This is a thorough review of the Nikon D850, based on my experience with the camera over the last few weeks. There is another, albeit shorter review, planned for January, once we have had a chance to use the camera for more landscape and wildlife photography during our trip to Costa Rica, so watch this space 🙂
This review is based on a combination of real-life shooting scenarios (events, portrait and a lot of casual shooting) and set up test scenarios, to see how far this camera can be pushed. I haven’t used the camera at a wedding yet, as until recently it was not supported by Lightroom and therefore camera profiles were not available.
Any photos in this review are included to illustrate a specific point, so if you are expecting dozens of pretty shots, I’m sorry to disappoint 😉 I think it’s beyond anyone’s doubt that this camera is capable of producing absolutely stunning images, hence why I decided that including such images wouldn’t be necessary. Some of the images were processed to show off the camera’s capabilities rather, so they are not necessary in line with our usual style.
The Nikon D850 has been around for almost 2 months now and if you are interested in this camera, you’ve probably read quite a few reviews already. I got the camera only a couple of days after its UK launch but I didn’t rush the review as I wanted to fully familiarise myself with it rather than just base my opinion on my initial thoughts during the inevitable ‘honeymoon period’.
Anna and I have been using mainly Nikon’s D750 for the last 2 years, as well as the Sony A7RII and the Nikon D5, so I’ll be making some comparisons, especially to the D750. Those of you looking to upgrade their D750 may find these bits particularly helpful. On the other hand I have no experience in using D800 or D810, so can’t really draw any comparisons to these cameras.
The camera used for the review was bought by us and we are in no way affiliated or commissioned by Nikon. We are also pretty much brand agnostic – as long as the bit of equipment does what we want, we don’t care what brand it is 🙂
Let’s get it out of the way first – the sensor in the D850 is absolutely epic. It’s also exclusive to Nikon. A while ago you might have heard the statement that the state-of-the-art Sony sensors would be exclusive to Sony cameras. This statement was rather worrying considering that Nikon is using Sony sensors. Well, this statement is not exactly true – it’s a combination of miscommunication, misinformation and the usual marketing fluff. It was Sony Imaging (the company that makes cameras) that started talking about their exclusive sensors, while the sensors are actually made by a different company – Sony Semiconductor. Why is this important? Well, in the past, Sony Semiconductor used to make state-of-the-art sensors, which were available off the shelf to camera manufacturers. These sensors included the well-known and widely used 24MP and 36MP full frame sensors used in the Nikon D610, D750, Sony A7, Nikon D800, D810 and Sony A7R to name a few. What is happening now is that such state-of the-art sensors are no longer available off the shelf. Camera manufacturers have to design their own sensors and Sony Semiconductors then make it for them – which is exactly the way Nikon have been working for years anyway. So, whilst it is true that no other manufacturer can use, for example, the exact sensor that’s used in Sony A9, they can still make their own (better or worse) version of that 20MP sensor. As a result, Nikon D850’s sensor is designed by Nikon and manufactured by Sony Semiconductors. And it is most definitely state-of-the-art 🙂
So, what actually makes this sensor so good? Let’s have a look…
mRAW and sRAW
The images below shows what happens when an image is seriously pushed in editing (almost 4 stops of underexposure with highlights then pulled all the way back; shot at ISO64).
14 bit RAW holds together very well. The colours remain natural and there’s pretty much no visible noise.
12 bit RAW loses integrity in the shadows – the greens are oversaturated and there is some noise.
mRAW has even more oversaturated shadows and more visible noise.
There will be another example showing this issue later in this review.
My another concern regarding mRAW is the fact that Nikon doesn’t make it public how exactly they accomplish making they RAW files smaller. There’s some clever algorithm running in the background and the files are certainly much better than Nikon’s previous attempts at creating smaller size RAWs. This doesn’t change the fact that some sort of resampling has to take place to reduce the file size and I don’t like not knowing what exactly is happening to the image before it’s written to the card.
As m- and sRAW files are 12 bit it is only fair to compare them to 12 bit full res RAW. Ok, you do save around 28% of the file size (25MB vs 35MB), but the buffer capacity remains the same in both cases (approximately 38 photos when writing to XQD and SD simultaneously). This is due to the additional work the processor has to do to make the files smaller.
So overall, the only benefit of mRAW is the somewhat smaller file size, at a cost of being limited to 12 bit and some small loss of image quality in extreme situations.
What’s my preference then? 14 bit lossless compressed full res RAW when shooting at up to 400 ISO and 12 bit lossless compressed full res RAW when shooting at higher ISO settings. While around 50MB makes it quite a hefty file at 14 bit, I can live with it… Sony A7RII files were around 80MB at 14 bit. The buffer and write speeds are good enough for my style of shooting and I know I’m getting the highest possible quality should I need it 🙂
Depending on the ISO and the scene, the file sizes for 14 bit RAW vary between around 45MB and 65MB. Nothing surprising here.
Same scene photographed at ISO64 gave the following results:
14 bit RAW – 47.5MB
12 bit RAW – 35.1MB
mRAW – 25.5MB
Again, no surprises 🙂
The below images have only very modest amount of processing applied to them. Skin tones look pretty awesome to me although I appreciate that pleasing skin tones can mean different things to different photographers. I’m certainly happy with the results.
The dynamic range is absolutely amazing – I’m sure no one was expecting anything less from this camera. At the base ISO and in 14 bit mode, the files are impressively malleable and shadows retain integrity even after pulling them up by 4-5 stops. There’s also a lot of detail preserved in the highlights. Switching to 12 bit or mRAW gives far less impressive results though. The image below is quite an extreme example – exposure was metered for the highlights, resulting in the foreground being underexposed by around 5 stops.
As expected, straight out of the camera there is absolutely no detail visible in the shadows. Let’s see what we can get out of it.
14 bit RAW file pulled up by 5 stops looks very good indeed. There is a little bit of noise and a tiny bit of green cast in the deepest shadows but otherwise the image holds up very well.
12 bit RAW… well, the image speaks for itself 😉 The amount of noise looks similar but there is a horrible green colour shift in the shadows and even in some of the mid-tones.
mRAW shows some more pronounced noise and the colour shift is even more visible. There’s certainly further loss of image quality when compared to 12 bit RAW, let alone the 14 bit version.
And here is the same shot from the Nikon D750 (14 bit RAW, at ISO100 rather than 64 though).
The image quality is certainly comparable to 14 bit RAW file from D850. Arguably even a touch better, but the difference is minimal and the lighting conditions have changed a little bit between the shots.
I’ve done a similar comparison at ISO400 and although the difference is not quite as dramatic, the same loss of image quality can be observed.
As I no longer have the Sony A7RII, I cannot make a direct comparison, but looking at the images shot with it in the past and the images coming out of D850, I can say with full confidence that both cameras have very similar dynamic range. Frankly, I’m not bothered about lab results showing a 0.5 stop difference in favour of one camera or the other, as both perform exceptionally well in this regard. And compared to D5? Well, there is no comparison really. Shadows in the D5 files were falling apart when pushed around 3-3.5 stops, but that camera was never made for maximum dynamic range.
Here’s another example, this time a bit more realistic, of just how much detail can be recovered from heavily underexposed shadows. Shot at sunset at ISO100, shadows lifted around 3.5 stops, only minimal post processing applied.
And one more 😉 This is some seriously impressive stuff, even if somewhat expected from this camera.
I tested the full native ISO range from 64 to 25,600. Up to around 800 ISO the noise is virtually non-existent. At values up to 6400 it is still very well controlled and limited to luminance noise. I’m certainly perfectly happy with using any of these settings if needed. In the 12,800 – 25,600 ISO range the noise becomes far more visible and the image quality deteriorates below the level I’m happy to accept. This is in line with what I expect to see in most cameras – the highest ‘clean’ ISO setting is 2 stops below the maximum native ISO setting.
I always find discussing high ISO settings a bit pointless as everyone has a different perception of what is an ‘acceptable’ amount of noise. I have to say though that at any setting, the noise is impressively well controlled, especially given the resolution of the camera.
ISO12.800 no noise reduction
ISO10.000 no noise reduction And a slightly different subject 😉
ISO3200, no noise reduction, pushed almost 3 stops in processing And below is a 100% crop of this photo… pretty awesome amount of detail and very well controlled noise.
Same as D5 what more is there to say? 😉 – this was the line I first wrote when I started making notes for this review based on my initial observations. Since then, other reviewers have pointed out that D850’s AF tracking is not quite as accurate. While this may be true for very dynamic action, in my experience I’ve noticed no difference in the AF performance between the D850 and the D5. I got plenty of keepers from both cameras and I got some out of focus shots from both. I must emphasise here that this is true in my experience and for my style of shooting. Your experience may be different. It’s beyond any doubt however, that this AF system is one of the most advanced systems currently available. It is also highly customisable so I’m sure most photographers will be able to tailor it perfectly to their needs.
The default card combination I use is the Sony XQD G Series card with the Lexar Professional UHS-II 2000x Speed SDXC card (both 64GB). This setup allows me to write RAW files to both cards without worrying about the buffer (this doesn’t mean that the buffer is unlimited – please see my notes below). I imagine that any combination of cards with similar speeds (like this or this) would give similar results.
I also did some additional tests for comparison. Unless stated otherwise, the tests were done with the camera set to CH mode (7 fps), 14-bit full res RAW with second card set as backup.
Here are the results:
To sum it up – if you do a lot of bursts in quick succession, then using the XQD card only will give you the best performance. When shooting weddings, I cannot be without a backup, so the next best thing is the XQD plus the UHS-II SD card.
What about the 9 fps capability when the grip is attached? Well, for me, the rather hefty cost of the grip, the battery and the charger, as well as the added bulk and weight plus dealing with different batteries and chargers (I still use the D750) is simply not worth it. But I can totally see how it can make sense for others. It’s just another element, which can make this already brilliant camera even more versatile.
I realise that this is a very basic way of using banks and they give potential for far more customisation, but this setup works perfectly for me, as it allows me to change between a few groups of main settings without having to worry that I will accidentally shoot a whole wedding on a small JPEG setting 😉
Nikon, you’ve created a truly amazing camera and I’m struggling to find anything to dislike about it. Yes, certain things could be improved, but this is true for every camera on the market. The D850 is one hell of a camera. In fact, I would go as far as saying that for the vast majority of photographers it is currently the best DSLR on the market, and by some margin (unless you need the extreme speed and durability of the PJ style bodies). For my needs and preferences the D850 is just about perfect.
This doesn’t, however, change the fact that the D750 is still an undisputed king when it comes to the combination of features, small size/weight and great price. Is the D850 a better camera? Absolutely yes. Is it worth over twice as much as the D750? That’s for you to decide. For me, yes, without any doubt it’s my favourite camera ever 🙂
Pure joy to use 🙂
Banding in silent shooting mode
Doesn’t have some of the fancy features the competition has (e.g. Sony’s Eye AF)
Could be a bit cheaper 😉
SD card impacts buffer clearing time (minor)