photozone tamron 17-50 test posted

http://www.photozone.de/8Reviews/lenses/tamron_1750_28/index.htm

By the Numbers

Except for in the center wide open, the Tamron has higher resolution everywhere than the Canon EF-S 17-55mm f/2.8 IS. At the center, both max out the 350D's resolution, but the Tamron's extreme corners are very nearly as sharp as the center, and should max out my 300D at most apertures and focal lengths. [Edit: "extreme corners" in PhotoZone tests are actually more like 80% to the corner - multiple DPReview forum members have reported that the very corners of the frame (i.e., 99% to the corner) are very soft.] Like the Canon 17-55, the Tamron is better than the 24mm f/1.4L and 35mm f/1.4L, but unlike the Canon 17-55, it is also a match for the 50mm f/1.4. Astoundingly impressive.

Vignetting and distortion are both slightly better than the Canon 17-55, except for at 17mm, where the distortion is slightly worse.

Chromatic aberration is significantly worse than the Canon 17-55, although it gets better when stopping down, unlike the Canon. At 17mm, the Tamron is much worse than the Canon and never catches up no matter how far you stop down. At 35 and 50mm, the Tamron is better than the Canon when stopped down.

Real World Shots

Unfortunately, Klaus took different subjects with each lens, but there are some comparisons that can be made with varying levels of confidence.

First, both lenses have bad bokeh. Both have bright ring highlights: the Canon seems to have a thinner bright ring, but the overall nervous effect is the same. Even worse is the astigmatism which both seem to share, turning the background to petroleum ooze.

Next, CA. The building shots have noticeable CA on the dark trees in the background. I can't see any CA at all in the Canon 17-55 shots. Klaus might be reducing the CA in RawShooter, but as long as he's doing the same post-processing on both sets of images, the Tamron is indeed much worse.

Then, contrast. Seems pretty good, although it's hard to tell without comparison shots of the same subject under the same lighting conditions.

Finally, sharpness. There seems to be something wrong with the sharpness of the Tamron. At 100% magnification, it is not as crisp as the Canon. The numbers say it should match or exceed the best primes in resolution, but I'm just not seeing it (I can see it on the Canon shots). Perhaps this is due to the strong CA. At any rate, it's not significantly better than my kit lens.

Other Considerations

Klaus mentions slight focus problems: "a little unsure indoors". This matches my experience with the Tamron 28-75mm f/2.8, which the new lens is supposed to be the "digital" equivalent of.

He also mentions field curvature bad enough to show in real world shots. Indeed, you can see the soft bottom corners and left side of the building in his 17mm f/2.8 architectural shot. My impression is that none of the Canon 17-55 shots yet posted have shown corners this bad, but it's a tough call without comparable compositions to judge.

Conclusion

In the end, this review was not overly helpful. The corner resolution numbers are phenomenal, but, whether because of CA, field curvature, or just plain experimental error, they don't appear to be backed up by the sample shots. If it is experimental error, then I trust the actual photos over the measurements. In that case, based on this review, the Tamron gets the thumbs down.