I have to say, seldom in my life as a writer about photography has an event occurred with the kind of serendipitous timing as Monday's announcement of DxO's new DxOmark.com.
To understand the reason why, I have to go back to a peculiar condition of reviewing. Very often, experienced writers of reviews can decide for themselves what's going on with a certain product far more easily than they can prove it to the satisfaction of others. For example: I've tested so many lenses in my life, and I know the visual telltales of the various optical aberrations and properties so thoroughly, that sometimes I feel I can get a pretty good handle on how a lens behaves just by looking at other peoples' pictures taken with it. There are even some things that you can tell about a lens by looking at pictures online. (Saying this, I'm aware of how perilously close I am to opening a can of worms.*) I know most of the websites where pictures are searchable by lens type, and I've spent happy but ultimately pointless hours sifting through pictures trying to detect what I want to know about lenses that I don't own and have never used.
But of course, I would never dream of making hard-and-fast statements as if they were fact from such suspect research, and still less of publishing such conclusions in a formal review.
And the ultimate "I'd never do that" is in cases where my own conclusions are diametrically opposed to conventional wisdom. If what you think is contrary to what the crowd believes, you'd better be prepared for arguments. Your proof has to be all that much more convincing, your procedures better armored against dispute.
But I'm also an old bull**. I've gotten lazy. It's hard to do all the work that has to go in to a formal review, and I've done it too many times for it to be exciting any more.
On the other hand, I do have my conclusions. I just don't always feel like I can talk about them.
This week I have a Nikon D3 in house. There are a lot of ways in which the D3 is superior to the A900: it's clearly quieter, and the mirror blackout is much shorter, and the camera as a whole is faster, crisper, more responsive (although the A900 is by no means bad). It's also bigger and heavier, and I don't like the user interface as much. Of course, those are things you get used to.
But the kicker is that, frankly, I think the A900's image quality beats the D3's.
Conventional wisdom holds that the D3 is superb at high ISOs, and it does do better than the Sony. But I like the Sony's high-ISO quality better than the Nikon's. Its sensor noise looks more like what I want sensor noise to look like: like grain that's more integral to the picture, more coherent. Does that mean it just looks more like film grain, and film grain is what I'm used to? Possibly.
So can I prove it to you? Well, no, I can't. I don't have the two cameras side-by-side to compare. I know what I think, but that doesn't necessarily convince you. I understand.
Scorn the dark
Incidentally, I think I'm approaching a personal "Theory of Überkameras." I was out photographing with the D3 the other night in what most folks would call "the dark." The Nikon was fazed by this usually daunting circumstance not one little bit. Remember how I was saying the other day that the A900 can see a lot better into the distance than I can? Well, the D3 can see in the dark a lot better than I can. So what's my theory? I'm not 100% sure of this yet, so don't hold me to it, but what I've been mulling over is that maybe I don't really need a camera that sees better than I do. I think maybe if it sees as well as I do, that's enough.
That's just an illustration, but I think you get the idea. (The D3 can actually see in the dark better than that.)
So what does any of this have to do with DxOmark.com? Just this: DxOmark gives the Nikon D3 a score of 2290 for "Low-Light ISO" and ranks it second of the 51 cameras it's tested so far, and the Sony A900 gets a score of 1431 and ranks fifth. That's wildly divergent from the conventional wisdom based on a couple of internet testing sites that judge the A900 based on how it will never be used. But my own subjective impression of the "low-light ISO" performance of the two cameras (shooting RAW and converting in ACR) is much more in line with DxOmark's ratings than it is with all those people shouting about how the A900 is unusable at ISO 400. Given how the conventional wisdom is already stacked against the Sony, and how strident the online arguments have been, would I have dared say this out loud and in public if DxOmark hadn't come along to back me up? Hmmm...maybe. But I would have had to be all mild and mealy-mouthed about it, because I know I didn't run strict enough tests with either camera to prove my impressions to anybody.
Anybody else, that is. I know what I thought.
It's still just my opinion, and needn't mean anything to anybody.
Another aspect of cameras that might be calcifying into a theory is the "shoot-to-carry ratio." When you're shooting, what percentage of the time do you have the camera in your hands, actively shooting with it, and what percentage of the time are you just carrying the camera around? I think pros generally have a high shoot-to-carry ratio. When they're working, they have the camera in their hands 80% or 90% of the time. (Is that fair to say?) I, on the other hand, have a low shoot-to-carry ratio. I probably spend 80 to 90% of the time carrying the camera, and only 10 to 20% actually shooting with it.
I suspect that people with a high shoot-to-carry ratio will like the D3. It's a great camera to shoot with. It's very fast, reasonably quiet (I can't hear the 24–70mm ƒ/2.8 focus, and the quick, instantaneous flutter of the mirror and shutter is addictive), and it has a really formidable flexibility in terms of operational parameters. If you can't do something you want to do with the D3, it's probably because you haven't learned how yet, not because the camera won't do it.
But when you're just carrying it, the D3 is...well, ridiculous. The D3/24–70mm is an order of magnitude bigger than my old Konica-Minolta 7D and 28–75mm ƒ/2.8, which is already too big for me. Note that this is just a preference, and yours might differ from mine. (The camera with the optimum shoot-to-carry ratio IMO, the perfect balance, is an M6 with a 35mm Summicron.) But when I'm carrying the D3 from place to place, I feel no more comfortable than I would with, say, a 10" circular saw dangling from my neck.
The A900 and Minolta 35mm ƒ/2 combo is still big, but lots better.
These are two of the best cameras money can buy at this moment in history, and each one is enormously impressive. Photography buffs who fantasize that either one has significant deficiencies are akin to upper-middle-class Americans obsessing that they just aren't rich enough: that is to say, they lack perspective. And there's no reason to "decide" between one and the other as to which is "best"—except, of course, that people shopping for a camera wouldn't buy both; as a practical matter, they'd choose. But either one is a tremendously flexible tool for picture-making.
Either one sees better than I do. Albeit with somewhat different emphases.
Still, the two cameras take a markedly different approach to image quality. The D3 is biased toward getting the picture no matter what; the A900 is more biased toward getting the best picture.
More or less the same area at more or less the same size: Nikon (top) at 200%, Sony (bottom) at 100%. Again, not a terribly well-done comparison, I admit, but I think you can sense the basis for my impression that the superior resolution of the Sony just does translate into better potential image quality ultimately.
I qualify the things I say too much, I know, but there are a lot of caveats you have to make when talking about digital image quality. Note that all of these distinctions I'm making are really just leanings, or tendencies. To say the D3 is faster doesn't mean the A900 isn't fast—it is—and to say the IQ of the D3 is a bit less in some cases than the A900's doesn't mean it isn't still superb; it is, it is. Either of these cameras presupposes a good deal of expertise on the part of its prospective owners, and, ideally, would demand significant amounts of time and effort on the part of the owner learning to get the most out of the camera and its files. Either one would lend itself to a long process of learning and optimization, and I think either one would, in the end, prove deeply rewarding.
If I were a pro, especially shooting mainly for reproduction, I'd get the D3. It's faster, more competent, better suited to getting the job done. The image quality is superb, and the sensor a wonder, especially the way it sees in the dark.
Still, I have to admit that I'm tremendously impressed by the image quality potential of the Sony A900. I feel like I did little more in the short week I had it in my grasp than begin to scratch the surface of what it can do. I've only made a few prints, but to say they're wonderful is an understatement. I'm left with the sense (again, a holdover from my own particular personal history with cameras, which includes lots of familiarity with MF cameras) that it's a bit more like a medium-format film camera—a bit slower, more deliberate, maybe a bit more finicky; but it's got that same "deep" enlargeability, that same feel of resolution to excess, the same pleasure in finding the"grain" deep down in the image morphology. A similar subtlety and richness.
If I were an art photographer, especially shooting for fine printmaking, I'd get the A900. Its image quality is superlative, yet it's still reasonably portable, fun and comfortable to shoot with, and easy to use.
Of course I am an art photographer, not a pro. If some munificent wizard appeared out of the mists holding out these two cameras one in each hand and asked me to take the one I wanted, I'd pick the A900. That's just me. Your mileage might vary. In any case I'm not buying either one any time soon, so rest assured none of the foregoing has anything to do with personal purchase decisions or any of the emotional baggage and status-claiming behavior that sometimes goes along with that.
I'll probably have more to say about the D3 anon, since I'm still shooting with it. Then, next up, the D700.
Mike (Thanks to John Camp)
*Doubtless, on some forum somewhere, some idiot will write, "Mike Johnston says he knows everything about how a lens performs by looking at online samples! What horseshit!" or words to that effect. It's inevitable. For the record, there are just a few things I can figure out from online samples...sometimes, if the samples show what I need them to.
**You might remember a certain dirty joke about a young bull and an old bull on a hillside catching sight of a herd of cows. In case you don't, no, I'm not going to tell it, so don't ask.
***Remember "satire alert"?