Although Canon's 17-55 does not fare too well in tests it may be better than the tests indicate, as appears to be the case with their 17-85. This happens a lot with audio equipment where the tests simply do not reveal many of the important characteristics of the equipment. Not that the tests ar poorly done or unimportant, just that there is more to a lens' performance than test charts
<In spite of my posting so many photozone charts - I do agree with Bob that the tests do not tell us EVERYTHING. They merely suggest an opinion. Several tests and sources - plus the comments accumulated at Fred Miranda and other similar reader opinion only sources - combined with PERSONAL OBSERVATION and HANDLING by respected associates, friends and YOURSELF in a local store should form a balanced initial impression . That was why I stated earlier that ...."I am not a good source for an opinion with NO real hands on experience with these lenses. ".... and that these were just results from ONLY one site ... photo zone>
It may be that Canon does not consider vignetting to be a big issue anymore since it is fairly easily corrected in post processing (btw, I do not agree with that idea).
Even with a possible 25% markup just for the "Canon" logo there is still a large price difference between the two lenses, and it seems unlikely that Canon put fancy glass elements in the lens just for bragging rights.
I am not a fan of third party lenses or much of any other third party equipment. I work for a major manufacturer and have examined many third party replacement parts for our equipment. Most are junk, few are very good, almost none meet all of the original specifications since we, like Canon,
do not share those specifications with outside sources.
<I susected this would be obvious. When my Tamaron was NOT functioning / focusing properly I thought that this must be a design flaw due to lack of full operational interface data! Reverse engineering is NOT expected to be a perfect approach to duplicating products in this arena!>
I have some experience with IS and consider it worthwhile. In my limited and completely unscientific tests it is good for at least two stops on my 24-105, at all focal lengths.
<I agree that 2 stops is conservative and reasonable. Nikon's claims of up to 4 stops of useful correction seem... excessive. And I used to be considered a NIKON SNOB about 15-20 years ago!>
My opinion is if you have the $$$ go with the Canon lens. If not, go for the less expensive option (which I did with a Zenitar 16, and was pleasantly surprised with how good it is, optically at least).
<I'm gonna look into a ZENITAR NOW... see what you've started Bob?
>
boB