Are there meaningful differences in Cameras and Lenses?
Original: 5/18/2010 on front page
Is There a Bad Camera?
May 18 (commentary). Right now there seems to be a lull in "which camera should I buy" questions (probably because there's a bit of a lull in major new cameras, the Sony NEX series notwithstanding), but the question still gets asked with regularity. Those that email me are usually surprised by my answer.
For several years now at the consumer level I've been writing "if you can't get good prints at the maximum size of any desktop inkjet (13x19") from any current DSLR, it's not the camera that's the problem." DSLRs got over a major hump sometime around the D70, and while they've gotten better with each new generation since, the differences between them are now mostly nuances rather than big issues.
In the pro ranks, there are some differences. The 5DII rocks on video. The D3s blows away the competition in low light. The D3x is the high pixel count winner both in actual number and in image quality. But pros pretty much know what they need and how to check whether a product has it or not. I'm not concerned about them. Moreover, even though a D3s has seriously mind-boggling low light performance, the 1DIV isn't exactly a slouch in low light. Nor is the 1DsIII a terrible performer at high resolution, either. It's just that there are some clear winners at any given time in the pro ranks, and if a pro absolutely needs an edge at something, they're likely to grab it and not worry about brand loyalty.
But let's get back to the headline question: are there any bad DSLRs these days? No. None. Even though I'd say that Sony botched the Alpha series and has produced an entire lineup of cameras where none top their competitors in image quality, that still doesn't make even the lowliest Sony A2xx model incompetent and incapable of taking seriously good photos. And even though I'd say that the Sigma SD14 seems a bit antiquated in features and performance, it too, can take perfectly decent photos.
What does differ between the brands is how you get the best results out of each of them. If you're shooting JPEG, almost none of the camera makers give you a "neutral" rendering at the default settings. Saturation and gamma tend to be boosted. That may be what you want, but the thing that really differentiates the cameras for JPEG shooters is in how easy is it to get the look you desire and achieve optimal results? I happen to like the Olympus JPEG defaults, but not the Nikon or Canon ones, for example. So the queston quickly becomes one of user interface. How easy is it to learn to dial in what I want?
Raw shooters have less worries in most respects, though not a single camera maker has given raw shooters the tools they need to evaluate whether we're getting optimal captures or not. Raw shooters are used to doing their tweaking during conversion, so the camera image quality controls start to be a bit less interesting.
Still, mark my words: "if you can't get good prints at the maximum size of any desktop inkjet printer, it's not the camera that's the problem." Thus, for most shooters who don't need that last drop of image quality, it's the camera's user interface experience that is the big differentiator. If you feel comfortable manhandling the camera settings in an intense shooting session on one camera versus another, that's likely to be more important
Is There a Bad Lens?
May 18 (commentary). Okay, if the sensors in the cameras all rise above a certain bar (see above), how about lenses?
I'm tempted to say there's no such thing as a bad lens (assuming its quality control is up to snuff so that it performs up to its design level). But I can't quite say that.
I have been struck at how good many low-cost lenses are. The 18-55mm and 55-200mm Nikkor kit lenses, for instance, are quite good performers overall. The new 35mm f/1.8G DX is a modest-priced prime that performs well.
But lurking behind the scenes are lots of issues where one lens can indeed be different than another. Simply put, there are a lot of ways in which a lens can impede the image quality you can achieve. We've known this pretty much forever, even back in the pre-film days. The goal with recent advances such as hybrid aspherical elements and different glass formulations is to try to pull up some of those problem areas so that lenses perform mostly "above the bar." Still, optical paths are pretty complex, tricky elements to get tuned perfectly.
A good case in point is longitudinal chromatic aberration. I see a lot of lenses well corrected for lateral CA these days, but not the other. Since lateral is easier to remove in software than longitudinal, I'd say that the design priority needs to be reversed. Of coure, then someone with a camera or converter that doesn't remove lateral chromatic aberration will cry bloody murder because the likely levels would be very obvious.
But the two real issues that separate "good lens" from "bad lens" these days are usually these: wide open performance, and corner performance. Consider these two lenses' performance wide open:
The lens on the left has a high degree of "sharpness" from center to corner. The lens at the right already starts at a lower MTF value but slides as you progressively head to the corners. Of course, one is a high-priced exotic f/2.8 lens (left) and the other is a much more modest priced f/1.4 lens (right). And at f/2.8, the lens on the right does much, much better.
So is the lens on the left a "good lens" and the one on the right a "bad lens"? Yes. No. Maybe. It depends upon what you need from it. The lens on the right is clearly struggling a bit at f/1.4, but those numbers still tell me that you should be able to get quite usable results out of it (though the extreme corners on an FX body might not hold up well). But is "usable" the same as "good"? Probably not. Here's what I had to say about the lens on the right: "...doesn't knock this test out of the park, but it does well. The older version was notoriously soft, especially in the corners, and the new version is better, but still far from perfect." Here's what I'd write about the lens on the left: "...has never disappointed me in any way."
While these are very different lenses in focal length and use, would there be a visible difference if I printed something shot wide open with both of them on the maximum size a desktop inkjet could print? Yes. And there-in lay several truths. For instance, when I talk about "upgrading," I say you first (skills), support second, lens third, and body last. That difference in a print is one reason why. I can see problems with your shooting discipline easily on a 13x19" print. Fix those and I can see problems with your support, still pretty easily. Fix those, and I can now start seeing the difference between a "good lens" and a "bad lens" fairly easily, especially when the lens is used near its extremes. Fix those, and, well, as the above article says, I don't see a lot of difference between Camera A and Camera B.
|
