At the edges there's no comparison with my Sigma 15-30 at any largish aperture. The Sigma is designed for FF 35mm. Comparing say 1/4 of the frame next to the short edge at f3.5, the difference between the two lenses is so great it's just ridiculous, yet in the centre the differences seem quite marginal with the EF-S lens being perhaps slightly sharper.
[a href=\"index.php?act=findpost&pid=172381\"][{POST_SNAPBACK}][/a]
Just for the record, I'll have to correct that statement. I was apparently comparing the 17-55 at full aperture (f2.
with the 24-105 at full aperture (f4) and 24mm, and depth of focus issues might have come into play.
Having spent more time checking my test images, the edge performance at 17mm and f2.8 compares very favourably with the Sigma at 17mm and f3.5 (its maximum aperture). In fact despite the Sigma being designed for FF 35mm this EF-S lens at f3.5 has the edge (at the edges), but at f2.8 it's apparently not as sharp as the Sigma at f3.5.
This creates a dilemma. The main attractions of the EF-S 17-55/2.8 IS are its wide maximum aperture and IS. However, if the lens is noticeably softer at that maximum aperture, one might prefer to stop down at least half a stop, thus removing one major advantage compared with a lens like the Sigma 15-30. There are other advantages though; less bulk, weight and a longer reach.
I think I might have to do further tests.