As promised, here’s an update to the Nikon D800 first impressions review I posted last night. I’ve now had the chance to shoot with the camera for several hours under bright daylight conditions (read: no problems with running out of light, base ISO and nice high shutter speeds) and want to share some more images, impressions and report back on a couple of things.
Firstly, I think I need to clarify a few points that have repeatedly come up on both the forums and in the comments to the first part of the review.
1. AF is not as fast as the D4, nor do we expect it to be.
The difference is similar to that between the D700 and D3 – probably down to battery voltage and the current available to drive the lens motor. It seems subjectively the same as the D700 under very low light conditions – specifically, speed and tracking ability. However, I was using an 85/1.4 G wide open to see if the camera could track moving objects with it – the answer is, hit and miss. Sometimes yes, sometimes no. I suspect the actual AF module performance is no worse than the D700, but with the increased pixel density of the D800, the demands on focus accuracy just got a lot higher – and that’s the shortfall we’re seeing here.
People seem to get angry and anxious when I say it isn’t as good as the D700 at the pixel level – I don’t know how it could be, the photosites occupy half the area! (Probably less, once you take into account the additional power and read circuitry required to run the sensor.) HOWEVER, if you downsize to 12MP to match the D700 (or print both at the same size) – the D800 is better, markedly so. I’d put it a stop ahead for noise, and there’s of course the extra detail.
3. A lot of people have said the images are noisy. Yes, they are at the pixel level, but remember a) mixed light sources; b) I run zero noise reduction and sharpen fairly aggressively – I much prefer grain to smearing and indistinct edges; c) this is a worst-case scenario, overall.
4. Usability and ‘demandingness’.
To achieve the same PIXEL LEVEL quality as the D700, you’re going to have to up your game. And the camera itself cannot deliver the same level of quality at the individual pixel level (think 100% enlargement on screen) because the pixels themselves are smaller. Those are laws of physics. In reality, this means shooting at one stop lower ISO, and taking care with camera shake. If you’re just talking about the overall appearance of a print at a given size, see #2.
With that settled, let’s move on to the update part of the review. This is to address performance of the camera under bright daylight, i.e. close to optimal conditions.
AF, under daylight
Focusing is snappy, positive, and noticeably faster than the D700. I was testing the AFS 28-300/3.5-5.6 VR today – not known for being a snappy lens – but nevertheless, the picture was in focus before I expected it to be. I often re-focused again, because I wasn’t sure it had locked – subsequent testing revealed it always had. AF-C mode is best described as being a little bit skittish – you can hear the AF motor chattering away as it tries to keep the lens elements in optimum position. Whether this is because of the new AF system, the interaction between the camera and the lens or something else, I don’t know. Tracking moving subjects – in this case motorbikes coming towards the camera – was no problem at any focal length using the 28-300VR. I’m certain performance would be better if a lens with a faster motor was used.
In short: at base ISO and sufficiently high shutter speeds that camera shake isn’t a concern, there’s more resolution here than you can shake a stick at, even with decidedly ordinary lenses: even the 28-300VR delivers pretty amazing levels of detail. With the Zeiss ZF.2 2/28 Distagon, it’s on par with the entry level medium format systems. Retouchers, beware. This thing is going to produce files that take two or three times as long to fix as previous cameras, simply because there’s so much more information here. I would honestly recommend NOT getting the D800E if you’re shooting portraits, because it’s going to produce downright unflattering results for anybody without absolutely perfect skin or makeup. Bottom line: the D800 delivers what you’d expect it to, and in a most impressive way. The anti-aliasing filter in the regular D800 is evidently very weak; fine detail remains well resolved, but simultaneously it’s just strong enough to prevent moire. I haven’t seen any evidence of it in the ~2000 images I’ve shot so far, even in fine repeating textures (which you’ve be surprised by how many of them there are when you have this much resolving power).
Some more resolution examples:
Nikon’s claims about more dynamic range than the D700 are true, but must be accompanied with a caveat. I’m finding that while there was a lot of recoverable headroom in D700 files, there isn’t so much in the D800 – however, there’s more useable shadow detail and less noise. Subjectively, I think we’ve lost 1 stop in the highlights and gained around 2 in the shadows; this at base ISO. You could probably pull a bit more out of it with judicious use of the right sliders in your raw converter, but then color accuracy starts to wane. At higher ISOs, color accuracy in the shadows is a bit suspect and heavily influenced by the ambient light source.
White balance, color and tonality
The D800 seems to deliver the same general white balance as the D700, but with a slightly different tonal response. I can’t put my finger on exactly how it’s different, but the files are quite reminiscent of the Leica M9′s output – my color profiles for that camera actually deliver better results during conversion than the D700′s profiles. I can only put it down to new sensor architecture, or perhaps a change in the filter pack in front of the sensor. It’s definitely more pleasing, that’s for sure. Note that I’m talking about RAW file output converted via ACR for both cameras, which removes any manufacturer-specific processing.
The other crop modes
I’m actually finding these surprisingly useful. The 15+MP file you get from the smallest DX crop is still a serious amount of resolution. On a personal note, I’m also starting to like 5:4 a lot. 3:2 is increasingly feeling like no-mans’-land between 16:9 and square for me.
Seems even better than yesterday after a full charge and cycle – I shot 500 frames today, and at the end was only down by 20% – again with heavy LCD use and mucking around in the menus. I was also using a VR lens, which wasn’t the case yesterday. That means an estimated 2,500 shots per charge – on par with my D3, as far as I recall. I don’t know what Nikon have done with power consumption, but it’s impressive. I don’t think I need to buy that third battery anymore.
Commentary on the AFS 28-300/3.5-5.6 VR
This lens has been a bit of a mixed bag for me in the past – it’s so-so on the D700, good at some focal lengths (the longer end) and downright unusable at 28mm until you hit f5.6 or preferably f8. Oddly, it did pretty well on my D5100; enough that I’d actually use it. On the D800, it’s a big, big surprise. The midrange is excellent at f5.6 and outstanding at f8 – we’re talking about 35-200mm or so here; the ends are slightly less good, with the 28mm lagging slightly behind the midrange (but useable wide open, if slightly hazy due to flare) and the 300mm end being just okay to good. Still, it’s a surprise given the resolution of the sensor. I didn’t think the lens was capable of resolving this well. Don’t get too excited though, while it delivers excellent macro contrast, micro contrast structure lags far behind the Nikon primes, let alone the Zeiss primes. Look out for a full review of this lens in the near future.
The shutter mechanism
Although it’s a bit more hollow-sounding than the D700, and frankly I was a little disappointed it wasn’t as smooth and well-damped as the D7000, it’s got one other trick up its sleeve: low vibration. Again, subjectively because I have no way of testing this, the D800′s shutter and mirror mechanism has a lot less recoil than the D700 – this is very, very, very important because it helps us reduce camera shake. Bravo.
Initially, I thought I’d shoot this camera at full size raw and then reduce by 50% to 18MP; not so. Instead, I’m processing at full resolution but forcing myself to be more selective about the keepers. I guess it’s a rare example of a camera actually driving you to be a better photographer – which isn’t necessarily a bad thing. I still need to go out to buy more hard drives, though. MT
Look out for more updates and images over the coming days and weeks. I’ll be shooting some studio assignments this week and next week, so I’ll report back after on how the camera performed. I’ll also try to make some video with it in the near future.
Visit our Teaching Store to up your photographic game – including Photoshop Workflow DVDs and customized Email School of Photography; or go mobile with the Photography Compendium for iPad. You can also get your gear from B&H and Amazon. Prices are the same as normal, however a small portion of your purchase value is referred back to me. Thanks!
Images and content copyright Ming Thein | mingthein.com 2012 onwards. All rights reserved