Full-Frame Group Test: Part 2
In part two we look at two great value ways of shooting 35mm full-frame digital, with both of our two subjects leaving you plenty of change from £2000.
We are, of course, still talking a great deal of money but shooting full-frame is an expensive business unless you consider secondhand where you can find products like the original Canon EOS 5D offering great value for money.
The camera itself is a significant part of the cost of going full-frame but the cost of lenses must be borne in mind. If you have been a 35mm film user you will have lenses that suit full-frame digital use, in theory at least. If you do have lenses from your film days it is worth remembering that some of these optics are unlikely to make the most of the larger sensor, especially as resolution continues to gain megapixels. This is why most notably Canon and Nikon have been gradually updating their respective lens ranges and using new coating formulations to minimise flare, ghosting and to help control optical aberrations. However, if you have been shooting APS-C format, you may have to dig into your pocket and start your lens collection afresh if you want to enjoy the larger format and make the most of its full picture potential.
Another aspect of shooting full-frame to consider is file size. Full-frame images, especially Raws, are big so you will have to invest in a few more storage cards.
Also, a little further down the workflow chain, the matter of long-term storage and external hard drives should not be forgotten either. The good news is that 2TB hard drives are available for £70 or less, so if you can afford one of these DSLRs, this is not too arduous.
Our test cameras tried in RAW and JPEG image MODES
The images shown below from the Canon EOS 5D Mk II and the Sony A850 were taken at the same time as those that featured in last month’s part one of this major test. The point of that, of course, is that you can make a direct comparison with the images from the Canon EOS 1Ds Mk III and the Nikon D3x from part one. Each camera was mounted on a Manfrotto 055 carbon-fibre tripod to ensure stability.
On the EOS 5D Mk II was fitted the £900 Canon 24-105mm f/4 lens while on the Sony A850 we had the £630 28-75mm f/2.8 SAM standard zoom.
Auto white-balance and aperture-priority AE were used with an aperture of f/11 for optimum lens quality. The Sony gave an exposure of 1/50sec while the Canon exposed the scene for 1/15sec. Both cameras were set up to deliver the highest quality Raw and JPEG files.
The Raws were converted in Adobe Lightroom 3 as well as their respective supplied softwares, Canon Digital Photo Professional and Sony Image Data Converter SR, in their default settings.
The EOS 5D Mk II gave files of 60MB in 8-bit measuring 47.55x31.7cm at 300ppi.The Sony yielded even bigger files at 69.8MB when opened up in Photoshop and you can get 51.21x34.14cm size prints at 300ppi without any interpolation.
Prints made on an Epson Pro 3880 from JPEGs straight out of the camera and converted Raws looked fabulous. There is no way anyone shooting with one of these cameras (or the two models featured last month) would be disappointed with the resulting images – at least with top-end lenses and sound shooting technique. But then you would not expect anything else for DSLRs at the top-end of the price scale.
ISO PERFORMANCE
Testing for digital noise after dark
You’ll notice here with the Canon EOS 5D Mk II that the first ISO shown is ISO 200. There was no space for ISO 50 and ISO 100 images – also they looked almost identical to the ISO 200 so nothing was to be gained by showing them. In the case of the Sony, its highest speed is ISO 6400.
Image quality is generally very impressive with ISO 400 and 800 – good enough for critical use. The Canon seemed to be slightly less noisy but the difference is fairly minor. That small difference remains at the faster ISOs, with the Canon maintaining its performance edge. Generally, though, both DSLRs impress.
High ISO noise reduction
For our High ISO Noise Reduction element of the test, we used both DSLRs at their highest proper ISO rating, so ISO 6400 in the case of the Canon and ISO 3200 for the Sony. Both cameras offer three levels of NR.
From our tests, the feature does offer a marginal benefit, as it did on the two DSLRs last month. The Sony seems to do a better job but then it needs to as its highest ISO is one-stop slower than the Canon.
Resolution of fine detail does suffer as the strength of NR increases but bearing in mind that you are shooting at the highest ISO settings, this is not an issue. Both are worth using if noise is an issue to you.
REAL WORLD PICTURE TEST
Colour and exposure performance
The first thing I noticed when I started shooting images for the test was that the Canon’s monitor wasn’t quite right. It was set to its default brightness level but the first few images I took looked overexposed. I set -1EV compensation and that gave me better looking previews. However, a quick check of the histograms revealed that the exposure was just about perfect at 0EV and there was nothing to worry about so I took out the compensation.
There was clearly an issue with the monitor on my sample. Despite this, it was immediately apparent that the Canon produced more friendly, warmer images compared with the Sony. The Sony seemed to impart everything with a cool, unwelcoming hue. In Raw this is easily fixed but with JPEGs the importance of correct colour set-up is clear.
In terms of exposure performance, it was difficult to fault either camera. I shot in program and aperture-priority modes using Evaluative and Honeycomb metering modes of the Canon and Sony respectively. If anything, the Sony leaned towards slight underexposure – about 0.5EV – and highlights could look flat and shadows heavy with shots straight out of the camera. It’s not a major issue and easily sorted on the computer.
Generally, both systems coped perfectly well in flat, even lighting, and acquitted themselves really well even with predominately light or dark scenes.
They struggled more in extreme lighting. Interestingly, in some direct into the afternoon sun shots taken in London’s Trafalgar Square, the Canon overexposed the scene quite drastically – drastic enough that even working on the Raw files in Canon DPP did not help much. Normally, to be honest, I would have expected underexposure. With the same scene, the Sony’s 40-zone Honeycomb meter did underexpose a little and it certainly coped better with an incredibly contrasty scene.
OUR VERDICT
Four full-frame DSLRs, and the winners are...
To be honest, picking an overall winner is a thankless and probably pointless task. If you are spending this sort of money, you shouldn’t be relying on a magazine review when you should be getting hands-on.
Moreover, when you’re at this stratospheric price level of the DSLR market, you will not be surprised to learn that the performance differences between them are very minor, and you really do have to be a seriously keen ‘pixel peeper’ to distinguish between them. Ultimately, it comes down to subjective issues like feel, looks and the sound of the shutter.
Taking the four cameras in their two respective price brackets, let’s start with the Canon EOS 5D Mk II and Sony A850. If you want to shoot full-frame and your budget is limited (!) to £2000 or so, then you will be delighted with either. Both of our samples worked well rather than flawlessly, but the images I got are beyond doubt straight out of the top drawer.
If you forced me to pick one, then I would go for the Canon. It sounds better, the AF is slicker and it feels great in the hands. The lens options are also broader, although that could change in time.
Onto the Canon EOS 1Ds Mk III and Nikon D3x, and choosing here is even more difficult. Both are remarkable cameras. They are built to last and it is worth stressing again that you are unlikely to be disappointed by the quality of the images either of them produce. The quality is amazing and getting seriously big enlargements from files is easy. It definitely helps to have a good workflow, a computer with lots of RAM and plenty of storage space. Work in 16-bit and you’re talking uncompressed files of nearly 140MB in the case of the Nikon.
The Nikon D3x definitely needs a sensor cleaning system and that is one thing the Canon is clearly superior on.
If I had five grand burning a hole in my pocket (I don’t) and if I did not have any commitment to a camera system (I do, Nikon), I think I would choose the Canon. But to be honest, that preference is purely subjective rather than based on performance differences and even that preference is fag paper thin so I could change my mind tomorrow. These cameras are all that good.
Taken from the May 2011 issue of Advanced Photographer magazine