tokina 11-16mm f/2.8

First item on the agenda is my XSi shipping notice — expected arrival is April 18, just in time for my birthday.

I also now know what I'm going to buy with one of my Sandisk-Adorama rebates (the first is still being processed): a Tokina AT-X 116 11-16mm f/2.8.

Canon EF-S 10-22mm f/3.5-4.5
@ 10mm f/16

With the June introduction of the Tokina 11-16mm f/2.8, there will be a half dozen ultrawide rectilinear lenses available for Canon APS-C camera owners:

• ($689) Canon 10-22mm f/3.5-4.5
• ($479) Sigma 10-20mm f/4-5.6
• ($659) Sigma 12-24mm f/4.5-5.6
• ($569) Tamron 11-18mm f/4.5-5.6
• ($499) Tokina 12-24mm f/4
• ($570) Tokina 11-16mm f/2.8.

(current prices at Adorama)

In addition, Tamron is working on a new lens which may or may not see the light of day, and if so, who knows when:

• Tamron 10-24mm f/3.5-4.5

Based on price, these are all in competition with each other — only $210 separates the most and least expensive. Looking closer, though, there are three price tiers: high $400s, high $500s, and high $600s.

Grouping by maximum aperture is a little fuzzier, but we can try for slow (Sigma 10-20mm f/4-5.6, Sigma 12-24mm f/4.5-5.6, and Tamron 11-18mm f/4.5-5.6), medium (Canon 10-22mm f/3.5-4.5, Tokina 12-24mm f/4, and Tamron 10-24mm f/3.5-4.5), and fast (Tokina 11-16mm f/2.8).

Cross-referencing these two criteria (price and max. aperture), it seems at first glance that one does not always get what one pays for. For instance, the Sigma 12-24mm f/4.5-5.6 is tied for slowest lens, yet is in the highest price tier. Conversely, the Tokina 11-16mm f/2.8 stands alone in the fast max. aperture bracket, yet has only mid-tier pricing.

The Sigma's abnormally high pricing is easily explained: it's the only lens of the group with full-frame coverage. Combine that with the lowest linear distortion of the bunch, and we see that the premium price is paying for a specialist lens. The only full-frame lenses close to that coverage cost around three times as much.

The Tokina f/2.8 lens, though, is harder to explain. Tokina says the optical design is based on their 12-24mm f/4. For a mere $70 more, and the loss of the long end of the zoom range, you gain a full stop and a little extra range at the wide end, where it really counts.

The Canon is significantly slower, and, if you include the lens hood, $153 more expensive. To its credit, the Canon is much lighter, has excellent flare resistance and color, unusually low linear distortion, the widest focal length, the largest zoom range (so far), Ring USM focus, and is fully supported by Digital Photo Professional 3.3's automatic lens aberration correction. And, of course, as a original manufacturer lens, one can rest assured that it will be compatible with all Canon cameras until the mount changes (and it hasn't in over two decades).

Canon EF-S 10-22mm f/3.5-4.5
@ 10mm f/5

Until someone compares the two side-by-side, their relative optical performance is unknown. Judging from early shots from the Tokina and my previous experience with the Canon, I can make some guesses though. Lateral chromatic aberration is going to be an issue for both at their widest focal length. Purple fringing is likely to be a bigger problem for the Tokina, with its larger maximum aperture. Vignetting is also a problem on the Canon, and despite the Tokina f/4 lens being a fairly good performer here, the extra stop is probably going to cause the new lens to have vignetting issues as well. Barrel distortion at the wide end should be a win for Canon, although the Tokina is not bad here. Although I have yet to see any clear flare tests for the Tokina, I suspect Canon will win that one as well, as it is very good in that department. Color should be close — the Canon has saturated colors with a neutral cast, but so does my Tokina 10-17mm, which is another recent design.

Finally, there's sharpness. I consider the corners of the Canon at 10mm just plain mediocre. Sure there are probably worse ultrawides out there, but compared to a good standard zoom or even just the center of the frame, they are not sharp. But then the Tokina is showing some not-so-great corners too. The difference is it looks like the Tokina corners improve greatly with smaller apertures, whereas the Canon remains sub-optimal at all apertures.

Canon EF-S 10-22mm f/3.5-4.5
@ 10mm f/16

More important than corner sharpness is center sharpness. Here, both seem to have good resolution — not as high as a prime, but good enough for most purposes. Contrast and micro-contrast are also good for a zoom, but the Canon at least doesn't bring a scene to life like a good prime does. It's hard to give a verdict on the Tokina since so far all shots I've seen have come from Nikons, which have noticeably different color responses and unknown contrast curves compared to Canon. Let's just say what I've seen so far is, if not quite amazing, at least promising.

There was a time when conventional wisdom said ultrawide lenses fundamentally couldn't have the edge-to-edge sharpness that longer lenses have at much lower prices. The Nikon 14-24mm f/2.8 debunks that myth — and it's wider than any zoom except the Sigma 12-24mm. But it does so at a price compared to the Tokina 11-16: it's twice the volume, weighs almost twice as much, costs almost three times as much, and is manual everything on a Canon. Canon's closest competitor, the 16-35mm f/2.8 II, combines the generous zoom range of the 10-22mm with the f/2.8 of the Tokina, but is almost as much as the Nikon, has significantly worse corner sharpness, takes expensive filters, and again only qualifies as a true ultrawide on a full-frame camera.

Which is all a long way of saying: if the Tokina even approaches the sharpness of a Canon 16-35mm f/2.8 II, it will be a revolution for the APS-C format, removing the last major reason to upgrade to full-frame for those on a budget.