Additional information on the low light Nikon DSLR camera

nikon-d7xx-dslr-camera-rumors-2 nikon-d7xx-dslr-camera-rumors
The tw0 low light pictures I published two days ago were not taken with a new camera, but with a new sensor if this makes any sense: the new sensor was installed inside a Nikon D750 test mule/body without any specific markings on the outside. The sensor had 20MP resolution (full frame) and was specifically designed for low light photography. There is no indication if Nikon will or will not go ahead with that sensor. This was simply a test and nothing more. Similar tests are done all the time in different locations and this is not really an indication of an upcoming camera. I was ensured that this was not a hoax and the pictures were not fabricated in order to gain online fame. I am hoping to get the full resolution JPG or NEF files in order to get additional information. Once and if I receive them, I will upload them here on the blog. This is really all I can share for now.

Update: here is the full resolution JPG photo I received today:


The EXIF shows just regular D750 data and I could not find anything unusual about it. I still hope to get the original NEF file that will provide us more details about the photo. As far as I am concerned, based on this photo, this could have been taken with a regular 24MP D750 camera (over the last two days I was told multiple times that the sensor was 20MP and it appears to be 24MP based on the EXIF).

This entry was posted in Nikon D760, Weird and tagged . Bookmark the permalink. Trackbacks are closed, but you can post a comment.
  • Steve

    If they go ahead with a 20MP sensor that is optimised for low light, it better have the AF module to back it up, ala the D5. The DF AF is terrible; and it’s such a shame as it’s high ISO performance is brilliant.

    • Dino Brusco

      Definitely agree, what always prevented me from taking the Df was the fact it had the same af of the D600 which is frankly nothing to brag about…. D750/D500/D5 are totally a different story !!!

      • Chris

        I have been using df along with d810 for a year before selling d810. At least when lighting is suitable for taking a picture df works right…. I handheld using 45/1.8 at f/2.8 and 1/15s at ISO 1600 once. You can figure how poor the lighting is.

    • That module was horrible–it’s the same from the D7000/D600.

      Of course, they somehow took the amazing 51pt from the D700/D3/D300 series, and turned it to garbage in the D750/D7100…so I’d wanna test it before buying, even if it’s the 153pt.

      • Eledeuh

        > Of course, they somehow took the amazing 51pt from the D700/D3/D300 series, and turned it to garbage in the D750/D7100…

        …what ?

        • After shooting a D300, D700, D3s, D3x…it’s just not the same. It’s not a reliable. It doesn’t nail focus the way I expect and need it to in the D7100 or D750. It’s the difference between popping off 3 good shots of a group and having three good shots to choose from, and needing to pop off 6 to get one that’s correctly focused. I learned not to trust the D7100 AF. I’ve demoed several D750s and it’s the same story. Even my first D300 was more reliable.

          • Eledeuh

            Two concepts to always have in mind when assessing something :
            – User error
            – Confirmation bias

            • I tested these things side by side on calibration charts. These were real, quantifiable differences in autofocus precision and accuracy.

              But thanks for the ad hominem.

              Lemme guess, I insulted a camera you’re in a close, personal relationship with?

            • Eledeuh

              I mean, look around you, you’d be hard pressed to find a reputable source (from people who know how to test such things, disclose their methodology AND make tests in the field for actual photography, showing the results at each step) that would say “The D750 [for instance] has worse AF performance than [any of the other bodies you mentioned]”.

              Either you didn’t test properly, or you had a lemon in hand. Stating that recent bodies like the D4 are better or worse than a D750 is kind of up for debate ; but those you mentioned, really ? That’s a first for me.

              But anyway, if it fits some narrative you want to buy into, then go ahead, I’m not going to bother.

            • The narrative I’d love to buy into is: my D7100 focuses good enough to get the job done. GREAT! I don’t need to spend any more money.

              …unfortunately, that didn’t happen. I HATE spending money on bodies. It’s always a money suck. So far, out of the box, this D500 is nailing it on the focus front.

              The single digit bodies have been consistently golden for focus, for me. Even my 11pt D2x was solid. But I’ve been hesitant to sink that kinda money into a body again, because, again: they’re a money suck.

              Maybe I’ve been spoiled by the single digit bodies. The D700 has been the only non-single-digit body to satisfy my focusing needs. Well…and now the D500. Ok, the D300/s wasn’t bad either.

              Maybe the D800/D810 is solid too. I haven’t used one enough to know.

              And maybe it’s just the old lenses I’m using. But there are plenty of bodies that focus these lenses just fine. Repeatedly.

      • Eno

        Please, keep in mind the D750 and D7200 do not share the D3, D700, D810 51 point AF sensor. It’s a different one, a little bit more sensitive in low light but with a narrower frame coverage.

        • ITN

          The D810 also doesn’t have the same AF sensor as the D3 or D700. The original version is called “Multi-CAM 3500”; the one in the D810 is “Advanced Multi-CAM 3500”, and the one in the D750 “Advanced Multi-CAM 3500 II”. Furthermore there are smaller differences in the AF between individual models using the same module designation. It is probably easiest to just consider each camera different, at least that way we don’t make assumptions that might not be true.

          • Max

            similarly i dont think the multicam 4800 in the D610 is exactly the same as in D5xxx bodies. or d7000 or DF.

            • Whether it’s exactly the same or not (and it’s easy to argue that it’s not, what with a different mirror to read off/through and all) my experience has been that they perform similarly, and inferior to the D700 era CAM3500.

          • Eno

            Your are correct, I was merely referring to AF point coverage.
            I saw differences in AF speed and performances even between D3, D300 and D700 cameras. 🙂

          • They give them similar names so we are being led to assume they’re closely related. And they add words like “advanced” and “II” to make us think they’re upgrades. Yet…my experience has been that the higher numbered “4800” is a dog. And the Newer versions of CAM3500 have been inferior.

            Except maybe the D810. I’ve only had limited stick time with one, and it seemed fine. As it ought to with that sensor!

            They appear to be back on track with the D500. I’m putting it through it’s paces right now, and it’s at least as precise as the D700. Maybe even a touch better! Finally some progress!

        • KnightPhoto

          Great illustration. By any chance do you have a helpful AF-coverage diagram like that that includes the D5 additional coverage?

          • Eno

            The image I show above was taken from the internet, tray a google search for more comparisons.

        • While there may be subtle differences in build, I’m talking about function.

          You may or may not be right about actual sensor. You could be right, and that could be why I’ve seen a difference. Or you could be wrong, and it could be a difference in AF logic/processing.

          It stumps me that the AF in the D7100 is tied to a fancier, allegedly higher performance light meter. Yet, my D700 tracks action a lot better with the same lenses.

      • Bob Riley Jr

        I have owned the D7100 since it came out. I shoot wildlife..birds in particular and there is absolutely nothing wrong with the AF module.

        • You might be less discerning.

      • Marc P.

        On a FF DSLR, the AF Module from the D7000 onto the D6x0 Series & DF, it’s not good certainly – but onto the D7000 itself, for what it was being developed (DX), it’s way good, period. Can’t complain about my D7000.

        • tobi

          I have the D7000. Its not bad.. I also have the D610 and the D610 is noticeably better. I also have the D7200. and that is much better than both the D610 and D7000. However in 97% of normal shooting situations where you are at ISO100 they all function adequately. PS: I have played with the D5500 and that is a bit better than my D610 and D7000. Just FYI

          • Marc P.

            I know…but the D7200 is no reason to upgrade my D7000. It’s good enough for my needs. Never shoot >1600 ISO, because i don’t need it. A 5×00 Series i’d never buy because they’re lacking too much things, 2nd Dial (for adjusting aperture/time in manual mode), 100% Pentaprism OVF (D3x00, D5x00 is a 95% tiny, dim lit Pentamirror OVF only), Top LCD Display, and so much know. ;).
            For other things i do have my FF Setup, but not from Nikon.

    • Chris

      You don’t buy the Df because of it’s “blazing fast AF”. It’s equipped for all kinds of legacy lenses. The sensor is great as well as the viewfinder.

      • Steve

        I would have thought having the best low light ISO performing camera; just might be a little useless if it struggles to actually focus, when the light drops, no?

        • tobi

          I would think with Manual focus lenses it would not matter 😉

        • You would have thought like someone who actually uses the gear. Clearly, the designers didn’t.

          Maybe they expected lots of manual focus users? But then, it doesn’t come pre-equipped with a prism/split finder, now does it?

      • TheMeckMan

        And most of those legacy lenses works on my d750; what’s your point…

  • Spy Black

    D700 replacement perhaps?

    • Eric Calabros

      The entire planet is hoping for that

      • Aldo

        Not me… that boat sailed long ago for me… nikon corrupted me. They got me hooked on resolution and now my shooting style with prime lenses often using dx mode depend on it. I can not afford to downgrade in res.

        • EnPassant

          With the Sony A99II shooting 12 fps with 42MP, twice as fast as the Nikon D5 which only manages the same fps with half the resolution, Nikon really have to increase the processing speed of their premium cameras.

          For me the “D700” replacement could very well come with an updated 36MP sensor and be capable of at least 8 fps.
          When a many (50-75?) MP sensor is ready Nikon can release that camera as the true D810 replacement.
          Then Nikon would beat both the Canon 5DIV and 5DS(R) on specs.

          • Eric Calabros

            12×42 is around 500MP/s. Nikon is already doing 600MP/s in its mirrorless series. Instant burst speeds can be very high, keeping that data flow for longer time is another thing. A99ii can only hold that for 4.5 seconds, or 54 raws.

            • EnPassant

              But so far Nikon haven’t been able to do that with their DSLRs.
              Nikon D5 only can handle 252MP/s with continuos AF.

              The big differentiatior is the size of the sensor.
              It’s one thing doing it with a 1″ sensor with most photos having very wide Depth Of Field, but quite a different thing doing it with the thin DOF associated with a Full Frame sensor.

              Sony apparently solved that calculation with the processors used in A99II.
              The length of the burst is a matter of buffer size and in the end the speed of the card used. 4.5 seconds should be enough for most uses.

            • Eric Calabros

              Data is data, no matter source is a FF sensor or a much smaller one. Nikon hasn’t anything bigger than 1 inch format that doesn’t have mirror inside. That mechanical thing is limiting them for higher burst, not bandwidth tech.

            • EnPassant

              If you are only looking at the data stream from the sensor, then yes.
              But if you want AF the camera must also process the data from the AF module/AF-points on sensor. And for a smaller sensor it is much easier to achive good (enough) focus as the margin of error is much greater because of more DOF.

              Nikon can do 12 fps with AF on the D5. That is the mirror limitation for Nikon while the mirrorless 20.8MP J5 can do 20 fps with AF, or 416MP/s.

              While I am no expert in Nikon’s use of processors it is quite clear that J5 needing less processing power for AF as well as being able to AF also at times when the mirror inside a DSLR is moving more of the processing speed can be used to shuffle data.

              With your logic we could ask why Nikon did not put a 24 or even 36MP sensor inside D5. Because after all using the processor in the cheap J5 a 36MP sensor with a 416MP/s able processor should be able to reach 11.5 fps, right?

              I think there are several answers to that:

              First, any slight focus error will be clearly more visiable on a 36 versus a 20.8MP sensor. To achive equally accurate focus much more processing power may be needed.

              Second, while low light capability mostly deponds on the size of the sensor there is a small advantage for bigger pixels when light levels get very low.

              Third, the clients may not want or need more MP. As they typically shoot a lot of photos big files would just slow down their workflow too much.

              I don’t know how the raw speed of Sony’s processors compares to Nikon. Even if they would be the same maybe Sony found better ways of programming and using the hardware? But asking for a rise from 12×20,8, 250MP/s to 8×36, 288MP/s does not seem to much in the DSLR world. Especially as Canon can do 14 fps with 20.2MP, 283MP/s.

            • My understanding is that the D500/D5 have a separate processor for AF.

              And the reason the 36mp sensor isn’t being used if because it’s not designed for fast frame rates–the sensor itself just won’t do it. It’s maxed at 6fps. In fact, that was the case from the beginning, and why Sony came up with a new sensor. The 40mp clearly is capable of those speeds.

              And yeah, I’d be happy with 8fps from a camera with that resolution!

            • EnPassant

              Yes, you are correct. Both D5 and D500 have a dedicated AF processor. But as far as I know they are the only Nikon cameras having one. It does of course take some work-load off the main image processor, Expeed 5.

              But as these cameras both use a 20.8MP sensor and have a mirror only experts can tell if 12 fps is the limitation of the AF-processor, the image processor or both, or because the mirror movement limit the measure time. Sony’s SLT solution for A-mount cameras is that AF can be measured constantly.

              Considering Nikon J5 use a variant of the same processor as in D5 and D500 that can shoot at the speed of 20fps at same resolution the main difference seem be the AF-speed.

              I rember reading Nikon themselves explained the sensor size choosen was part of the Nikon 1 concept. Between the lines this means it wouldn’t (at that time) work with a bigger sensor as AF with a thinner DOF would take longer time. It’s the difference between focusing with a very wide angle (for APS-C and bigger sensors) where at medium distances almost everything is in focus unlike for a longer tele lens.

              You are also correct the build structure of the sensor itself seem to be limiting the speed of the old 36MP sensor. Actually googling a bit it seem be even worse than 6fps! I found the information it is actually max at 4.7fps for the full sensor, which of course Nikon for D810 round off to 5fps, with faster speeds, up to 7fps, only available in crop mode.

              That’s the reason I wrote UPDATED 36MP sensor!
              For me 36MP seem be the ideal compromise for FF sensors, just like 20-24MP seem be perfect for APS-C sensors.

              While I don’t have a 42MP Sony camera to compare myself the impression I get is that the 6MP extra hardly is notable and the images from D810 are just as good if not better than from Sony A7RII.
              Because more MP is not always better. Yes, resolution in the centre of the image should be increased as expected. But the off centre and corner results may not be as impressive. Look for exemple at some of photozones latest Canon lens test using both 50 and 21MP sensors where many results are about the same despite the big difference in resolution, and in some results the 21MP sensor measure better!
              The reason is the steep ray angles combined with small pixels in the corner. A digital problem that will remain as long as sensors are not flat like film.

              The reason Sony can get 12fps from A99II (rather than 5fps from A7RII) seem be the addition of a new front-end LSI unit. I would guess that is similar to the AF prcessor in D5 and D500? If something also changed with the sensor compared to the one used in A7RII I do not know.

              Assuming Nikon used a new faster 36MP sensor with the D5 AF processor its not unreasonable to think it could be as fast as the J5 which doesn’t need a sparate AF processor to reach 416MP/s. As I wrote in a post above that would translate to 11.5 fps.
              But such a camera could kill a lot of D5 and D500 sales! Therefore I doubt it will happen and just ask for a “modest” 8fps!

            • I think noise may be impacting resolution with that Canon sensor. It’s really pretty rough, even at base ISO. That 40mp Sony sensor is really close in performance to the D810, to my eye. I expect with dash of Nikon’s sensor magic, they could make it happen.

              At speed? That seems to be the open question.

              My guess is: YES!

              …but when?

              BTW, about film: film emulsions are actually LESS flat. They had a depth and layers to them. Their imaging plane is actually, effectively thicker. That was actually advantageous–they captured a little more information that way. And then…lost some of it to dye bloom in color film…

              Really, if steep angles are an issue, they’re a worse issue for film. And there’s only so steep the angles can be in SLRs, with their mirror restricted flange distance.

              That said, I used to use some wide lenses on 4×5 with really short distances between the back element and the film plane, and I never really noticed much difference between them and longer lenses. Nor did they improve in the corners and closer focus distances in ways I’d attribute to such angles.

              In short: I think I’d have seen that with my experiences. If it was there, it didn’t jump out at me.

            • Captain Insane-O

              The buffer allows files to bottleneck in front of the processor. You’ll need to look at sustained shooting to more properly quantify the processor.

              Also, you’re missing file compression.

    • TX-Night-Sky

      Please put this sensor in d810 body with near D5 specs. Make sure it is best 20MP for low light. Having at least 8fps. Having the best autofocus. This would make a true D700 replacement. D800/D810 was just replacing D3x without the integrated battery grip. I do not need 36+ MP filling up my Hard Drives. I do not need or want a pop-up flash which unusable with big glass. The camera needs be weather proofed. I just want the best possible 20MP in real body like the D810. Nikon you can drop video and optimize controls for still photography instead making a DF2 with its funky aperture wheel. This would make the 100 year anniversary camera that I love to would buy and keep as milestone camera. I could deal with video as long it does not sacrifice still photography performance.

      I will pay same price as a D810 since It be same level as D810 just optimized for low light and sports rather than pixel count. Oh and please give the early adopters a battery grip since they will be beta testing this camera. Nikon please make this camera since I can not afford to buy cameras missing my mark like the D810 and D750.

      Nikon thank you thank you for making the D500. Bummer that I am no longer can afford to buy two bodies. I currently own the old dream team of D300 and D700. Before that I owed D200 and D2h.

  • stormwatch

    Astrophotographers say goodbye to astrophotography.

    • Steve

      Seriously no…

    • nwcs

      Not a chance 🙂

      • stormwatch

        Ok, but explain me how are you going to do the astrophotography with such a souped up sensor. According to this review, if in the middle of the night you put it on for 1/8000 at f/22 ISO 50 with ND8 attached, you will spend the rest of the day removing overexposured parts.

        • David

          You’ll be able to see the remnants of the Big Bang! And still complain about blown highlights.

          • stormwatch

            Yes, you’re absolutely right, remnants of the Big Bang are real money makers, but I would not like to catch a Supernova forming at the same time, then I will have a clean 255, and no PP can make that up, maybe some shadow recovery, but I’m not sure.

          • TwoStrayCats

            Those are already visible on your old television as static noise. Its a great radio telescope.

        • Steve

          Jesus, Mary and Joseph…

          • stormwatch

            Yes yes, We’re happy to have Them, especially Jesus as our savior. Merry Christmas everyone!

        • nwcs

          First of all I really doubt you’ll be able to do 1/8000 and ISO 50 and f22. Second, most of the AP is not done with camera lenses but scopes and many can be stopped down far more than f22 equivalent although that caries its own issues. Third, filters that reduce light pollution drop the amount of light significantly, especially in certain spectrum regions. Lastly, you have to see what programs like PixInsight can do with stacked exposures to tease out signal from what seems to be washed out.

          • stormwatch

            I don’t doubt it at all, actually this sensor and some app for identyfying space trash that orbits around the planet will be a breaktrough in the space imagery.

        • Michiel953


        • Bob Thane

          As far as I can tell, this review didn’t use those settings. I believe it was a 2.5 second exposure on one of the shots, and high ISO for both. You can’t mess with physics and make stars emit more light, the only way 1/8000 f/22 ISO 50 with ND8 would possibly overexpose a night sky would be if the definitions of ISO or aperture were changed so that it’s really f0.4 and/or ISO 256000.

    • Spy Black

      Tough to do something with equipment that can do something when you don’t know how to do something with the equipment…

  • Admin, thanks for following up! No matter the misunderstanding, the last post had so many comments, just like in the good old days 🙂

    • Eric Calabros

      with T.I.M as nuclear comment generator, I never miss the old days.

      • T.I.M

        Don’t even think about Astro photography in Florida (or maybe the moon in a cold windy “winter” night).

      • Ha ha, this gave me a good laugh, thanks 🙂

        • T.I.M

          I’m still laughing !

  • Andrew

    I wonder how good this sensor is in normal light, though. I.e. what sacrifices they had to make for such low-light performance? On D5 they had to give up some DR in order to get slightly better low-light capabilities. I hope it is not the case here.

    • Eric Calabros

      I still don’t believe D5 base ISO issue (if you call it an issue) is related to its performance at high ISO.

      • Thom Hogan

        Believe it. It has to do with how you tune the gain mechanism in the sensor.

        Nikon themselves know what I’m about to write: “noise” will be predominated by one of three things: shot, read, or pattern noise. If you tune to make read and pattern noise low, shot noise dominates. How you tune determines which “noise” predominates. Nikon had a little Java tool on their site almost two decades ago showing the relationship of how when you tune for one thing you get another thing dominating.

        On the D5, the gain system is optimized for getting accurate values in low light. Because of that, read noise predominates at low ISO values and keeps the sensor from having exceptional low ISO dynamic range.

        The curious thing is that the D3/D3s/D700 used this approach, but the D4/D4s didn’t. Now the D5 does so again. I don’t know why Nikon changed their approach. But I suspect it was because the D4/D4s was already close to the best you can do with the ISO-less approach. In order to make the D5 look a little better at high ISO values, they had to return to the trick they used with the D3/D3s.

        • Max

          I think CMOS sensors have reached a tech ceiling. There’s nothing you can do to make it better, unless someone comes up with some radically different technology to increase light sensitivity..
          But then low light is low light – you can’t add more photons to describe a scene, unless going for bigger sensor and bigger lenses.
          I wonder why DXO won’t test the 645

          • Hector Cordero Espaillat

            the d500 used some tricks to push the sensor near the FF performance, probably this same technology we`ll see on next FF cameras. The night vision devices uses other technology that (i suspect) can be put inside a camera… is nikon developing a color night vision camera???

          • silmasan

            They tested Phase One P65+ and IQ180. Both are CCD of course…

          • harvey

            wasn’t there a DXO rating of the Pentax 645z of 101 with a ‘sport’ rating of about 4400, if I remember correctly?

          • Thom Hogan

            No, CMOS hasn’t reached a ceiling yet. Moreover, you also get into the age old “optimize for A or optimize for B” type of conundrum. You find that you are in a place where you can’t do both.

            A lot of the gains we’ve seen in the recent past are more attributable to things that start on the sensor but aren’t solely constrained to it, particularly the data processing chain.

            We have a couple of areas that are still not near the physical boundaries, such as efficiency. Plenty of work is active on that, including the Quantum Dot approach.

            Still, what a lot of folk fail to understand is that the random nature of photons is where most of the noise you see originates these days. You could make a “perfect” sensor, but randomness will still produce noise.

            • Spy Black

              Although somewhat of a long shot, there’s also the possibility of finally mass-producing technology like this:

              …or this:

              Yes, I wouldn’t hold my breath on it, but the the technology exists.

        • Eric Calabros

          Why they didn’t use dual gain tech (or whatever they call it) that Sony used in A7rii? is it related to pixel saturation level, so better suited for high res sensors?

          • Eno

            My thoughts exactly!
            Probably because a dual gain version must appear in D5s…

          • Thom Hogan

            The D3/D3s/D5 sensors are all multi-gain sensors with very tuned small gain adjustments suited primarily to high ISO.

            The Sony advances on the A7rII roll through some other things, like BSI, while the D5 sensor is a traditional front side one. As such, the D5 sensor is probably cheaper to produce than the Sony one, and that’s especially true of low volumes.

            Sensor tuning is a very tricky subject. There are experts that believe that no one is getting it perfectly right ;~).

    • stormwatch

      It’s not the daylight sensor, there is still no lens + ND combination that can handle it.

  • Sawyerspadre

    Interesting. I would think anybody that Nikon would send a test mule camera to, would have signed an NDA. I am sure that Nikon employees and reps have a confidentiality agreement for intellectual property and business secrets. Outsiders would have to sign an NDA to get their hands on a mule.

    • Thom Hogan

      Absolutely. Moreover, it’s extremely weird that you’d send the mule to that country to that photographer to do that kind of shot to test a low-light mule.

      More interesting is the “20mp” comment coupled with fact that these are supposedly 2015 photos. Makes it sound like a tuning mule for the D5 to me. See my comment above to Eric: you can indeed drive sensors in slightly different ways to get different results.

      • This could probably be confirmed by culling exif data from the D5 promo shots. If there were folks in the field, shooting with actual D5s during the period in 2015 these were shot, then this is probably not that sensor.

        I’ll be shocked if Nikon doesn’t have some sort of body with the 40mp Sony sensor in the near future, whether it’s a D7__, D8__, or D_x. 20mp so soon after the D5 doesn’t make a lot of sense to me.

  • It would make for a sweet, street-like “anniversary” camera though. Full-frame, retro looking body, optional grip … mirrorless? Who cares… it wouldn’t have to be part of any new “line”, it could just stand out on it’s own as a magical unicorn. Come on Nikon, give me a crazy cool unicorn to attach my odd little 58mm to.

    • If they do another retro body, I hope they give a retro style viewfinder for all my AI-S lenses.

  • Allan

    Is there any precedence for photos from a test sensor or camera being “leaked”, and the test sensor or camera not eventually being released in one form or another?

  • Eric Calabros

    so we no longer spot any secret body. New sensors are inside bodies we see every day. Jesus is among us!

    • Leica does that all the time and then they sell the demo body at auctions:

    • Thom Hogan

      Nikon has done this for a long time. I can’t say which model it was because the person who handed me the mule could get into trouble, but mules are very common. At the pro level (D3, D4, D5), the body doesn’t really change much between generations, anyway, so you can easily build mules to test with. I’m pretty sure the D5 sensor was tested with D4 mules.

  • KS

    I’m still calling bullshit. Stop giving attention to this guy, obviously he gets it that his made-up story is falling apart and is now trying to explain his ass.

    • Well, the next step is to get the original files. I have no control over that, but I hope he does send them to me.

      • Eric Calabros

        Why he should do that? If it was a mistake, so it was unintentional. Now if he gives files to anybody, he does an intentional version of same act.

        • Aldo


        • Because he was upset that some readers did not believe him and called him a fraud. I hope he sends me the files.

          • Aldo

            The only way that story works is to assume he is incredibly naive (web wise). I too hope he ‘leaks’ the raw files though.

          • Thom Hogan

            He should be upset with himself first. (1) he didn’t do enough to establish any meaningful believability to his claim; (2) he’s almost certainly stepping on his own NDA or abusing someone else’s. And to what end?

            • Fujimoto

              For a guy that said this is fake, you sure seem interested.

            • virage

              +1. Thom is full of himself.

            • Steve Hughes

              Thom also has a boatload more industry experience, and especially Nikon experience, than most of us.

            • Hans

              Oh! Really? With boat loads of shit that has proven wrong over time.

            • Allan

              Thom is generous with his time, and contributes both well reasoned industry insights, as well as multiple teaching points about digital photography.

              The negative comments are inappropriate.

            • Tom Co

              With the exception of my ex wife I don’t know of anyone on this planet who is right 100% of the time.

            • Steve

              Ha Ha!

            • Tom Bruno

              Sure! My ex wife! 100 percent, always right.

            • whisky

              i don’t always agree with Thom’s analysis, but his credibility is tops. it’s hard to find anyone with more credibility, and integrity, than Thom.

              if that breeds a little ego — so be it. JMO.

            • decentrist

              you have a low bar

            • Bill Ferris

              Thom was far from alone in being skeptical if the authenticity of the story surrounding the images. Let’s be frank: it looked, smelled and tasted like a hoax. And in a sense, the original hypothetical – these are exposures from a new, soon-to-be-released Nikon camera – has been shown to be “fake.” If this new story is accurate (let’s wait on analysis of the original files), then it will be confirmed. The images were made as part of field testing of sensor tech; not as beta testing of a soon-to-be-released super low light camera body. If that’s confirmed, it’s an interesting peak behind the scenes at how a major camera manufacturer tests concepts. How is that not interesting?

            • EnPassant

              That about sums it up.
              But this sensor or a similar based on the tech may still end up in a real camera.
              My guess as I wrote in another posting is that this is testing for a new improved sensor to be used in D5s that is to be expected in early 2018.

              Remember that Nikon came with a redesigned sensor for the D3s with much improved low light performance compared to D3 and D700. I would bet Nikon will do the same for D5s. Maybe ISO above 102,400 now actually will be usable?

            • azur

              “For a guy that said this is fake, you sure seem interested.”

              I think Thom is just explaining why he doesn’t think the Facebook leaker has any reason to be upset with our doubts.
              On which I fully concur. The story doesn’t have much credibility, in my opinion.
              On the one hand it isn’t all that likely that the Nikon Corporation would arrange such a leak teaser-stunt of an isolated, early sensor test.
              On the other hand it is also not all that likely that a sensor test driver would by his own initiative post test results on Facebook (of all medias).

            • Gary

              But isn’t that what you’d expect from someone who has an enquiring mind? Thom stated his view, and remained engaged as more evidence became available, but without putting people down who expressed contrary opinions.

              I’d like to think that we’d all do the same – and if the evidence instead proved our initial assessment to be wrong, then we’d acknowledge that too and learn from it.

          • Spy Black

            Inasmuch as I would like to think this is really happening, I still think this is all suspect.

          • Spy Black

            Well, your update pretty much confirms this whole thing was a hoax. You should really have made that a separate posting instead of an update to this one, as new comments will be buried in the past comments.

            • I updated my post already. I am still waiting to see if I will get the NEF file.

            • Even if he could supply an NEF file, you likely wouldn’t be able to open it (if it were from a new sensor) – you’d need a RAW converter updated to handle the file, wouldn’t you? If you could open it, it would just go further to prove that the file not from a new camera.

            • Hans

              Pete should have removed the initial post when the owner came forward to say that the images belongs to him and was leaked by someone without his consent. It really doesn’t matter if the assumed information of a new camera is real or not, the pictures still belongs to him and he got the right to file claim for distribution without his consent.

            • I don’t know how many times I have to repeat this: I was asked to remove the screenshots because they contained the name of the photographer. He actually sent me the actual pictures to post online instead. Are we all clear on that now?

            • Hans

              Sorry, I wasn’t aware of him sending you those pictures to replace those with his name on until your reply. Maybe it will help if you post a reply directly under his initial post and state that these pictures are provided by him to replace those he objected. This will surely reduce the kind of misunderstanding like mine.

            • This whole thing became a big mess. Btw, the Facebook post was public too and everyone could see it.

            • Spy Black

              The original comments were of the owner, taken off his fb page. So he was full of it.

      • Davo

        What software can read the NEFs if this is from an unreleased sensor?

        • tobi

          Read the zeros and ones..for sure.. and lots of information can be gathered.. you wont be able to “process” it in to a good image, but the geeks will be able to “see” stuff.

  • Aldo

    I think we are nearing the end of the SLR… and just like with film cameras we will see some really amazing dslr’s right before they become expensive door stoppers. How long will this take? Idk but I know we will see the f6 of digital pretty soon.

    • T.I.M

      That’s what people said about DSLR when the first Polaroid camera was released.

      • Aldo

        you mean SLR? polaroids came out before DSLRs

        • T.I.M


      • harvey

        with Polaroid coming out in the late 40s, this predated the popularity of slrs by a few years …

    • purenupe1

      door stop is an over exaggeration, DSLR’s can be used for as long as they power up, unlike film cameras and the loss of availabilty for common and necessary film & processing chemicals

      • Max

        Good point. The change to mirrorless is not nearly as big a difference as film to digital.
        A lot of people still use old rebels 350D’s and D40’s and D70’s etc..
        If you look at Flickr’s camera finder, one of the most used Nikons on there is the D90.

        By the time mirrorless cameras are responsive enough, have good enough EVF’s, quicker AF and the battery life’s improved, it will still take another long time before the then-new DSLRs will go unused.

        • azur

          “The change to mirrorless is not nearly as big a difference as film to digital.”

          Unless mirrorless means a new mount and therefore a completely new lineup of native lenses.
          That would be a sad day for my collection of F-mount lenses.
          I’m not a big fan of adapter solutions adding additional tolerances to the system.

    • Spy Black

      As far as I’m concerned, that would depend on the advancement of the EVF. The primary issue is you’d have to overcome image lag. You’d also need to have sufficient brightness in daylight. You would additionally need an increase in resolution and frame rate. All this, and you’d also need to significantly increase battery life on top of the requirements I just made.

      • John Albino

        I also think battery technology is a big obstacle. One of the biggest limitations of renewable power such as solar farms and wind farms is what to do when the sun doesn’t shine and the wind doesn’t blow. There are no huge battery breakthroughs on the horizon to solve that problem, much less provide enough capacity in a small form factor to power a fully electronic (EVF, no mechanical mirror box, electronic shutter, etc.) cameras.

  • Yang Shunyong

    Actually, I really hope the previous message is from a new NIKON camera. Does anyone have the same feeling with me?

  • Bill Ferris

    Does Nikon use mules to test tech other than sensors? AF systems? Touch screen menus? Lens designs?

  • EnPassant

    If the sensor used was 20MP (Or rather closer to 21?) and the photos were taken recently and not in 2015 it could be testing for the D5s expected to be released early 2018. Or for some special low light (and 4K video?) camera.

  • sickheadache

    I went to bed with another sick headache.

    • FountainHead

      You need to lie down.

  • Štefan Herics

    I still believe, that the first leak is true and we will see a replacement for D750… right now, the guy just got shit himself, because the lawyers from Nikon contacted him.. 😀

  • RC Jenkins

    What could have been useful from the source between now and the time we’ll apparently get the JPEG’s & NEF’s would be the settings that were used…

  • Captain Megaton

    It’s more plausible that he was testing a soon-to-be released D750 replacement camera and now trying to do damage control by claiming it’s just an experimental sensor.

    Sorry, but Nikon does not need to send a sensor to Scandinavia to test the low light performance. They can just turn off the lights at the Tokyo office. You might send a new camera out into the snowy wilderness to get last minute feedback on low temp, low light conditions, firmware options, but just a sensor in an D750 test mule? Nope.

    • Maybe this was a camera for astrophotography and they wanted to test in the Northern Lights. I am just guessing here.

      • CERO

        D5a for astro then?

      • Bill Ferris

        That’s an interesting idea. However, astrophotography is more about putting as much light as possible on the sensor. Astrophotographers would likely be more interested in a body with built-in OIII and H-alpha filters, maybe a broadband light pollution filter or even an in-camera app that overlays an outline of celestial objects present within the FOV. There are several features an astrophotographer would prefer to super high ISO performance.

        • nwcs

          Right on. Plus we would want Nikon to be more favorable to bring controlled by programs like BackyardNikon and no more artificial limits on shutter speed in the camera’s intervalometer. As an example anyway…

      • Salty-snack

        Why on earth would anyone want to use this for astrophotography?

  • Beso

    I’m just glad Peter was “ensured” this was not a hoax. Hopefully he was assured as well. Whatever sells page hits, eh?

    • Mato34

      I believe in admin’s sincerity more than in that of others rumor’s sites. At least he is telling us what he can share with no hype, and in a quite “cold” and objective manner.

    • harvey

      feeling better about yourself now after all those corrections?

  • GapUp

    The pic with girl has to be fake ….. no way sensor can pull colors in total darkness.

    • EnPassant

      Not seeing colors in the night is a limitation of the human eye.
      Some (propably many, but yet to be confirmed in tests) night active animals have been tested and found able to see different colors in night who for humans would appear the same grey.

      Digital sensors can see much more light than the human eye. That’s why they need filters for UV and IR so that the images not will have a color cast humand not can see with their eyes.

      • RC Jenkins

        In order to see anything, there has to be a source of radiation with those frequencies in the first place for the object(s) to either reflect or propagate. In many cases, IR radiation is propagated (eg. from a heat source), not reflected–and so you wouldn’t see things like flesh tones. To see flesh tones, you’d have to have some external light with flesh-tone frequencies. To see flesh tones + colors of hair + hat + jacket + trees + snow, you’d need at least some visible white light.

        The animal ‘color’ thing is typically (but not always) more about depth of colors than ability to see them–kind of like a camera sensor that can detect 20 bits vs 24 bits. Colors don’t always appear as grey–sometimes they’ll be a close neighbor. Some animals can additionally see things like polarization.

        In any event, neither of these would have the effect of producing the colors in the image above.

        • RC Jenkins

          I thought I’d back up my statements with proof… 🙂

          Here’s an image I took last year, in close to pitch black darkness. I couldn’t see the trees other than silhouettes–it was very dark.

          I googled around, and here’s someone else’s image of that same scene, which is closer to how it appeared in real life:

          What settings did I use? ISO 800, 30 seconds, F/11 on a D750.

          I’ve got 4 stops before I get to F/2.8. And 3 stops before ISO 6400. So just by opening up the lens, I could have shot at 2 seconds. By upping the ISO, I could have shot at 1/5 of a second. He said he shot his ‘girl’ picture at 2.5 seconds. And there weren’t little dim lights all over the ground, but there was a lot of snow.

          Any my image as I shot it is very clean. Also attaching a 1:1 so you can see this.

          So until I see some settings, I am not particularly impressed; nor do I think this is impossible. Quite the opposite, in fact.

        • EnPassant

          I think any conclusions about the light conditions for the photo with the girl rest on a very loose ground. Without being there it is hard to make a reference point based on a description claiming it was pitch black.

          And where exactly was the photo taken. How far from the nearest town or village with lights on were they? Light from a town can be reflected in clouds very long distances. So there can be light in the forest (helped by the very reflecting snow!) even if it is such weak humans don’t notice and think it is completely dark. That can explain the (for the camera sensor) visible colors plus propably clever signalprocessing adapted for low light photography.

          • RC Jenkins

            At this point, the EXIF data has already given the settings, and there’s nothing particularly special at all. All evidence points to the fact that this was just a standard Nikon D750 shot at F/2.8, ISO 12,800, 3 second exposure. It wasn’t anywhere close to ‘pitch black’.

            Here’s an exercise: take a shot outside at night with those settings. Or light a larger room with a only candle…it might look dark and almost pitch black, but that’s plenty of light for many cameras with 3 seconds of exposure time and F/2.8. 🙂

            • EnPassant

              In a villa area, as the one I live in, with only street lights on, such an exposure will turn night into day and make even walls of houses that for the human eye look black or almost black appear in their natural colors.

              Actually I just tested. It’s late in night here! With a D750 of course! So I see your point.
              However lights are on and scattered around, so far from conditions in a forest with no artifical lights on in miles distance.

              To see what kind of images I would get in such light I would have to drive many miles from here. After next snowfall! Because the 2dm we got a couple of weeks ago have mostly melted away, and now it’s raining… 🙁

  • Max

    Could this maybe have been the D5 sensor?

  • Nikos Skartsilas

    Mirrorless DF II with this sensor…..said, and woke up.

  • Fishguy

    The first post mentioned “total darkness” – yet the second photo of the trees clearly shows distinct shadows. It is either a “day for night” shot, or using moonlight with the camera on the tripod – move along, nothing to see here….

    • EnPassant

      From translatad description of the second image (first in the posting two days ago): Moonlight, but still very dark
      Lear to read before you fart!

      • Fishguy

        Shadows are way to sharp.

  • Update: here is the full resolution JPG photo I receive today:

    The EXIF shows just regular D750 data and I could not find anything unusual about it. I still hope to get the original NEF file that will provide us more details about the photo. As far as I am concerned, based on this photo, this could have been taken with a regular 24mp D750 camera.

    • So unless I receive the NEF file that shows the same level of low light capabilities shown in the JPG (meaning the photos was not edited), I would say all this was fake.

    • Clubber Lang

      Looks like a Camera company CEO pondering what the heck they should do to sell more cameras.

    • RC Jenkins

      Just to highlight something for the readers: This JPEG is at 6016 x 4016 pixels. This is 24MP, and the exact native resolution of a standard Nikon D750. This is a higher resolution than what the supposed 20MP sensor would provide.

      Also, The image of the girl was taken at ISO 12,800, F/2.8, 3 seconds. Moonlight with snow on the ground, and plenty of light to take those images.

      The EXIF data shows that it was modified in Adobe Photoshop CC 2017 on a Mac.

      This to me means this was a standard D750. See my post below for similar (if not worse) lighting conditions.

      • Yes, but this is not what I was told in the past two days while I was talking to this guy. This latest picture shows a completely different story compared to what I was told.

        • RC Jenkins

          Sure…please don’t take my comment the wrong way. I know you didn’t have the right info from the source, and you were very clear about that from the start! You did a good job separating fact from fiction!

          My comment was just to highlight these facts for other readers who may gloss over these points when skimming. 🙂

        • AC

          Is the guy still in contact with you? You could ask him why he’s so sure that this is a new camera test body, given the fact that it shows only data from a common D750. He claims that he owns (or has shot with) D5, D500, D810, D750 and so on. What is it that should be revolutionary in this “mule” that he claims he put his hands on. Resolution is D750 24Mpixel, exif is regular, High ISO seems normal, shot seems taken a year ago. What is his evidence that we should believe his story???

    • Aldo

      only thing that photo is missing is a caption that reads:
      ‘ Trollololol’

      Reminds of of older retired military personal doing interviews about UFO stories…

      • EnPassant

        I saw a troll yesterday evening in the centre of my town. It wore a big, red santa’s hat, had a big nose, the size and shape of a thick cut branch without bark, and wore green spruce clothing.
        Next time I go to the forest I will see trolls in every tree!
        Merry Christmas to you!

  • DSP~

    Can’t we just get over this stupid attention seeking guy??
    I love rumors but this one is total BS!!
    All evidence shows a rather unknown photographer with a D750 and some shots in low light.
    There is nothing at all, no information to support his claims. Please stop giving this guy attention.

  • Brubabs

    And the date in the EXIF data is indeed December 2015 – a year ago. Of course, I suppose the date could have been set incorrectly in the camera.

  • Bob Thane

    Oh man, the original’s hilarious. Nothing special there, just a typical ISO 12800 file with lots of noise reduction.

    • RC Jenkins

      Lol! That’s one way to think about it…but look on the “bright” side (pun intended):

      1) the Nikon D750 continues to have great low-light performance, as do many cameras.

      2) there are a lot of people who don’t shoot in low light often enough to be familiar with or to take advantage of the low light capabilities of their cameras.

      Think ISO 12,800 is too noisy on your camera, but 3,200 isn’t? Shoot 10 second exposure instead of 2.5. Problem solved.

      As a reminder, the D750 is even specced to Autofocus high contrast subjects in -3EV

      …which is approx. the light from the full moon…

      ..(without snow on the ground)…

      (which is probably at least as much light as there was in the scene above.


      • Bob Thane

        Totally, the D750’s a phenomenal camera. But this shot is either from a D750, or from a prototype that’s not noticeably better.

  • silmasan

    What I meant by “both are CCD” in this case is that in terms of low light performance none of them would be a match for the CMOS sensors, despite being a larger sensor.

  • Politics_Nerd

    I hope this is a new Sony-derived but Nikon-developed with Nikon firmware back side illuminated sensor. That would make perfect sense. I have a Sony camcorder with a BSI sensor that albeit tiny, does alright in low light. It looks like D3200 quality more or less in low light. Which ain’t all that bad. This would be an AWESOME upgrade for the D500 in the form of the D500s. Or a D750s. That would be an awesome companion to my existing D750. But a D500 speed with the low light sensor would be an awesome indoor sports camera…

  • Spy Black

    Yeah, I’m still not buying his rap. He was full of it.

    • Hans

      Just back spending the Christmas on an island and catamaran, so that’s the reason for the late reply.
      I wouldn’t like to join in the discussion on either the article writer is telling the truth or not because I have absolutely no way of knowing it. But I do know that developers and R&D engineers do at times use an existing product to test their latest product with modifications. And since cameras are very similar there is always a high chance of doing just that. This is based on close to 30 years of product development and management. The next thing is this particular photographer is good and specialized in low light and long exposure, I wouldn’t be surprised if Nikon pick him as a product evaluator to gather feedback and to promote their product. Again this is based on my experience of having participated in a couple of events organized by Nikon to promote their latest products. For example they use a National Geographic photographer who specializes in landscape, nature and wildlife to promote their D800. A couple of wedding photographers to promote their D750. So really I am not surprised if they really did that.

  • Mike Ver Sprill

    I call BS… Look at the exif ISO 12,800 f 2.8 Shutter 3 seconds and this guy claims it was pitch black with no moon light. Now go look up night photos taken at 12,800 f 2.8 shutter 25- 30 seconds long with no moon and they are not this bright. I think this photographer is trolling us to get views with his photography or something because a 3 second exposure in pitch black with an ISO of only 12,800 would not render those results. He probably took that picture during a cloudy day. The nikon d750 at ISO 12,800 is fantastic… especially if you are properly exposed, hence the clean image. I think we are being trolled.. just my 2 cents. I hope I am proven wrong because I’m all for a high ISO monster!!! we shall see…

  • Aldo

    At least now there is an alternative to becoming quickly famous by making a sex tape…

  • Aldo

    Just came across one of the Angry Photographer videos saying that Nikon will kill their d750 series with a baby D5… with the same sensor as the D5. Not sure if he actually has a source or he is merely speculating.

    • Spy Black

      Well then, the price of D750s will become very attractive.

    • RC Jenkins

      Probably just reading this site or other sites, which is why I don’t ever pay attention to angry photographer.

      For example, see here:

      Poor, irresponsible journalism by those sites. According to that site, NikonRumors (this site) has confirmed that this image was taken with a D750 with a 20MP sensor.

    • if he is not repeating NikonRumors, he is making stuff up

  • Adnan

    Maybe the “new” sensor in D750 was of the D5 , some geeks do alterations like that just for fun and “what if”.

  • this has been going on for while and it’s not only this site

  • Rick Ram

    I could have taken this picture with my d810 so the pictures are not showing anything special..

  • peter w

    Well, we had nothing else to do, it was great fun. Happy christmas to all of you.

    (When do you present a D820?)
    (what could be improved on D810 except for autofocus? Live view functionality?)

  • Ok, so no new camera from Nikon then. I guess the only surprise left in the near future is the mysterious Sony A9.

  • DafOwen

    IF this is a brand new sensor – I don’t expect ACR etc would be able to process a new NEF file.
    And if the pixels dimensions in the jpeg are correct (24 rather than 20Mpx) I’d doubt the story of it being a new chip.

  • Piooof

    Honestly, I can’t understand why this photo is still on NR. This was supposedly taken in a pitch black forest at moonlight. But a defining characteristic of moonlight is -OMG- shadows. Yep, like the sun (it’s the same apparent size!). I can’t see any. Conclusion: light source is much, much larger, like clouds backlighted by the sun. What we commonly call “cloudy daylight”, but maybe that bears a different name in Swedish which might be decomposed into pitch, tar, black and silvery moon.

    • I think we all agreed that the photo is fake.

  • Back to top