Camera Reviews
A full review of ISO & Resolution Performance of the Canon 5D Mark 2, Nikon D700 & Nikon D3x
THE LINKED WEB GALLERIES IN THIS ARTICLE ARE PASSWORD PROTECTED, ENTER ALI FOR ACCESS (Case sensitive)
When I bought my 1Ds Mark I in January 2005, I marveled at the crisp, detailed images it produced coupled with a Canon 500 f4 IS lens. Using it for bird photography was a joy, as long as I kept it below 400 ISO, and even then the noise needed to be subdued in post processing.
http://www.availablelightimages.com/galleries/portfolio/content/B91U0016_1_Edit_large.html
Then, as I switched to Nikon in the summer of 2009 in search of the blistering performance of their 14-24 f2.8 wide-angled zoom combined with the High ISO noise capabilities of the D700, I was further amazed when I was shooting at 2500 ISO at 17500 feet in Western Tibet to capture the Milky Way over Mount Kailash.
http://www.availablelightimages.com/galleries/portfolio/content/_DSC3391_large.html
Cameras have come a long, long way in the last half dozen years and envelopes are being pushed with each new release. The Nikon D3s now holds the title of the best performer at high ISO in the current DSLR line-up, who knows what 2011 shall bring?
Now, in the New Year of 2010 we are in a position to do a side-by-side test of three current DSLR bodies to compare their attributes, particularly noise as the ISO is pushed up.
My wife has held onto her Canon lenses and we recently bought her a 5D MarkII as her primary body. It seemed the perfect combination of resolution and performance to meet her needs; macro, butterflies, portraits and some landscape work. I still have the D700 I bought in July 2009 and this week I bought a second hand D3x. Many reviewers of the latter camera in the past bemoaned its high cost and slow speed as a negative and I am not too keen to go into that first issue, but having previous owned a Hasselblad H2 and a Phase One P45+ 39MP back, by comparison, THAT is expensive kit.
The main purpose of the following tests was for me to take a good look at two primary things that are my top priorities: Performance at higher ISO and resolution/sharpness. For landscape, studio and macro work all three bodies are highly capable, and even at a now relatively slow 5fps they are faster than the Phase One by a mile. I sold the latter last year, so unfortunately couldn’t include it in tests, but I do have many files and include one in full size just for a benchmark.
This is the first time I have conducted such tests and doubtless there are flaws in my methodology, for example, I shot in Manual mode in all cases, but let the cameras meter themselves, maybe I should have shot all three each at a selected shutter speed and aperture instead. There are differences in the way the three bodies see the world, and this was evident in the RAW files, but I did manually standardize the White Balance when I converted them all in ACR to 3400K to match the lights in my office.
Here is a shot of the scene I used to test the cameras with, set up in a dim corner of my office, no natural light, lust a couple of small spots.

With the Canon 5D Mark II I used a Canon 24-105 f4L @ 50mm
For the D700 & D3x shots I used a 24-70/2.8 also at about 50mm
In all cases I shot @ f8
I shot each camera at 200/400/800/1600/3200 & 6400 ISO and posted all the full-sized images here:
http://www.pbase.com/availablelightimages/full_size_scene password: ALI
It saves my bandwidth on my website and also allows you to download the images for your own analysis.
The three cameras were basically shot in “Normal” mode. As follows:
Long exposure Noise reduction OFF but I had to expand the ISO on the D3x to H2 as their default maximum is 1600.
The second sets of images are here:
http://www.pbase.com/availablelightimages/12mp_equivalent_crops password: ALI
I cropped all the cameras to the equivalent of the D700’s 12MP and downsized the D3x and 5D2 images to the same dimensions as the D700. These are 100% crops, grouped by ISO
Finally, I re-shot the scene using a Sigma 150 f2.8 macro that we have for both Canon & Nikon.
These crops are of the same section, but are at the cameras native resolutions; the size difference is quite evident.
http://www.pbase.com/availablelightimages/native_resolution_crops password: ALI
The shots in many ways speak for themselves, all three are clearly very capable bodies and I would have no hesitation being put in a tough spot with any of them. I am also aware that there are endless reviews of them all over the Internet, so I’ll save myself the trouble of writing it all again, but here are my 2c’s worth.
On the Nikon D700 the viewfinder is considerably less than the 95% that Nikon claim it to be, it is probably between 90-92% of the scene, to the extent that I had to recompose whenever I put that camera on the tripod. The viewfinder is also a lot darker than either the 5D2 or the D3x.
The Canon does not have the same build quality as the D700, feeling more plastic and less durable than the Nikon. The D3x is obviously at the pinnacle of DSLR build quality and instills faith as soon as you hold it.
Finally, coming from a Canon background and using their remote timer cable releases (TC-80N3 & RS-80N3) that just clip into the socket a bit like a Apple MacBook pro magnetic power cable, having to now deal with the Nikons MC-36 that you have to screw into the body is a pain. At high altitude in -25C conditions in the dark, this is a hassle. I have to prepare that before I leave the tent and hike to the scene, which is not my preferred method.
Finally, the D700 was noted for it’s high ISO performance and that is backed up by my own experiences after 11000 frames since July 2009. It is however, a little softer than the two big brother cameras. The D3x delivers spectacular resolution and I feel up to 800 ISO is unsurpassed. The 5D Mark II is an incredible performer in both ISO and resolution and at a very friendly price. The way Nikon deal with noise is quite evident, with Canon allowing more color noise than the two Nikons.
My final word is this, splitting hairs at 6400 ISO is a remarkable place to be, I would use any of these cameras in the 1600 to 2500 ISO range with no hesitation, and for most of us, that is a pretty amazing solution to some age-old problems. Photography is moving into a new age, with shots now possible that were in the realms of fantasy a year ago.
Please respect my copyright, the images can be used for personal use for your own evaluation only, please consult me for any other use.
Hi Alibenn,
While I’m still reading through your review (and so far I feet it is very well put together!!) I take the opportunity to notify you of a possible mistake in the naming of your pictures. Eg: the ISO6400 tests have the D3x images with D700 Exif, and vice-versa. Can you check it out and clear up any possible confusion?
Thanks.
Fabian was quite correct, I had mis-labelled the images on the PBase gallery and have corrected it now, the EXIF data does clearly state the files true identity. Thanks for spotting that, it is now correct.
Nice article.
I like the D700 and 5D2 also, packages that offer great value for money.
D3x is like a 1Ds3, an indulgence in indulgences……
An H4D perhaps??
Excellent as usual sir!!
Cheers Rob, when you factor in a paid US$5100 for the D3x and it’s in perfect shape, the indulgence seems a bit more reasonable.
I would observe that in most of the reviews I’ve read, they either capture to RAW or to RAW + JPG. Nikon JPGs tend to always be softer than the same scene captured and recorded RAW. The processing of the RAW down to JPG compromises the capture a bit. May affect noise, also. You might want to do a similar series, but record to RAW, then open (no processing) and show 100% crops of representative areas. Just a thought.