LENS - | User manual | Sigma 15-30mm vs. Canon 17-40mm

Sigma 15-30mm vs. Canon 17-40mm

David Burren, June 2003

This is a review/comparison between these "ultra-wide" lenses in the context of using them on a Canon EOS D30, D60, or 10D

(which have a crop factor of 1.6). All the photos through the lenses were taken with a Canon EOS 10D. This article should probably be read in conjunction with my article on choosing a wide-angle lens for these cameras .

I'll say up-front that I own the Canon 17-40mm lens. I have previously owned a Sigma 17-35mm . The Sigma 15-30mm was borrowed from a friend for about a day and a half for testing and comparison. I took the lenses down to a park near my home for some simple tests.

Some readers may find faults in my tests or conclusions, and I welcome feedback. But please keep in mind that this is not meant to be a complete, scientific, and exhaustive test. This was a test comparing how they "felt" and performed in a few examples of field conditions.

The Sigma 15-30mm f/3.5-4.5 EX Aspherical DG DF

This is a fairly big and heavy lens. As can be seen in this animation (sorry about the different perspectives, I only had one tripod on hand) the front element of the Sigma lens moves in and out a fair way. As it zooms to

30mm the element is retracted into the hood, and at 15mm it's fairly exposed.

The "petal" hood on this lens is permanently attached, to protect that huge convex piece of glass, both from sunlight and from physical damage. On a full-frame camera such as an EOS 1Ds or a film body, the cut-outs in the side of the hood are needed to avoid vignetting the image. But on a camera with a 1.6x crop factor (where the lens acts like a 24-48mm zoom) we can use a longer hood and still not cause any vignetting.

In fact, the lens is shipped with a hood/ring that slips over the petal and provides an 82mm filter thread (as shipped this is occupied by a lens cap). When attached to the lens, this extra hood does NOT vignette the image on these cameras. This extra hood is not just a lens cap (the cap portion comes off) to be left in the camera bag.

My advice is to always leave this attached to the lens, and just remove the lens cap itself. Not only will this allow you to attach an 82mm polarising filter (or in fact any other 82mm filter you want to use) but it provides important protection from flare, as can be seen further below.

Focusing the 15-30mm

The lens has two switches to control auto/manual focus operation. Firstly there's the usual Sigma AF/MF switch on the left of the lens, just near the lens mount (this is the same as found on the Sigma 17-35mm). Secondly, the focus ring can be pulled towards the camera to engage manual focus, or pushed away to disengage the focus ring from the internal focus mechanism.

So there are four possible combinations:

Switch Focus ring

AF

AF

MF

MF

Effect

AF (disengaged) Auto focus, with the focus ring "free-wheeling".

MF (engaged)

Auto focus, with the ability to manually adjust the focus.

The focus ring turns during AF.

AF is slower than with the focus ring disengaged.

CAUTION : "fighting" the AF (for example by resting your hand on the ring) can damage the lens

AF (disengaged) Neither the camera nor you can affect the focus.

MF (engaged) Manual focus, with the camera knowing it's in manual.

This is a great improvement over that of the Sigma 17-35mm, which only had the AF/MF switch near the body, had a focus ring that turned during AF, and did not allow manual override. It's a flexible arrangement, although it's not as good as Sigma's

HSM or Canon's USM (which both offer full-time manual override).

The Canon EF 17-40mm f/4 L USM

This lens was introduced in May or June (depending on where you live)

2003, and is one of Canon's "L" ("professional" lenses). This lens is smaller than the Sigma, although the hood is quite bulky. On an EOS D30,

D60, or 10D, this lens acts like a 27-64mm f/4. It's not quite as wide as the

"24-48mm" Sigma, but it's still a very useful range.

Like many of the recent L lenses, there is a rubber gasket around the lens mount. While this is of most relevance to use on an EOS 1D/1Ds/1V with their weatherproofing everywhere else, it does provide extra protection against dust and moisture even for the D30/D60/10D. Some lenses expand and contract during operation (for example the large "push-pull" zooms) and thus can draw dust and moisture into them when used "in the elements". Not so with this lens. Actually, the front element moves in and out slightly during focussing, but this is inside the filter thread. I have a UV filter permanently attached to mine to provide protection for the lens glass, and this also seals the moving element in.

The controls on the lens are simple: the zoom ring, an AF/MF switch, and the focus ring. Like most USM lenses the lens offers

"FTM" (full-time manual focus override). The focus ring is not directly attached to the internal focus mechanism, but sends electrical signals to the same mechanism controlled by the camera's AF. Once the camera has achieved focus via AF you can tweak the focus simply by turning the focus ring. On my cameras I have set a custom function to move AF from the halfpressed shutter button to a button under my thumb, and then can AF with my right thumb or MF with my left hand. This gives me great flexibility, especially when using a telephoto lens, and having the same system on this lens makes things easy to work with.

Comparing flare

The next photos compare the performance of these lenses in a fairly extreme condition. I carefully set up this scene to try to

maximise the opportunity for flare without actually having the sun in the frame.

The camera was set up on a tripod quite low to the ground, with a 420EX flash on a cord off to the left to fill in the shadows in the foreground. The camera body was not moved between exposures or when the lenses were changed. Both lenses were focussed at or close to their minimum focus distance.

Canon 17-40mm @ 17mm

1/125s, f/11

In this image you can see flare effects (yes, the 17-40mm can suffer from flare although this is an extreme situation). A pink streak in the upper right, with some green haze below it, and a green blotch in the lower left. Click on the image to see a larger version if you can't make them out, and then use your browser's BACK button to return.

The sun was shining full on the front element of the lens, so the hood was hardly providing any benefit. The lens had a Hoya HMC (multi-coated) UV filter attached. I left this attached as this is meant to be somewhat of a "real world" test.

Sigma 15-30mm @ 15mm

1/125s, f/11

The extra hood was attached to the lens.

The perspective is slightly different here for two reasons. In the top half of the frame you can see that we have a wider view due to the use of 15mm instead of 17mm, but in the bottom of the frame you can see that we almost seem to have a narrower view. The cause for this is simply that the 15-30mm is a longer lens than the 17-40mm, so the front element is physically further forward into the scene.

We can see a few obvious examples of flare here: a large red blob in the lower left, and blue and purple dots against the dark tree toward the upper right. As before, click on the image to see a larger version if you can't make them out, and then use your browser's

BACK button to return.

Sigma 15-30mm @ 15mm

1/125s, f/11

The extra hood was NOT attached to the lens.

Here I took the extra hood off the lens just to see what would happen.

The answer: LOTS of flare.

Not only are there more coloured blobs in a line across the centre of the image, but the contrast has been reduced by a haze across the entire frame. Compare the distant tree between this image and the previous.

From the above example it should be clear that having the extra hood attached to the 15-30mm lens when at 15mm is very important. When at 30mm the front element is retracted somewhat into the lens, so how important is it at the "long" end of the lens? Judge for yourself. Both of these photos were taken at 30mm with an exposure of 1/400s, f/8, ISO 100:

As you can see, the image on the left

(without the hood) has significantly reduced contrast, with a blue flare in the lower right. The only thing different about the image on the right is that the extra hood was attached.

The blue object in the lower left of both images is just a children's slide in a playground.

I didn't take a comparable image at 30mm with the 17-40mm, but this is with the lens set to 40mm. No flare that I could notice:

Comparing sharpness

Just a simple test, photographing a brick wall at each end of the zooms, wide open as well as stopped down to f/16. The centre of the image isn't very interesting in terms of sharpness, so I'll just show the overview and a 100% crop from a corner.

All these images were taken on a tripod at ISO 100, using mirror lockup (at f/16 the shutter speed drops off somewhat). The

10D's Contrast and Sharpness parameters were set to "Normal", and the 10D was saving "Large Fine" JPEGs.

No processing has been applied to these images apart from cropping or resizing. Normally I would apply some UnSharp Mask to images before using them, but here I wanted to show the raw camera output.

full view Corner crop

Sigma 15-

30mm

@ 15mm, f/16, 1/3s

At 15mm there's some barrel distortion visible.

Sigma 15-

30mm

@ 15mm, f/3.5, 1/60s

It's noticably softer around the edges.

Sigma 15-

30mm

@ 30mm, f/16, 1/3s

At the other end of the zoom there might be some pincushion distortion, but not much.

Sigma 15-

30mm

@ 30mm, f/4.5, 1/40s

Again it's softer around the edges, but not badly.

Canon 17-

40mm

@ 17mm, f/16, 1/5s

Again there is a bit of barrel distortion.

The exposures for these images are slightly different from that used for the

Sigma lens, simply because the sun moved.

Canon 17-

40mm

@ 17mm, f/4, 1/80s

It's hard to quantify from

this view, but there seems to be very little change between f/16 and f/4.

Canon 17-

40mm

@ 40mm, f/16, 1/5s

Canon 17-

40mm

@ 40mm, f/4, 1/80s

Again, aperture seems to have had little effect on sharpness.

The Canon lens seems crisper "wide open" at f/4 than the Sigma lens, but even so the Sigma performs fairly well.

Conclusions

They're both quite capable lenses, although my feeling is that the Canon offers a slightly better image.

When using the 15-30mm, the extra hood should always be used to reduce the incidence of flare. With it in place I think it

performs quite well. It's still possible to put the lens in situations which result in flare, but this is true of the 17-40mm also.

If you want the extra 2mm (closer to 3mm when you take into account the 1.6x crop factor) then the Sigma will give you that. If you want a Canon "L" lens that has FTM, "feels" like your other Canon lenses, is smaller and lighter than the Sigma and has a

77mm filter size, then the 17-40mm will give you those.

By the way, the cost difference between the lenses is quite small, as can be seen in another article .

Which one should you get?

Sorry, I'm not going to tell you. Personally I'm happy with my own decision to purchase the Canon 17-40mm, but if I had the

Sigma 15-30mm I wouldn't be too upset either. You'll need to make your own decision as to which one is better for you.

Hopefully I've contributed some useful information!