Sigma 17-70mm f/2.8-4 DC Macro OS HSM

SLRgear Review
April 5, 2010
by Andrew Alexander

The Sigma 17-70mm ƒ/2.8-4 DC Macro OS HSM was released at the end of 2009, updating the previously unstabilized ƒ/2.8-4.5 version of the same lens. The lens is designed to be used on APS-C sized digital sensor bodies only; users of full-frame cameras will encounter hard vignetting at all focal length settings.

The lens will produce an effective field of view of 27-112mm (Canon) or 26-105mm (Nikon and others). It isn't a ''constant'' lens, in that as you increase the focal length, both the maximum and minimum aperture sizes decrease. The following table reflects the change in aperture with focal length:

The Sigma 17-70mm ƒ/2.8-4 DC Macro OS HSM ships with a lens hood, takes 72mm filters and is currently available in Sigma, Canon, Nikon, Sony and Pentax mounts for approximately $450.

Sharpness
When used wide open (ƒ/2.8) at its widest aperture (17mm), our evaluation lens showed slightly de-centered results, but fairly good performance: a large central sweet spot of sharpness in the range of 1-1.5 blur units, tapering out to slightlly soft corners in the range of 2-3 blur units. Zooming out things get a little better, especially in the 23-35mm range, where the sharp sweet spot gets larger and there's little corner softness to speak of. At 50mm (we're at ƒ/4 as a maximum aperture at this point) we encounter de-centering again and some degradation in the results: our sweet spot has moved to the ''southeast'' of the frame, and while it's a fairly good area of sharpness, corner softness has increased, again showing 2-3 blur units. Performance is comparatively poor at 70mm: de-centering has moved the area of sharpness to the extreme lower right corner, and sharpness is generally around 2 blur units across the frame.

Stopping down improves sharpness for this lens, but it never produces tack-sharp results across the entire frame. Stopping down to ƒ/5.6 produces the ''sharpest'' images until 50mm and greater, where you need to stop down to ƒ/8 for the best performance. Fully stopped-down, the lens produces mediocre results: 3-4 blur units at ƒ/22, all the way to 5 blur units at ƒ/32 and 70mm.

Chromatic Aberration
CA is quite well-controlled in the Sigma 17-70mm ƒ/2.8-4 OS, showing up only slightly in the corners of the 17mm focal length setting, in areas of high contrast. There's no practical difference between aperture settings, but according to the numbers, you get slightly less CA when using the lens at wider aperture settings (ie., ƒ/2.8 instead of ƒ/22).

Shading (''Vignetting'')
Corner shading isn't a real problem for this lens, with slight shading visible only when the lens is used at its widest aperture settings. The most prominent light falloff is visible at the 17mm focal length setting (at ƒ/2.8), where the extreme corners will be a half-stop darker than the center. At other focal lengths the differential is around 1/3EV; stopping down to ƒ/4 or smaller, corner shading disappears below the quarter-stop differential.

Distortion
When used in the wide angle configuration, the lens provides uniform barrel (''bloat'') distortion up to around 21mm, with noteable distortion in the corners at 17mm (+0.75%). After 21mm, distortion across the frame remains consistently barrel-distorted, but at a moderately low level (around 0.2%, on average) and the extreme corner distortion turns into the pincushion (''squeeze'') style. The worst results are seen at 34mm and greater, where the corners show almost -0.5% pincushion distortion. Post-processing would be required to correct for these effects.

Autofocus Operation
The Sigma 17-70mm ƒ/2.8-4 OS employs Sigma's HSM technology, which allows for fast and quiet autofocusing. Indeed, the lens was able to focus between infinity, close-focus and back to infinity in around one second. Attached 72mm filters won't rotate during focus operations, and there is no lens extension either.

Macro
Sigma has qualified the 17-70mm ƒ/2.8-4 OS as a macro lens, though it can only achieve 0.37x magnification (1:2.7 reproduction) at a minimum focusing distance of 22cm (just under 9 inches).

Build Quality and Handling
The Sigma 17-70mm ƒ/2.8-4 OS uses largely plastic construction, however at 535 grams (18.9 ounces), the lens is quite a handfull: used with an APS-C sensor body, the camera/lens combination becomes slightly front-heavy. The lens is finished with an all-black matte texture that is very tactile and easy to grip. The lens mount is metal; the filter threads are plastic. There is a distance scale etched onto the focus ring, marked in feet and meters, and there are macro reproduction ratios marked on the lens barrel, visible when the lens is zoomed out (and the lens extends).

The lens has a few control switches. One switch enables or disables autofocus; the second enables or disables optical stabilization. There's also a zoom lock switch to keep the lens at 17mm, though in our casual testing we didn't note any zoom creep that would warrant the switch (though perhaps over time, the zoom lock switch could be useful to have).

The zoom ring is rubber with raised ribs, approximately 3/4 inches wide. The lens extends as it is zoomed in towards 70mm, growing 1 5/8 inches in length. The ring is smooth to turn, nicely cammed, going from 17mm to 70mm with a 75-degree turn of the ring.

The focus ring is also rubber with raised ribs, about 3/8 inches wide. The ring only provides about 45 degrees of turning action, making it slightly challenging for manual focus operations, but providing for faster autofocus speed. The focus range ends in hard stops on either end.

The lens hood is petal-shaped and can be reversed onto the lens for storage. When attached, the lens hood adds 1 3/4 inches to the overall length of the lens.

We didn't perform extensive testing of the optical stabilization system for this lens, but we did find that its implementation made it difficult for us to conduct our lens testing. It's noteworthy that the optical stabilization system is always active in the optical design of the lens, even if it's turned off: we noted that even in manual focus, with the OS system set to off, the OS system moves its elements into position when the meter is activated, and it doesn't move them into the same place every time. Thus, while it probably won't have an impact on practical photography, if you're a macro shooter, you may find large inconsistencies in manual focus performance.

The actual effectiveness of the OS system, in contrast, is fairly good, providing (anecdotally) 2-3 stops of hand-holding stability.

Alternatives

Sigma 17-70mm ƒ/2.8-4.5 DC Macro ~$380
While the older version of Sigma's 17-70mm doesn't have the same aperture capability of the new version or optical stabilization, it produced markedly sharper images - though it could be that there are sharper copies of the 17-70mm OS we tested. CA resistance is a little better, distortion is a bit lower, but corner shading is a bit higher. Macro performance is also a bit better and the whole package weighs a bit less; it's also a little less expensive.

Canon EF-S 15-85mm ƒ/3.5-5.6 IS USM ~$720
There are actually several options to choose from in the Canon line (such as the 17-85mm ƒ/4-5.6): in this case, the Canon 15-85mm is definitely sharper, though it doesn't have quite the same aperture capability (ƒ/5.6 at 70mm versus Sigma's ƒ/4) though it could be said that this enforced stopping-down is what keeps the Canon sharper. CA resistance and distortion are about the same, though corner shading is slightly more prominent in the Canon.

Nikon 16-85mm ƒ/3.5-5.6G ED VR DX AF-S ~$630
Similar to Canon, there are several options to choose from in the Nikon line (such as the 18-105mm ƒ/3.5-5.6 VR): in this case, the Nikon 16-85mm is sharper than the Sigma, though (again) it doesn't have quite the same aperture capability. CA resistance is much better in the Nikon. Distortion is about the same, though corner shading is slightly more prominent in the Nikon.

Pentax 17-70mm ƒ/4 AL IF SDM SMC DA ~$480
Pentax bodies are image stabilized, so the lack of image stabilization isn't noteworthy here. The lens uses a constant ƒ/4 aperture compared to Sigma's variable aperture; it's much sharper used wide open than the Sigma. Chromatic aberration is somewhat problematic at the wide and tele ends of the focal range, much more so than the Sigma. Corner shading is less pronounced on the Pentax, and distortion is similarly complex.

Sony 16-80mm ƒ/3.5-4.5 DT Carl Zeiss Vario-Sonnar T* ~$750
There are a few Sony options in this category, with OS not being an important consideration as Sony bodies are image stabilized. The Sony is generally sharper than the Sigma, though CA is more noticeable on the Sony at 16mm than the Sigma at 17mm. Corner shading is about the same on the Sony, perhaps a bit more prominent at 16mm, and distortion is similar between the two lenses - perhaps a bit more forgiving on the Sony.

Conclusion
Given the price point and intended market, the Sigma 17-70mm ƒ/2.8-4 OS performed quite well. Good performance for sharpness, excellent resistance to chromatic aberration, low corner shading - the lens has test results which commend it. Many users may be tempted in by the ''faster'' aperture performance offered by the lens - ƒ/2.8-4 versus ƒ/3.5-5.6 of most kit lenses - though at 50mm and longer, they may find that they must stop down to achieve the sharpest images.

Sample Photos

The VFA target should give you a good idea of sharpness in the center and corners, as well as some idea of the extent of barrel or pincushion distortion and chromatic aberration, while the Still Life subject may help in judging contrast and color. We shoot both images using the default JPEG settings and manual white balance of our test bodies, so the images should be quite consistent from lens to lens.

Still Life shot

VFA target

As appropriate, we shoot these with both full-frame and sub-frame bodies, at a range of focal lengths, and at both maximum aperture and ƒ/8. For the ''VFA'' target (the viewfinder accuracy target from Imaging Resource), we also provide sample crops from the center and upper-left corner of each shot, so you can quickly get a sense of relative sharpness, without having to download and inspect the full-res images. To avoid space limitations with the layout of our review pages, indexes to the test shots launch in separate windows.