samsung can’t handle the current demand for hbm memory and nand already.
      I doubt they are going to be wasting capacity producing sensors for Nikon.
      Only way there is a Samsung sensor in there is that if someone bought an old Samsung production line

      • I think he meant that there licensed some technology from Samsung and integrated it into the D850 sensor. I’m not agreeing with that, just clarifying what I think he was saying.

  • MB

    Sony was a pioneer making first practical BSI CMOS sensor back in 2009, but today every CMOS fab can do it.
    Gap less micro lens BSI sensor was Micron (Aptina and now On semi) patent but these days it is also common technology…
    I am certain Samsung is not making sensors for Nikon, but we will now for sure only after Chipworks dismantles one 😉

  • They do that all the time – it’s a way to attract attention with made up stories aka clickbait. They twist every news story and try to make it beneficiary to Sony… even if it has nothing to do with Sony. They also reported that the D850 has an EVF lol… and much more BS over the years.

  • Nobody Cares

    If I contact BH, will they tell me my position in the queue?

  • David

    Nice.

  • John Alexander

    I designed my car
    Toyota made it

    • ToastyFlake

      Is that how the dealer suckered you? “You can design your own car with all these ‘low cost’ options!!”

  • Ralf

    Between the lines it seems some folks are a bit obscessed with the thought Nikon might fab their sensor inhouse. More or less nobody in their right mind fabs inhouse anymore except the very largest users or those specialising in contrat fabbing. A modern fab costs a handy few billions and the few sensors Nikon needs just won’t saturate a fab line into profitability.

    • Yeah, no kidding. Most people are clueless about what it takes to make these kinds of devices. Not only does it cost billions just to build and equip the factory itself, but to bring one through process engineering, qualification and production is probably another billion. You better have many hundreds of thousands of devices per month to get that to pencil out.

  • Azmodan

    So Sony improved DR at ISO800+ by a full stop with their new 42MP sensor with BSI compared to the 36MP sensor and Nikon’s version of BSI doesn’t help at all and they claim it’s only for speed. When they say image quality will be the same at twice the ISO as D810 how are they comparing, 100% or equalised resolution or is this jpg only. 1 stop improvement in RAW would be amazing.

  • jtang128

    will pay the price $3300 if Nikon D850 have D810 sensor in the camera

  • Spock

    On paper, the D850 is an amazing DSLR. Way to go Nikon!

  • Alan Wilder

    My understanding is that the sensor is actually made by “Tower Jazz”, an Israeli company affiliated with Panasonic that owns 49% of the company. Nikon does not actually make or has never made sensors but probably had a hand in providing the desired specs to Tower Jazz.

    • if we consider the sources, I say I am 100% sure this info is false/fake

  • Back to top