Depth of field for different sensor sizes

40D - 24mm f/4
GF1 - 20mm f/4


There is a lot of talk about how people love micro 4/3s becuase of the ability to get more out of focus backgrounds (shallow/less depth of field). At the same time, the argument goes against micro 4/3s in favor of APS-C. Then again, in favor of full frame over APS-C. I've seen both side of the argument, and decided to do a quick comparison between my Canon 40D and Panasonic DMC-GF1. The problem with these kinds of tests:

1. Different focal lengths. Without going into great detail, you have to choose between using the same focal length, or same perspective/angle of view. I went with the later, choosing a focal length on my 40D that approximated the GF1's 20mm lens. I eyeballed the 40D and ended up using 24mm - about 38mm in 35mm equivalent field of view terms. The Panny had the 20mm x 2 for about 40mm. Pretty close.
2. Aperture. When it comes to compact point and shoots versus an SLR it is unfair - SLRs have lenses avaiable with much larger maximum apertures so you'll always be able to isolated better. In the GF1's case I happen to have a wider aperture there than most of my SLR lenses (85mm f/1.8 excepted). I chose an aperture of f/4 so I could compare with a similar field of view. It won't be as dramatic, but that is kind of the point - take a look at what I can get with my normally used lens on the 40D (24-105mm).
3. Foramt. The GF1 has an image sensor size ratio of 4:3, while the 40D's is 3:2. This creates some minor differences, but for this example I just tried to keep the subject the same size.
4. Resolution. The two cameras have slightly different resolutions - the 40D's 10MP compared to the GF1's 12. As you can see, they aren't that different, and really if you cropped the GF1 to 3:2 it would be close to 10MP.

My main question was how did the backgrounds compare - did the Panny have visibly more sharpness resulting in a more distracting background? Sort of. Sure the 40D has a little more blurring, but that is partly because the lens is wide open (circular aperture) versus the Panny's stopped down f/4. All in all I wouldn't really notice this difference in practical terms. Of course, all things are never equal - I can get MORE blur out of the Panny at that focal length becuase I don't even own a f/1.7 lens in that range. Or I could slip a 50 f/1.4 on the GF1 and walk a little further away. In my opinion, the difference over a compact is there, but against the slightly larger APS-C sensor lens and aperture available is more important the sensor differences. I could affect it more by moving position a bit and selecting focal length. With the 20mm f/1.7 Panny lens I can get enough separation for my taste...

I really don't think it is worth the fuss - you are much more hampered by the 4/3s lack of lens selection that a slight difference in blur. This is a knock against the Olympus system (by some), but who makes an f/2 standard zoom (Olympus!) versus everyone else's f/2.8...