August News – Nikon D850 Thoughts

Posted on by

 Nikon D850 – Initial Thoughts.

Nikon D850

Before we get into my initial thoughts about the D850 “leaked specs”, Roger Styles read my D5 post from the other week and asks:

“Very interesting…and I wonder if you would care to suggest how the D500 with a 300mm f4 lens would have performed? Similar? Worse? or heaven forbid better?”

Well Roger, all I can say is that I’m not really in a position to comment on how the MultiCam 20K system worked out on the D500/300mmPF combo (I didn’t use it) but I do have  D4S, D5 and D500 7200ISO shots so you can compare the image quality with regard to noise etc:

Nikon D500, 300mm f4 PF, 1/2500th @ f8, ISO 7200. Click to view full size

Nikon D5, 500mm 1/2000th @ f8, ISO 7200. Click to view full size

Nikon D4S, 400mm f2.8 1/2500th @ f8, ISO 7200. Click to view full size

Bare in mind I’m only illustrating IQ here – so look at the out-of-focus areas and darker tones to see the differences.

Roger – I can’t offer you any real comparisons between the D5 and D500 AF performance,  but from other tests I’ve done with the D500/300PF combo I’d say it performs the same or slightly better than the D5.  But only because you are using a shorter focal length lens with theoretically greater depth of field for any given aperture and distance – therefore more AF errors are masked by DoF.

Why does the D500 image look so crappy?

The answer is simple – too many mega pixels and not enough light!

The more megapixels you squeeze into a fixed area, the smaller each one of those photosites has to be.

There are two main problems with making photosites smaller:

  1. Reduced Dynamic Range
  2. Increased Diffraction

Overall, the sensor becomes more light-hungry.

Let’s put these three sensors on an even playing field with regard to crop factor:

D4/D4S = 16MpFX = 8Mp x1.5 crop

D5 = 20.8MpFX = 10.4Mp x1.5 crop

D500 = 20.9Mp x1.5 crop = 41.8MpFX

The IQ implications of these figures are illustrated in the images above!

And this brings me nicely around to the new Nikon D850.

Nikon D850

I got rather excited about the idea of this camera when it was first thought to have a hybrid OVF/EVF – the implications for using the plethora of super-sharp older manual lenses with modern focus-peaking in an EVF made me go all swoony!

But alas, this was not to be, and instead, all we have is a pumped up FX D500 – if the leaked specifications are to be believed.

The D850 is NOT a replacement for the D810 – anyone who thinks that is an idiot.

Let’s look at these leaked specifications:

  • 45.75MP FX full frame CMOS sensor – clipped Dynamic Range then, nice one Nikon
  • 180,000 RGB sensor that’s same as the D5, with better face detection and enhanced scene recognition – really?
  • Native ISO range of 64-25600 (expandable to 32-108400) – meaningless at the top end, and I doubt the base ISO will actually be 64ISO
  • 153-point AF system with 30% more frame coverage than the D5 – a higher resolution sack of angry weasels!
  • Center AF point -4EV, and all others -3EV – same as the D5
  • 8K timelapse shooting – Who in their right mind shoots time lapse and allows the camera to process and assemble it? Oh yeah, that’s right – dickheads!
  • 4K UHD video recording in FX with no crop – pass
  • 51-photo buffer when shooting in 14-bit uncompressed RAW – GOOD. That’s really a data-pushing miracle, to be honest
  • 3.2″, 2.36-million-dot tilting LCD touchscreen with improved gesture control – tilty screens are useful but straight away are a weak point.  But what use is gesture control when you’ve got gloves on ‘cos it’s -30 below?
  • 7fps continuous shooting standard, 6fps with autofocus, 9fps when using a battery grip – here’s where the price tag will go over £4000, because the grip will be £400 plus if I know Nikon!
  • 30fps at 8MP using the electronic shutter – 8Mp raws from a 48Mp sensor – what a spiffingly top notch idea.  And is that 30fps available silently?
  • RAW can be small, medium, and large resolutions – For F***s SAKE WHY would you buy a huge capacity camera and then shoot small files with it?  Has the world gone bloody mad?
  • 0.75x magnification viewfinder, the first for a full-frame DSLR – GREAT, but you can buy an adaptor to do the same thing to the majority of existing Nikons.
  • Focus stacking. The camera can shoot up to 300 photos with 10 levels of bracketed focus from nearest to infinity for software to stack afterwards – I have every confidence that this will turn out to be crap!  It’s a gimmick to get the unskilled to part with their money.  Aimed at macro and landscape photographers who can’t be bothered to tweak their focus manually.
  • Natural Light AWB achieves better white balancing in natural light – stupid gimmick
  • Completely silent electronic shutter while shooting in live view. – could be useful for sports if it works with fast shutter speeds
  • There’s no low-pass filter – GOOD – why couldn’t they do that on the D5?
  • SD + XQD card slots – Jesus Christ – Nikon need to grow up and stop mixing media

So as I’m sure you can tell, the Nikon D850 is not setting my world on fire.

What could Nikon have given us?

A hybrid OVF/EVF with an RGBW sensor and keep the capacity down to 36Mp or a tad less would have made a good impression with me for starters.

You have to have been asleep for months to not have heard something about the Fuji GFX medium format.  That’s running at 51.4Mp on a 1441mm2 sensor, which is roughly 1.67x the area of an FX 35mm camera.

Simple maths tells us that if we trimmed the GFX sensor to fit in a 35mm DSLR then it would be – that’s right, 30Mp.  The world of photography is populated by frigging idiots who just keep clamouring for more megs – and the camera manufacturers give them what they ask for simply because the idiots spend money like it’s going out of fashion.

Listen, if you want 50 megapixels or more, then go and buy a medium format camera and get 50 megs worth of good dynamic range with nominal diffraction.

Do not buy a Nikon D850 then stick a wide angle lens on and stop down to f22 – the image will be unusable at full resolution – and I don’t need to see a raw file to know that; it’s simple physics.

How this camera will stack up on the sports/action/wildlife front remains to be seen, but I don’t see how it can even be as good as a D5 – and that’s not brilliant.

To get the full potential out of the D850 for sports/action/wildlife then you will need the vertical grip AND an ENEL18A battery or two, and a charger, because I don’t think the D850 has USB charging.

An ENEL18A battery at Park Cameras is £169.00 and a genuine MH-26a charger is crazy money anywhere!

So you will be looking at more than £4000 – and I can think of far more sensible ways to spend that lump of cash.

Nikon promised us something really special to celebrate their 100th aniversary – this ain’t special Nikon!  It’s nothing more than the DSLR equivalent of a click-bait video.

But then again, I’m going on “leaked specifications” – and they could all be lies, smoke and mirrors.  We will have to wait and see what the real specs are when Nikon officially announce the D850.

Become a patron from as little as $1 per month, and help me produce more free content.

Patrons gain access to a variety of FREE rewards, discounts and bonuses.

 

8 thoughts on “August News – Nikon D850 Thoughts”

      • Hi, thanks for adding that. It’s interesting to see how the D500 performs. At the risk of straying of the D850 thoughts topic I was interested in more of your thoughts on the D500 with a smaller/ faster lens as a possibly more affordable set up.

        I appreciate the photos were only examples of equivalent ISO performance but if you were by the side of the guy with the D500* (f8/ISO 7,200) with your D5 when that picture was taken you would have to set your aperture to f12 to get similar DOF in your FF picture. Therefore your ISO would have to be increased to over 16,000.

        Similarly if the D500 guy was by the D5 when it shot at f8/ISO 7,200 the D500 would have to be at f5 and ISO 3,200 to get the same picture.

        If the crop and full frame are taking pictures at the same aperture and ISO then they will be different pictures (different DOF in the image). If you correct to get the two cameras in the same available light taking the same picture (field of view, DOF, shutter blur) then the 1 stop drop in aperture in dx gives 1 stop back in ISO noise. DX dropping the aperture will also get your diffusion issues cancelled out.

        I’m not saying the 300 at f5 is as good as the 400 or 500 at f8 or that the D500 sensor is only a stop worse at these higher ISO’s, I don’t know.

        In the real world if one guy had a DX and the other a FF taking the same picture of the same bird and assuming the DX guy had a fast enough lens and the FF guy has a long enough one to get equivalent settings then in theory they will produce the exact same quality image.

        I’d be curious to see true side by side D500 v D5 equivalent picture comparisons if you have any. Maybe the differences will be all in the lenss, both the lens focus and optics qualities. Is a D500 Sigma 120-300/2.8 S a “poor mans” D5 200-400/4 ?

        Maybe I’ve digressed too much and it’s a topic for another day!

        * 300mm dx v 500mm ff not perfect focal length equivalent but I assume you’ll understand the point

        Reply ↓

        • Hi John

          The D500 is a cracking camera – throw a battery grip under it and a 300mm PF and you’ve got a really useful wildlife/sports/action setup. But it needs light to work at its best, and being a crop sensor it lacks the wide angle versatility of an FX body.

          Most lenses tend to deliver a good sharpness/DoF balance at f8, but using smaller apertures on the likes of a 500mm prime is pointless, so I would never advise anyone to go to f12 because of diffraction/higher ISO/slower shutter.

          Here’s a shot from the D5 to make a comparrison – https://www.wildlifeinpixels.net/blog/wp-content/uploads/2017/08/D5A7857.jpg – but it’s 5600 ISO, because I’m at 1/2000th shutter, and the D5 seems to consistently meter +0.33Ev higher than the D500, hence the 2/3 stop difference.

          If we were to take an FX and DX camera with identical focal length lenses, and shoot the same subject at the same distance and aperture, then the DoF would be the same in both resulting images. If we were to move the FX camera closer to the subject to match the composition of the DX camera then the FX would have LESS DoF. That’s the advantage of DX shooting – you don’t need to be so close.

          But that is DX’s ONLY advantage.

          A 20Mp DX will always lose out to a 20Mp FX in terms of dynamic range (Edit 23/8: this should read DR/DoF – thanks Grammarly!) simply because of photosite size differential. Add to that the differences in ADC/SNR firmware that Nikon programme the camera with, and you have the full explanation for the overall IQ differences.

          I reffuse to comment on Sigma tele-zooms! But a D500 is not a “poor mans D5”, and such a thing does not exist. And a Nikon 200-400 is a pig to use – I should know as, I’ve owned one for years!

          For some, a D500 DX camera is a wise choice, and it’s an epic little camera if used within its ‘operational envelope’ so to speak. But that envelope is not as big as an older FX D4 or D4S, so for others the purchase of an older FX body will fullfill their needs better.

          Reply ↓

  2. Hi Andy,

    Consider this an off line note as I don’t want to clog your comments up but I think you might be basing your some of your logic on an incorrect assumption.

    Note that having just said that your conclusions may well still be right based on your hardware results you’ve tested! So I don’t mean to offend but I thought I should point out the assumption I don’t think is quite right.

    Your comment “If we were to take an FX and DX camera with identical focal length lenses, and shoot the same subject at the same distance and aperture, then the DoF would be the same in both resulting images” I don’t think is correct.

    In photography terms, i.e. taking into account the circle of confusion, then they do not have the same DoF in the resulting image.

    For example take 20MP DX (CoC 19um, pixel 4.4um) and FX (CoC 29um, pixel 6.6um) sensors.

    And both with 300mm at f5.6, iso 3200 pointing at the same bird at 10m:

    For DX that gives a 20MP final image with FoV 5.4 deg and DoF 23cm, airy disc to pixel size ratio 1.72

    For FX that gives a 20MP final image with FoV 8.0 deg and DoF 35cm, airy disc to pixel size ratio 1.14.

    Because FX has a much wider FoV the bird is smaller covering less pixels in the final image so it takes 35cm of bird to tell it’s OOF. For DX you have a bigger image of the bird filling your final image covering more pixels so you can see that at 35cm it’s getting soft and in this image it looks sharp for 23cm (that’s where the CoC comes in). So the image DoF reports larger for FX even though they were taken with the same physical lens.

    But if you make the DX lens equivalent to FX (one that results in the same looking final image, bird the same size/pixels as FX and looking sharp for 35cm) then in this case a 200mm set at f3.8 would do it.

    Now you have to drop the iso to ~1400 (so you get your noise reduced). Because you dropped the aperture your airy disc gets smaller. Then you get pretty much the exact same image:

    FoV 8.1 deg and DoF 36cm. Now the airy disc to pixel size ration of 1.16 is the same so they have same diffraction performance.

    On line calculator here:

    https://jscalc.io/calc/Dt49M7yj19JtxhQQ#%7B%221%22:200,%222%22:3.8,%223%22:2000,%224%22:1400,%225%22:20,%226%22:10,%227%22:44,%228%22:33%7D

    I know some of the values from different places are very slightly different but that is down to the optical approximations and rounding in the different models used. I have a simple Excel based sensor optical model I can e-mail you if you are interested in playing around with combinations.

    What’s interesting is that the physical aperture is the same at these two equivalent settings (300/5.6 = 54mm, 200/3.8=53mm), the lenses are just focusing the same field of view onto a different image circle size. DX is letting in the same amount of light as the FX, a 53mm hole for 1/2000th of a second. FX has to spread the same amount of light out thinner over the larger sensor hence the lens is moved back to 300mm and the f ratio changes. DX can spread it “thicker” as it focuses closer on a smaller area and so can use a lower ISO. The 4.4um pixel is getting the same total light/number of photons as the 6.6um pixel (both 53mm x 1/2000th worth of light shared across 20Million pixels).

    It’s swings and roundabouts. For a fixed ambient light there is only so much “3D information” you can capture through a 53mm hole in 1/2000th of a second. 3D being x,y = FoV and z= DoF across 20M pixels. You share it around between FoV (FL), DoF (f-stop) and image circle size/iso.

    So in theory if you choose the focal length and apertures to be truly equivalent you can get the exact same image (including noise and diffraction performance) on DX and FX. If you don’t make them equivalent then you will be just looking at different image quality only because they were given different amounts of light to work with.

    Theory is one thing, manufacturing limits, costs and technology are obviously another! So the difference between DX and FX images should be down to the camera and lens and their technology/performance compromises and not in theory just due the physical differences in sensor size.

    There is a big overlap in the normal focal ranges and apertures where you can get good equivalent images on DX and FX. I think real world DX suffers for lens quality at UWA but I’m not sure for the longer FL ends. But typically DX lenses are usually built to a lower cost unlike some of the more exotic FX lenses. That’s why I’m curious to see just how close in system performance say a D500 and a 300/2.8 might be compare with say a D5 and a 500/f4 (as you can’t get a 450/f4). One day I might be able to afford the former set up.

    Anyway, I don’t mean to be critical or offend. It’s just that you might be one of the few people with the opportunity and knowledge (if not the time or inclination!) to test real world differences for truly equivalent settings and see just how close the modern DX systems can get.

    These are some interesting reads:

    http://www.josephjamesphotography.com/equivalence/

    https://www.dpreview.com/articles/2666934640/what-is-equivalence-and-why-should-i-care

    https://www.dpreview.com/articles/5365920428/the-effect-of-pixel-and-sensor-sizes-on-noise

    Cheers,
    John

    Reply ↓

    • Hi John, we are both right! I, in my blurry eyed quick reply I never noticed that the damned Grammarly plugin in my WP editor deleted the /DoF after dynamic range – I actually typed DR/DoF and the plugin went nuts! Apologies!

      One thing though, be careful with third party DoF calcs as they sometimes use generic CoC values, not true ones. With traditional FSI Bayer sensors the surface tech and wiring substrate can both effect the true CoC value.

      This means that the sensor CoC in one model of say, 20Mp DX camera, will not have the same exact CoC as the sensor in another 20Mp DX model. And we can say exactly the same for FX as well, it makes no difference.

      Sensor format has no effect on DoF or anything else, other than AoV/FoV. Pixel density is what makes or breaks the deal.

      Reply ↓

Add your comments and feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.