Let's review some outdoor image quality test results obtained during evaluation of a handful of 17-somthing mm lenses. To comparison test outdoors, a clear day is the first requirement. Any clouds in the sky, even if they appear to be even, can change the lighting between shots and cause one lens to appear better or worse than it is.
The first comparison test results I am sharing are from the absolute top left corner of the frame out of a Canon EOS 60D using the Standard Picture Style and a sharpness setting of "1" (very low). Lenses were set to approximately 28mm. The same tripod-mounted camera position was used, but this does not mean identical corner framing due to lens focal length marking discrepancies and my guess at the focal length setting on lenses not having 28mm marked.
Specific image content aside, I think that you will be able to discern a reasonable difference between the lenses in this example. These 28mm results are generally indicative of the performance of these lenses over their entire focal length range. Because of the large number of results, a mouse click is required on the aperture labels.
The Canon EF-S 17-55mm f/2.8 IS USM Lens indeed turned in better corner performance at a wide open aperture than stopped down one or perhaps even two stops in this test. This of course means the other lenses are better performers at f/4 and f/5.6 with the Sigma 17-70 being a possible exception.
The Sigma 17-70 would better fit in a difference class of lenses due to its narrower aperture and longer focal length range, but including it here shows me what it can do as I prepare to write the Sigma 17-70mm f/2.8-4 DC Macro OS Lens review.
With light being least redirected in the center of their image circles, most lenses perform best in the center of the frame. As I said earlier in this review, the Canon EF-S 17-55mm f/2.8 IS USM Lens turns in very impressive center-of-the-frame performance. Even at its center image quality weakest focal length, 50mm, the Sigma performs on par with the Canon 17-55 (which is set to 55mm in the following comparison).
The above comparison images were shots with the Canon EOS Rebel T3i/600D using the Standard Picture Style and a sharpness setting of "1" (very low). The Sigma 17-50 shows a more-clear advantage in the center of the frame at the lower end of its focal length range.
One difference I've noticed in the various comparison tests I've shot is that the Canon EF-S 17-55mm f/2.8 IS USM Lens consistently produces warmer image colors than the Canon 17-55. I've actually preferred the warmer image in the majority of comparisons I've noticed this difference in.
From a flare standpoint, the Sigma 17-50 OS compares favorably to the rest of these lenses. Perhaps slightly better overall than the Canon 17-55 and the Tamron 17-50 non-VC. The Tamron 17-50 VC is the clear loser in the flare competition.
The Sigma 17-50 OS and the Tamron 17-50 non-VC lenses exhibit the least vignetting with the Canon 17-55 and the Tamron 17-50 VC showing the most.
The Canon EF-S 17-55 has less barrel distortion at 17mm than the other lenses, but more pincushion distortion at longer focal lengths. The other lenses are similar in their distortion amounts. I didn't see any significant differentiators from a CA standpoint - these lenses are all similar in this regard.
The Canon 17-55 wins the AF performance category - this is a big advantage in my opinion. The Canon 17-55, however, is the loser in the price comparison. It costs far more than the Sigma 17-50. The Sigma 17-50 is slightly more expensive than the Tamron 17-50 VC, but the better-optically-performing Tamron 17-50 non-VC is considerably less expensive. It, of course, lacks image stabilization - another big advantage in my opinion.
Back to the Canon EF-S 17-55mm f/2.8 IS USM Lens Review
Back to the Sigma 17-50mm f/2.8 EX DC OS HSM Lens Review
Back to the Tamron 17-50mm f/2.8 XR Di II VC Lens Review
Back to the Tamron 17-50mm f/2.8 XR Di II Lens Review
Back to the Sigma 17-70mm f/2.8-4 DC Macro OS Lens Review
Bringing you this site is my full time job (typically 60-80 hours per week). Thus, I depend solely on the commissions received from you using the links on this site to make any purchase. I am grateful for your support! - Bryan