Sigma 17-50mm f2.8 OS EX on SD1 Merrill - Good and Bad

Recently I got an SD1 Merrill with a 17-50mm f2.8 OS EX lens. I love them both, but I have a love-hate relationship with the 17-50 lens. Here's why:

50mm set to auto-focus (back-focuses in the middle, but not at the edges)

Settings from processing above image in SPP 6.4.0

35mm auto-focused (same processing as 50mm shot)

Here's a shot made with infinity focus, done manually:

35mm manually focused at the infinity mark

I know I can fix the focus problem in the middle of the shot, using the SD1's micro-focus adjustment feature, but that would only solve one problem. It wouldn't solve the problem which is inherent in the lens, which is also an advantage, which gives the lens a special ability . . . the ability to get more in focus when shooting inside buildings, like cathedrals, or even inside a function room. A flat field of focus in such a situation isn't exactly ideal, so this lens is actually designed right for certain situations, but wrong for others.

What have I learned? You have to be careful when testing and viewing tests by others. This lens is incredibly sharp at the corners, but I didn't think so at first, because I didn't realize that it has such a semi-spherical (bent) plane of focus (not really a plane at all, I guess).

Here is a shot of the same scene made at 17mm:

17mm auto-focused

All these were shot at f6.3 (to avoid diffraction as much as possible) and exported to jpegs at level 10 compression, using the same processing settings in SPP 6.4.0 (version 6.5.0 won't work on my MacBook Air, because I have an old version of OSX on my computer) for every shot. The idea was to eliminate differences in processing, of course. Since the lens is f2.8 throughout its range, I thought that f6.3 was a reasonable aperture to use for the whole range, though I could have opened up the aperture at little at wider angles. Still, I: was surprised to find that even f6.3 didn't seem to help much at 17mm, as you can tell by the corners of this photo immediately above this paragraph.

I was testing this lens against my 18-50mm lenses (I have two of them). Here's a shot from one of those at 18mm:

18mm with 18-50. Surprisingly good in the middle, for such a cheap lens, huh?

How about that color cast?!? Yes, the camera settings AND processing settings AND sunlight were identical from one lens to another.

One thing I was really surprised at was how good the middle (see the entryway at the bottom middle of the photo) of the photos from the cheap 18-50mm lens turned out to be. That lens produces a very detailed image in the center, as far as I'm concerned. Could it be better? Probably not . . . for the price. Distortion was pretty minimal too, don't you think?

Overall, I like not only the detail from the 17-50 better, but the color too. Now I see where the SD14 might have gotten some of the "green cast" we see in the photos from those cameras.