Sigma's attempt at a fast DX standard lens
elements in 7 groups.
HSM lens focus motor, internal
focus, 62mm filter size. 8-blade aperture. Focuses to
About 2" (52.5mm) long, 15 ounces (530g).
Since posting the original review, I've gotten a number of "you must have a bad sample" emails from people. A few others challenged my testing. One of the reasons this review came quite a bit later than my two 35mm reviews was that I was indeed worried that I had a bad sample, so I quickly arranged to borrow another. But I got virtually the same results from it, so either Sigma's quality control is quite bad or my results were correct. You can form whatever opinion you'd like to about that.
As to the testing methodologies, I do not simply fill the frame with a chart, and I don't simply trust the phase detect AF to get things right. My assistant and I work hard to get proper chart alignment. The images I publish in reviews are without any post processing or sharpening, which is probably what is throwing many of you. You can't actually get real, comparable resolution numbers by running things through post. Let me illustrate. Here's the same image: what I posted below in the Results section is on the left; what you can obtain from that same image with careful post processing is on the right.
This illustrates something that a lot of people don't get about the Web and why I don't put a lot of numbers or images in my reviews: it's really easy to make things into something they're not. Everyone would say the right-hand lens is sharp and contrasty, but the left-hand lens is unsharp and has low contrast. But they're the same image from the same lens. However, consistent and correct testing reveals relative differences between lenses, and the only way I can even begin to show that on the Web is via unprocessed shots. As I replied to one person, there is a difference between a "good enough" lens and a "good" lens. The Sigma can be used in ways that make it "good enough," but it is not a "good" lens at f/1.4 in my testing. When I look at the raw numbers out of Imatest, it is down significantly in MTF in the central area compared to the Nikkor 35mm f/1.8G DX, and way down at the edges. It simply isn't as sharp a lens as the Nikkor wide open, and until someone can show me otherwise, I'll stand by that statement.
The 30mm focal range is close to "normal" for the DX format (some would say it's closer than 35mm is). You'll get some arguments over what the definition of "normal" should be, but I'd say that anything in the 28-35mm range for DX certainly falls into this category, and this lens is squarely in that range.
The first thing you notice with this lens is its bulk and weight. At about 2.5" long and 3" in diameter and weighing in at almost a pound, this is a lens that is relatively small, but feels dense and heavy for its size. The reason is there's a lot of glass in this lens compared to the other two 35mm lenses I reviewed (mostly) simultaneously. That f/1.4 aperture produces a lens that's 62mm in size for filters compared to 52mm for the others, and that pattern of "larger glass" repeats throughout the design.
The 30mm focal length provides 42° of horizontal angle on DX. One reason why some people think this is "normal" is that the most central area our eyes take in--anywhere from 40-60°--is where most of perception of the scene in front of us is. Areas outside that tend to be more "motion detectors" as far as our eyes are concerned (which is why we glance to the side or turn or head when something is not happening right in front of us). The nice thing about 42° is that it is easy to "guess": it's close enough to half of 90° that anyone ought to be able to walk into a scene and visualize how much the 30mm will see on their DX body.
The f/1.4 aperture is relatively fast. It's fast enough to give a three-stop+ advantage over most of the kit zooms (which would be at least f/5 at 35mm). The focus ring is a nice size, taking about half of the depth of the lens. The lens has a distance scale but no DOF or IR markings. There are no switches on the lens. The lens does not have VR or an aperture ring (this last limits the number of older film SLRs that the lens can be used on). The aperture blade is eight-sided.
Close focus is a about 16" (0.4m).
The lens cap on my sample is the old-style edge pinch. The lens uses 62mm filters, a common filter size for Nikon.
The lens is made in Japan.
With a lens this simple, handling isn't exactly a deep subject.
The focus ring is wide and reasonably smooth for an autofocus lens. The front element does not rotate during focus. The focus ring rotates about a quarter turn from near to far focus, which feels a bit constrained and makes it difficult for precision manual focusing. The slightly petaled hood is actually quite deep and substantive, making the lens stick out quite a bit more than the competitive lenses with their hoods.
The Sigma is a relatively heavy lens, despite its smallish size. It's slightly bigger and considerably heavier than the Nikkor 35mm f/1.8G, for instance. Thus, on the light bodies, like the D40x, D60, and D3000, the Sigma makes for a front-heavy camera. On the heavier D300 it balances fine, though. Build quality is quite good, almost high-end pro.
The Sigma is on the right, the Nikkor on the left (the 35mm f/2 Nikkor is in the middle). This is a single shot, so the scale is correct.
What do you expect from a very fast prime? Well, lots of light, but historically, poor wide open performance.
Autofocus: The HSM on this lens is fast, faster than many Nikkor primes with AF-S.
Sharpness: The weakness of this lens, and the reason I'm agnostic about a recommendation, is that it simply isn't up to Sigma's best standards in terms of resolution. The lens is quite soft wide open (abysmally so in the corners), and improves to about f/8 where it is its sharpest in the center (f/5.6 in the corners). My sample was ever so slightly sharper on the right side than the left, but not enough to judge it as misaligned.
f/1.4 center f/8 center
f/1.4 extreme corner f/8 extreme corner
The edges just never get anywhere near approaching the central sharpness. I'd judge the absolute corners to be soft at every aperture. The central area is very usable from from about f/2 to f/11, though. Indeed, at it's peak it is what I'd regard as excellent in the center. For landscapes and things where you need good edges, this isn't the lens. For indoor shooting of events where the corners may already be dark, this lens performs decently.
Note: my image samples come from a slightly different set of testing than many other sites and are presented with no processing and sharpening. With wide angle and even "normal" lenses you can't test them at distances where the "chart fills the frame" (which would often put you close to minimum focus distance). You need to test them at far longer distances, which means that the chart itself starts to become an issue. In my testing of wider lenses, I use two different charts, one of which I position at the extreme corner. But the images I show won't look the images that you see from others testing this lens. If they're using the standard test charts, they may be testing as close a 36" to the chart. Put another way, that's testing the close focus ability of the lens, and the normal and long distance abilities of the lens may be better or worse. Yet you're less likely to use your lens at close focus distances. I've tried to pick a chart sample that shows what happens at more normal usage distances. These particular samples were taken at 12 feet from the charts.
Light falloff: Fairly decent considering that this is f/1.4. We get about a stop of vignetting wide open, which is good. But the image circle of the lens is very tight to the DX frame, and thus vignetting doesn't get into what I'd call "control" until about f/4. Moreover, you can see almost nothing beyond the DX frame on the sides if you use this lens on an FX body (third image, below).
f/1.4 falloff (DX)
f/11 falloff (DX)
f/11 FX coverage
Chromatic aberration: Lateral chromatic aberration (side to side) was clearly present at all apertures, though not in excessive amounts. Moreover, the chromatic aberration was fairly constant across the apertures, which is a little surprising. There isn't an aperture at which I wouldn't want to do some correction in post processing if I could.
Flare: Not really an issue that I can see. The hood does an excellent job of keeping stray light off the front element. Only direct into the sun does flare become an issue.
Distortion: Obvious barrel distortion (a bit less than 2%) with what looks like just a minor touch of wave to it (the barrel isn't perfectly linear, it tapers faster near the edges while the middles seem to hold straight a bit longer than they should). I was able to correct the distortion without sophisticated software, though.
Bokeh: The eight-blade aperture diaphragm in my sample is slightly lop-sided. It produces a slightly narrow octagon from mid-range apertures down to the smallest ones. The blades are straight, so at all but the widest aperture there is a distinct geometric pattern to the bokeh. Nevertheless, the bokeh looks decent, and approaches good. Still, I'd shoot with this lens wide open if I wanted the best out-of-focus look possible.
Final comments: this doesn't feel like one of Sigma's best efforts. Nowhere do they hit it out of the park with an attribute on this lens. It feels a bit old school in that respect: a prime where they paid some attention to the center but didn't really worry much about getting any of the other details optimized. I do note that Sigma's literature talks about the aspherical element being used to correct chromatic aberration. I suspect they should have had another aspherical element trying to pull in the corner performance. Overall, this lens is a disappointment. The build quality is better than the optical quality that's delivered, in my opinion, which seems like an odd mismatch.
- Don't look to the sides. Corner sharpness never hits a high enough level.
- DX only. The lens just doesn't have the coverage to make it usable on FX, even at 5:4 crop.
- A so-so performer in most respects. About the only thing that stands out is mid-aperture center-of-frame sharpness.
- Build. Built like a brick, with a very nice feel to the focus ring for a low cost AF lens.
- Goes to f/1.4. If you need it, you need it.
- Well hooded. A robust and complete lens hood for a change.