The whole crop-sensor versus full frame sensor thing never quite made sense to me, ’til I saw the difference between a 50mm lens on a crop camera versus the exact same lens on a full frame sensor. If this whole thing already made perfect sense to you, feel free to skip this post. For those of you to whom the whole thing makes about as much sense as Finnegan’s Wake — in Swahili — read on.
Sensors come in several sizes, from the thumbnail-sized sensors in your average point-and-shoot to the 120mm sensors in medium-format cameras. If we’re taking a 35mm film frame as our point of reference (also the size of the sensor in a “full frame” camera), any sensor smaller than 35mm is going to have a crop/multiplier factor when used with 35mm lenses. You’ll recall that some time back we talked about the difference between digital and optical zoom, where digital zoom essentially crops the image captured by a sensor at its native resolution; the crop factor introduced by a smaller sensor does the same thing, minus the software trickery.
Here’s why it didn’t exactly make sense to me. Lenses have different fields of view at different focal lengths. A fisheye lens (say, 8mm) can give you a 180 degree angle of view. A 50mm lens, sometimes called a “normal” lens, closely approximates your natural field of vision. A much longer lens, like a 300mm lens, gives a much tighter field of view (around 8 degrees).* You’d think (or I thought, at least) that regardless of the size of the sensor, the photo would be the same because the lens’s field of view at a given focal length would be the same in any case, so a lens racked out to 300mm would have the same FOV whether you used it on a full-frame sensor, a crop sensor, or a point-and-shoot.
Only it isn’t quite. So how’s this crop thing work? By way of analogy, think of it like this. Let’s say you’ve got a slide projector that’s ten feet from a three foot wide screen. The image fills the screen with no problems. The projector is your lens; the image coming out of the projector is the “image circle”; the screen is your sensor. When you’re using a crop sensor, you’re not moving the projector relative to the screen; it’s simply changing the area covered by the projected image. So if you put a two foot wide screen in front of the projector, you’re going to notice that a much smaller part of the image is visible on the screen (with some of the image spilling over to the area beyond it). A DX lens has a smaller image circle (in essence, focusing the “projector’s” beam more tightly), so it’s going to fill a smaller “screen” (sensor) easily enough, but it’s going to come up short on anything larger.
This is also, incidentally, why a lens that exhibits light falloff or softness in the edges and/or corners on a FF camera generally looks better on a crop camera. Most lenses — at least once you stop them down a bit — are going to be reasonably sharp in the center. The part that’s sharp is the part that’s being projected onto the smaller sensor, whereas a larger sensor’s going to also incorporate the dodgy bits from the perimeter of the frame.
Therein lies a lesson. Some people — and I was one of them — purchase full frame lenses when they have a crop sensor camera just in case we decide at some point to jump to a full frame camera. When you’re reading reviews of lenses, therefore, one of the things to pay attention to is who’s using the lens as they review it. It’s not unheard of for a lens to get rave reviews from DX/crop users only for the FF people to point out flaws in the lens’s image quality. If you have no plans to switch formats, you may not have much to worry about (though other issues, like lens flare, coma and color fringing will typically manifest no matter what body you’re using). But if you’re going to be switching at some point, pay attention to those flaws. You may be willing to put up with them, but at the very least, go in with eyes open.
Oh, and about the images accompanying this post: the camera wasn’t moved relative to the bookcase (it was on a tripod). I’m also using the same lens (a 50mm 1.8) in both shots. The only change is that the first shot was taken in the camera’s full-frame mode, while the other uses its crop mode. So on full frame, the 50mm looks… well, like a 50mm looks. In crop mode, it acts more like a 75mm.
If I haven’t been as clear on this as I’ve tried to be, feel free to sound off with your questions (or better examples) in the comments section below. We’ll be revisiting this topic (albeit from a different angle) soon.**
*If reading this is making you trade your confusion over sensors for confusion over angles of view, there’s a very good explanation at the always-reliable Mansurovs Photography: http://photographylife.com/equivalent-focal-length-and-field-of-view
**The “different angle” is a post taking up whether a crop sensor or full frame is “better.” (click the link for my take)