Oh No! Not a Nikon / Canon Comparison !
Yes. I'm Afraid So
with 24-120mm f/3.5-5.6 @ ISO 25,600
Attendees of the Jay Maisel seminar at The
Luminous Landscape Gallery watching a
video slide show.
The illumination in the gallery was so low when this shot was taken that it was too low to read.
Regular readers will have noted that I shoot primarily with Canon cameras and lenses.
Why? Because I have to shoot with something – right? In about 1996 I switched from 30 years of shooting Nikon because I was attracted to lenses which Nikon didn't offer at the time, the Canon's T/S and IS lines.
I had few regrets at the switch through the first 7 years of this decade during the transition to digital, because it's my opinion that with its full frame sensors and superior high ISO noise capabilities Canon had a clear edge. Many other pros apparently thought so as well, and as a consequence Nikon lost some market share in this segment.
But with the introduction of the D300 and D3 a few months ago, and the commencement of their shipping during the last week of November, '07, I now believe that Nikon is back in the game. For this reason I recently purchased a D300 and a few lenses to begin refamiliarizing myself with the current Nikon line, and mentioned this on What's New.
As could be expected, a number of ill-mannered individuals with an apparent axe to grind started commenting on other discussion forums (no guts to do so here, it would seem) ascribing all sorts of motivations to this, most of them uncomplimentary if not downright rude and unkind.
No rebuttal is warranted, but since words on the net last forever I will make the following comments, just to set the record straight. This site is designed to reflect my personal interests in photography, both the art and the craft. That means I write about the products and software that cross my desk, or which I personally own and use. But with some one million unique readers a month this site needs to reflect from time to time the interests of a broader constituency, and in this case I felt that my lack of Nikon coverage was doing a disservice to that segment of the community, especially pros for whom Nikon is their tool of choice.
So, this isn't about my needing more cameras, or switching from Canon to Nikon, (which I'm not), or having money burning a hole in my pocket. And lets not forget that by adding Nikon to the mix on a regular basis I'm increasing my work load, not to mention the confusion factor when working in the field.
Rather, this move is intended to expand my abilities, coverage, and therefore audience. This obviously isn't about advertising, because the only ads on this site (other than my own) are from B&H Photo, and they'll happily sell you whatever brand you wish. They're platform agnostic. So those that think there's a commercial motivation to this are simply being malicious and are spouting uninformed nonsense.
I just had to get that off my chest. End of preamble.
Jay Maisel – Toronto, December, 2007
Over the weekend of December 1 – 2, 2007, my friend and colleague Jay Maisel taught a seminar at my Toronto Gallery / Studio. I mention this because the day before he arrived in Toronto I took delivery of the Canon 1Ds MKIII that I had ordered after my Madagascar shoot, where I took some some 5,500 frames with a pre-production MKIIIs. And then, on the first day of the seminar, one of the attendees brought in his just acquired Nikon D3, among the first to be delivered in the US. Add 20 enthusiastic amateur and professional photographers, and what did you think everyone (including me, a Canon shooter, and Jay, a Nikon shooter) wanted to do? You got it. Compare the Nikon D3 and the Canon 1Ds MKIII, especially at high ISO, since the buzz about the new Nikon is its claimed very low noise capability at high ISO (up to ISO 25,600).
And so I did – different ISO's, different lenses, different subjects. Nothing terribly formal, but enough tests, with enough pairs of experienced eyes, both Canon and Nikon users, to come to some preliminary conclusions.
The photograph of Jay Maisel above isn't terribly flattering. The lighting in my gallery is designed to illuminate framed photographs, not for shooting portraits. But a low light level / high contrast situation like this was pretty good for testing available light type shooting.
This is where we run into trouble. We can't shoot with the same lens on both cameras, at least not with the lens selection we had available that day. So I picked two lenses which were at least reasonably close, the Nikkor 24-120mm f/3.5-5.6 VR and the Canon 24-105mm f/4 IS. Aperture was set to f/6.3 and the shots were taken in aperture priority mode at ISO 1600. The Nikon exposed at 1/50th second and the Canon at 1/40th second – not identical but close enough.
Why ISO 1600? Because that's the Canon's highest setting without going into a boost mode. The Nikon doesn't go into a boost mode until above ISO 6400. I know that if we did the test with one camera at a settable ISO and the other boosted I'd hear no end of it. I also did shooting at other speeds, but not in a somewhat controlled comparison like this.
Of course the Canon 1Ds MKIII is a 21MP camera and the Nikon D3 is a 12MP camera. This made for a difficult if not unfair comparison in many respects. How to you frame the shot? Same image size or same resolution? I don't know. Both approaches are flawed.
In the end I said, screw it. I don't have all the answers. Let's just shoot a head and shoulder and see what we get. Whichever way I do it someone in the online world with greater tribal loyalties than common sense will get pissed off and claim foul, so I'll just do a test that satisfies my curiosity, and people can natter about it one way or the other if they wish.
Canon 1Ds MKIII @ ISO 1600. 100% enlargement
Nikon D3 @ ISO 1600, 100% enlargement
Both cameras produced ISO 1600 images of astonishing quality. Would a Canon 5D, reputed to be to date the lowest noise 35mm DSLR yet, be any better at this speed? I have no idea, since we didn't have one handy that day. What about the Canon 1D MKIII? Ditto. There wasn't one on hand. Would that have been a fairer point of comparison? Probably. The 1D MKIII's sensor at 10MP is closer to the D3's 12MP than the 1Ds MKIII's 21MP, and the pixels of the Nikon are bigger, so theoretically of lower noise. The D3 is also closer in price to the 1D MKIII than to the 1Ds MKIII.
Right. We agree. But, the D3 is Nikon's current flagship, and the 1Ds MKIII is Canon's. Both started shipping within a day or two of each other in late November. How could we not resist the comparison?
And what did we see? What you see above. Put aside the use of different lenses and the different sensor resolutions and just look at the noise and IQ characteristics.
I see comparable resolving power and clarity (not absolute resolution, of course), and at least a one stop noise advantage to the new Nikon. Possibly there's a bit more shadow detail in the Canon frame though (look at the iris). The Nikon D3's low noise edge is also apparent at ISO 3200. The speeds of ISO 6400, 12,800 and 25,600 are also possible on the Nikon D3, but not on the Canon, so there's no comparison to be made at these levels. And of course at the same magnifcation level the 1Ds MKIII provides a larger image because it has more pixles. There's no free lunch.
The photograph at the top of the page was shot at ISO 25,600. Is it noisy? Yes, of course it is, but it's freek'n ISO 25,600. I don't care that it's noisy. This kind of sensitivity provides the ability to capture an image when another camera would simply record soot and shadow, or be forced to use a shutter speed so low as to render an unusable image. And unless you're a total gearheaded pixel-peeper, getting a photograph in tough conditions is much more important than any noise complaints.
Nikon D3 at ISO 25,600. 100% Crop of image at top of this page.
What About The Nikon D300?
The Nikon D300 was also on hand, and so I did some comparison shots with it as well.
Nikon D300 @ ISO 1600, 100% enlargement
Sad to say, it's not in the same low noise league as either the D3 or the 1Ds MKIII. Noise is more apparent, and though the exposure was 1/50th second vs the D3's 1/60th second (essentially identical), the image was overall quite a bit darker, indicating that the ISO was likely not what the camera indicated it was; maybe ISO 1000 in reality, and the apparent dynamic range somewhat less as well.
Nikon D300 with Sigma 10-20mm f/4-5.6 @ 10mm and ISO 3200
And again, before anyone disses the D300, I have been doing quite a few ISO 1600 and ISO 3200 images during the week prior to writing this article, and find them to be more than acceptable. If you're a Nikon user, or thinking of becoming one and don't have the budget for a D3, the D300 is an extremely capable camera, as will be seen in my forthcoming review.
I also thought it worth throwing the D300 into this comparison because clearly the D3 is not the end of the road for Nikon's flagship camera line. A higher megapixel full frame chip is absolutely inevitable, and the D300's noise characteristics may give us a hint at what to expect from a higher resolution D3 sensor. The only question is – when might it be introduced, and what will be the pixel count? Some informed sources have said that it will be 24MP and that introduction will be Q1 2008, with first shipments in Q2. We'll see.
As for pixel count, 24MP makes sense if you scale up the 12MP D300's 1.5X chip. So, does this mean that the noise characteristics of the D3(?) will be similar to the D300? Possibly, and so that's why I thought it worth including the D300 in our little experiment.
And, as for those that have asked why, since I'm now going to be putting together a Nikon outfit to use along with my Canon gear, I didn't get a D3 instead of a D300, the answer is that just as I own a 1Ds MkIII rather than a 1D MkIII, I also will want a D3(?) rather than a D3. It better suites my needs.
The Bottom Line
As I wrote a little while earlier, based on initial impressions of the D300, and now their first full-frame CMOS camera, the D3, Nikon is now back in the IQ game. Canon has enjoyed a deserved reputation for high image quality, especially at high ISO. But now with the D3 Nikon has upped the ante, producing a camera that appears to offer lower noise than Canon (at least against the flagship 1Ds MKIII), and higher available speeds as well.
Canon 1Ds MKIII @ ISO 3200
100% crop at ISO 3200
Before people draw any unwarranted conclusions it needs to be said that the 1Ds MKIII is no slouch when it comes to high ISO performance. The shot above and its 100% crop show what the camera is capable of at ISO 3200. In a 16X20" print none of the artifacts seen in the 100% crop above is visible, and noise is extremely low. As I found on my Madagascar shoot, you can work at virtually any ISO under any conditions and get usable, saleable images. And don't lose sight of the fact that you can do so at 5 FPS in 14 bit mode, and with 21 Megapixel files – the previous domain of medium format.
And what about other performance and operational aspects? Those judgments will have to wait for me to spend more time with a D300 and D3, but will follow in the weeks and months ahead.
One final comment if I may. If you're one of those people that takes exception to either my tests, my methodologies, or my conclusions, do us both a favour. Come onto this site's forum and state your objections here. Don't do it on some other site where you know I won't respond. If you're going to disagree with me, do so to my face. Regulars on my forum know that I never censor anything other than spam or obscene language. But be prepared to defend your position, just as I will mine. And, whatever you decide to do, do try and be grown up and polite. Anything less just obscures and defeats whatever validity there me be in what you have to say.
I forgot to mention that all images were shot raw and were processed in Lightroom 1.3.1. I don't care for Canon's DPP, and haven't yet become familiar with Nikon's Capture NX. Both of these might have done a different job on the raw files, but that's something for debate.