Comparing Cameras
I’m still trying to make up my mind as to which DSLR to buy. I long ago decided on Canon, and the Rebel’s a little small and a little too low on the megapixels for my tastes. In my price range, that leaves the 50D and the 5D Mark II (or equivalent older models, but I like newer toys. :-) ). The primary difference between these two are:
- The 50D is an APS-C sensor with a 1.6x Field-of-view crop. The 5D is a full 35mm camera.
- The 5D has 21 megapixels vs. 15 megapixels on the 5D.
I like to take pictures of two things: birds and bugs. (For family, friends, and pets I can use a point-and-shoot.) Keeping that in mind what makes sense?
The smaller sensor on the 50D means that the image is cropped and you get a closer picture. (These are really two sides of the same coin.) Whether this is a good thing or a bad thing depends on what you’re shooting. If you like wide-angle landscape shots, you hate the cropping and buy the 5D, but I take pictures of birds and bugs. For birds, you want the crop, but only if the megapixels are the same size. That is, a 15MP APS-C camera will outperform a 15MP full-frame camera. However the megapixels aren’t the same. What does that mean?
In essence, set up side by side with the same lens, and pointing at the same bird, the full frame 5D will pull in more background than the APS-C 50D because it has a wider field of view. That’s great if you don’t lose megapixels on the bird. Do you? Let’s do a little math.
Let’s assume we’re looking at a Sanderling far enough away that it fully fills the frame on the APS-C sensor. At maximum resolution, the Sanderling will be 4752 pixels wide and 3168 pixels high (honestly less because sanderlings are wider than tall, but assume a rectangular bird.) By contrast on the 5D we’ll have 5616 x 3744 pixels, but with the same lens not all of those will be focused on the bird. How many bird pixels do we lose? If we end up with 4752 by 3168 bird pixels or more, then the 5D makes sense. But if we end up with fewer than 15.1 mega-bird pixels, then I should prefer the 50D. (Of course I could just screw a 2X teleconverter onto the 50D, but where would the fun be in that?)
The 5D and the 50D will both place the bird on a sensor of 22.3mm * 14.9mm. That’s 100% of the 50D’s sensor, but only 22.3mm * 14.9mm / (36mm*24mm) == 38.46% of the 5D’s. So where the 50D is using 15.1 megapixels for the shot, the 5D is using 38.46% of 21 megapixels. Thus side by side with the same lens at the same distance the 5D only gives me about 8.1 megapixels, barely more than half of the 50D’s. Thus to take pictures of birds I should definitely prefer the 50D.
Check my math. Does that make sense?
Anyway, this is all for the birds. For the bugs, especially ones you can get close to, I think the full frame just works better. In that scenario, (1:1 macro) it’s a straight calculation that larger pixels are better and larger sensor sizes are better. But wait a minute: the 5D has a much larger sensor (256% larger) but only 35% more pixels. What does that imply about pixel size?
Another quick calculation and we find that on the 5D we have 5616 * 3744 pixels/(36mm * 24mm) = 24,336 pixels/ mm2. Assuming square pixels, each pixel is 1/24336 == .000041 mm2. On the 50D we have 4752 * 3168 / (22.3mm * 14.9mm) == 45,307 pixels/ mm2. Thus each pixel is approximately .000022 mm2. Larger pixels are better (less noise) so indeed it does look like the 5D handily beats the 50D by this measure.
Also, since a macro lens is 1:1, the 5D can full frame a bug up to about 4cm long, while the 50D can only go to about 3cm. That matters for grasshoppers and such, but for the smaller bugs, it’s not so relevant.
Quick summary: shoot birds with a 50D and bugs with a 5D.
Update: According to dpreview, “Apparently the 50D’s sensor has what are effectively gapless micro lenses, which significantly increases the efficiency of each pixel.” So possibly the quality difference between the 5D and the 50D isn’t as large as the sensor sizes might suggest. Also see this forum post.
October 13th, 2008 at 6:57 AM
Assume a spherical cow… Sorry, couldn’t resist. If you don’t get the joke, google is your friend.
Anyway, do you have a good camera store near you? One that would let you actually try the cameras out, in-store, with the lenses you want to buy (and your own memory card)? Bring in a picture of a bird, tape it to a far wall, and go to town. Try that 2x adapter out. Try the macro lens. Bring the images home and look at them. Remember that the odds are that the lens will make an ENORMOUS amount of difference in the final image, so don’t stint on the cash there.
I wind up doing a fair amount of photography in fairly low-light situations, so the bigger sensor with less noise is a clear win for me. Your mileage more than varies.
October 13th, 2008 at 9:11 AM
My suggestion is to spend a little more time investigating the lenses you want to buy and which camera bodies work (and work well) with those lenses. Camera companies are constantly upgrading the bodies, but the lenses are pretty stable and will last far longer than the body.
You like newer toys, and I don’t blame you (I’m the same way). In a year or two you’ll see another new camera from Canon that you’ll like much more than the one you buy now. I’m not suggesting you’ll run out and buy it, but you need to keep in mind that you’re most likely going to upgrade over time. After only 2 years of owning a Nikon D80, I want a shiny new D700 or D3.
October 13th, 2008 at 11:25 AM
Well, I’d suggest a Nikon D300 for birding, but if you have to go Canon, you’re probably best off with the 50D (the 40D would be fine, too, but I buy your logic). That, or a secondhand 1D variant. The reason I bring that up is it seems for birding good AF goes a long way, and almost any 1 series’ AF (even the old ones) will blow the doors off of the 9 point AF in the 50D. I’d take an in focus, lower res shot any day over a misfocused high res shot, but that’s just me :)
October 13th, 2008 at 11:40 AM
I actually do have a pretty good idea of the lenses I want. That’s why I’m specifying Canon. Their high end telephoto and macro lenses seem to be a step or two beyond what I’ve seen for Nikon.
And yes, I am familiar with spherical cows. :-)
May 12th, 2009 at 7:16 PM
Elliotte- It’s not correct to use the same lens to compare different sized sensors abilities’ to render an image. It may be practical fiscally or weight wise in the field, but for the bigger sensor, you really should be using a lens with a longer focal length- one that matches the increase in size of the sensor. Otherwise you’re comparing apples to oranges. For the cameras you list, a lens 1.5x should do the trick.