Canon EF-S 17-55/2.8 IS USM vs. Sigma 17-50/2.8

The premium standard zoom for APS-C sensor cameras usually covers the range from around 17mm to around 50mm and has a maximum aperture of f/2.8. In 35mm terms this coverage is equivalent to something like that which a 27-80/2.8 zoom would have on a full frame camera.

Such standard zoom lenses are frequently used by event photographers (e.g. weddings) and cover the range from wide angle to short telephoto. At the telephoto end of the range they are good for “head and shoulders” portraits and at the wise end of the range they are good for group shots. This range is also very useful for travel photography and the f/2.8 maximum aperture can create attractive background blur when desired as well as allowing shooting in low light (an ability that is enhanced by the built-in optical stabilization of both lenses).


In this review I’ll take a look at two pretty similar premium standard zooms, the canon_17-55 and the sigma_17-50. Both lenses are APS-C only, have an f/2.8 maximum aperture, have built-in optical image stabilization and use a ultrasonic/hypersonic focusing motor. Both take 77mm filters but only the Sigma lens includes a lens hood and soft case. An optional lens hood is available for the Canon EF-S 17-55/2.8 (EW-83J, approx. $45 street price).

Where to Buy’s partners have both the Canon 17-55/2.8 and the Sigma 17-50/2.8 lenses available. Their prices are fair and you help to support

  • canon_17-55
  • sigma_17-50

The table below compares the optical and physical specifications of the two lenses:

Canon EF-S 17-55/2.8 IS USM and Sigma 17-50/2.8 Specifications

Canon EF-S 17-55/2.8 IS USM Sigma 17-50/2.8 EX DC OS HSM
Lens Construction 19 elements in 12 groups 17 Elements in 13 Groups
Angle of View (diag) 78.5° – 27.9° 72.4° – 27.9°
# of Diaphragm Blades 7 7
Mininum Aperture f22 f22
Minimum Focusing Distance 35 cm / 13 in 28 cm / 11 in
Filter Size (mm) 77 77
Maximum Magnification 0.17x 0.20x
(Diameter x Length)
83.5  x 110.6 mm / 3.3 × 4.4 in 83.5 × 91.8 mm / 3.3 × 3.6 in
Weight 645g / 22.8 oz. 565g / 19.9oz.
Lens Hood Optional Supplied
Mounts Canon Canon, Nikon, Sigma,
Pentax, Sony/Minolta

As you can see the Sigma lens is slightly smaller and slightly lighter – and of course it has a slightly smaller zoom range. The Canon lens has the typical semi matte black finish, while the Sigma lens has their matte slightly rubberized “zen” finish. Both zoom and focus smoothly.

The Sigma 17-50/2.8 uses FLD glass (similar to fluorite), one hybrid aspheric element and two molded glass aspherics to minimize aberrations. The Canon EF-S 17-55/2.8 uses UD glass Ultra low dispersion) and three aspheric elements.

Image Stabilization Mechanics

Both lenses seem well built and both rattle a little when shaken, presumably due to the movable optical group which is used to achieve optical stabilization. The rattle on the Sigma is a little louder and more noticeable than that on the Canon.

One thing I did notice is that the Sigma 17-50/2.8 always makes a very slight noise when attached to the camera and the camera is turned on. When the shutter is 1/2 pressed there’s another noise that joins in with the original noise as the stabilization system begins to operate. It sounds like something is spinning all the time that the lens is getting power from the camera, not just when the stabilization is active. On the other hand, the Canon EF-S 17-55/2.8 is totally silent when mounted on the camera until the shutter is 1/2 pressed when you can hear the control gyros start up. The noise level during IS (image stabilization) operation is low, but not as low as the Sigma lens, which is pretty silent unless you have your ear up to it. I’m not quite sure what to make of the constant low level noise of the Sigma lens. If it is spinning gyros at all times it might consume more power then the Canon, but I can’t really measure that and neither lens seemed particularly power hungry.


Both lenses have a zoom ring, which turns in the same direction, with 17mm on the left and 50/55mm on the right.

The Canon EF-S 17-55/2.8 has focal lengths marked at 17, 20, 28, 35 and 55mm, while the Sigma 17-50/2.8 is marked at 17, 21, 28, 35 and 50mm. Both lenses extend when zoomed out from 17mm.

Zoom Creep

The Sigma lens I tested appeared to be brand new, while the Canon lens appeared to be used and from their “loan pool” (lenses loaned to professional for evaluation). Bearing this in mind, I found that the zoom on the Canon lens was a little loose, i.e. there was some zoom creep. With the lens held vertically the focal length setting would change. The Sigma lens zoom ring was tighter and the lens exhibited no zoom creep.

An informal unscientific poll of Canon EF-S 17-55/2.8 owners suggests that although zoom creep is not unknown, it’s not “normal” for this lens. 12 out of 15 owners reported no zoom creep at all while there were 3 reports of zoom creep under some conditions (such as when carrying the lens on the camera with the lens pointed down).

Manual Focus


The Canon EF-S 17-55/2.8 has a manual focus throw of approximately 90 degrees going from infinity to closest focus (35cm). The focus ring of the Sigma 17-55/2.8 only rotates though about 40 degrees to go from infinity to close focus (28cm). This means that manual focus on the sigma lens is faster, but manual focus on the Canon lens is capable of finer adjustment.

The Canon 17-55/2.8 has full time manual focus, which means that the focus can be manually “tweaked” at any time without having to switch the lens from AF to MF. Also the manual focusing ring does not spin during autofocusing. In contrast the manual focusing ring of the Sigma lens does spin, even in AF mode and in order to manually adjust focus the AF/MF switch must be set to MF. Actually you can turn the focus ring in AF mode , but it’s very stiff and obviously not meant to be used that way.


Both lenses have optical stabilization and both have a stabilization “ON/OFF” switch on the lens barrel just below the AF/MF switch

Optical Performance



As would expected both of these lenses do show some vignetting throughout the focal length range when used wide open at f/2.8. Overall the Canon EF-S 17-55/2.8 shows very slightly more vignetting then the Sigma 17-55/2.8, but both lenses are pretty good and most of the visible vignetting goes away if either lens is stopped down by a stop, especially at longer focal lengths. In both cases the vignetting is most noticeable at 17mm.

Canon’s DPP program will automatically correct vignetting when you shoot in RAW with the Canon EF-S 17-55/2.8 IS but will not correct images shot with the Sigma 17-50. Some Canon DSLRs can also correct vignetting in-camera when Canon lenses are used. However, other RAW converters and image editors may be able to apply correction to images shot with both lenses.



Again both lenses show fairly good control of distortion. The table below gives an rough estimate of the distortion based on my measurements and using the SIMA definition 5.20 for TV image distortion (based on the edge and center height of a box which fills the frame). This model assumes simple 3rd order distortion (i.e. simple barrel or pincushion rather than complex), which appears to be a good assumption for both of these lenses.

  Canon Sigma
50mm 1.6% (pincushion) 1% (pincushion)
35mm 1.7% (pincushion) 0.8% (pincushion)
17mm 0.6% (barrel) 1% (barrel)

Just as a reference point, the EF-S 17-85/3.5-5.6IS USM at 17mm shows about 3.2% barrel distortion measured using the same technique.

Image Stabilization

Both lenses have built in optical stabilization achieved by moving a small group of lens elements to correct lens motion detected by built in sensors. Sigma claim 4 stops of stabilization while Canon claim 3 stops. However there is no well defined test for the degree of stabilization a lens can deliver. “X” stops is always really “up to X stops” and it’s a probability game anyway. 2 stops of stabilization may indicate a probability of 70% that an image will be 90% as sharp as a “perfect” image – or it may be a 60% probability that an image will be at least 75% as sharp as a perfect image. No manufacturer has yet published a set of standards on which they base their claims so I don’t put a huge amount of faith in manufacturers claims when it comes to image stabilization.

In testing both lenses I found that the actual stabilization achieved amounted to between 2 and 3 stops in “real world” tests, based on getting maybe 60% of the images as sharp or almost as sharp as tripod mounted shots. If I had to say one lens was better stabilized than the other, I’d probably give a slight edge to the Canon, but it’s really hard to quantify the difference. I’d really just say that you can expect somewhere between 2 and 3 stops most of the time. Occasionally you might see 4 stops with either lens, though that’s probably as much due to luck as it is to optical stabilization!

Color Balance


While I don’t really regard color balance as an important characteristic of a lens when shooting digital (since color changes can be made so easily), with the same camera settings (default white balance), the Sigma lens generally gave slightly warmer looking images than the Canon.


Neither of these lenses makes any claim to be a macro lens. The Sigma 17-50/2.8 has a close focus distance of about 11" at all focal lengths, resulting in a maximum magnification of approximately 0.2x (1/5 life size). The Canon lens doesn’t focus quite so close, 13" at all focal lengths, so despite the slightly longer maximum focal length the magnification is slightly lower at 0.17×.

Image Sharpness


Both lenses are reasonably sharp in the center and pretty good at the edges at all focal lengths. As you would expect with a fast lens, the image quality in both cases improves when they are stopped down (gaining in both contrast and resolution). However, unless large prints are required, both lenses can be used wide open without major concerns over image sharpness.


At 17mm with both lenses wide open I’d say that the Sigma may have a very slight sharpness/contrast advantage, but the differences are small and even a very small amount of focus error can easily mask any differences.

At 35mm both lenses were pretty sharp wide open in the center.

11680310 11680311

At the edge the Canon was slightly sharper but both were still pretty good. There was no visible chromatic aberration. As expected, stopping down does improve sharpness a little, but if you are buying and f/2.8 lens, f/2.8 is where you want to do the comparison. The images on the right are 100% crops shot using and EOS 40D and each of the two lenses at 35mm and f/2.8.


At 50mm wide open (f/2.8) the Canon EF-S 17-55/2.8 seems to have a slight sharpness edge over the Sigma 17-55/2.8, perhaps more noticeably at the edge of the frame than in the center, but as the lenses are stopped down the differences disappear. Again, at the long end of their zoom range these two lenses were very close in performance

Chromatic Aberration


Chromatic aberration (CA) is well controlled in both lenses and is unlikely to be a significant issue with either lens. The image on the right links to a comparison of the two lenses shot wide open at 17mm. They are 200% blowups because at 100% it’s hard to see the CA. As you can see, the Sigma lens does show slightly higher CA, but both lenses are pretty good. These crops were taken from the extreme edge of the image.

At 35mm CA was very hard to see, in fact I couldn’t detect any in real world images even at extreme magnification on a monitor screen.


At 50mm CA was hard to see. The images on the left are taken from the extreme corners of images shot with the two lenses and displayed as 200% crops so the small amount of CA can be seen. In this case, the CA of the Canon lens is perhaps a little more visible.

The bottom Line

The bottom line is that both the canon_17-55 and the sigma_17-50 are excellent lenses and their overall performance is very similar. There are some small differences in sharpness, vignetting, color, distortion etc., but the differences are small and overall don’t strongly favor either lens.

The major difference between these lenses is price more than performance. The Sigma can be obtained for around $400 less then the Canon, plus it comes with a lens hood, a soft case and a 4 year warranty (1 year + 3 year extended). The Canon warranty is 1 year, there’s no case supplied and the lens hood is $45 extra.

Given the choice of either lens at the same cost, I’d go for the Canon for several reasons. First is guaranteed compatibility with past, present and future EOS bodies. Second, the IS may be slightly better, the zoom range is slightly longer and it does have full time manual focus with a focus ring that doesn’t spin during AF. However, if I had to buy the lens with my own money, I’m not sure I’d be willing to pay an extra $400 for the Canon lens (and an extra $45 if I wanted the hood).

So while the Canon may overall be the slightly more desirable lens, the Sigma is a better value and certainly seems to provide the most “bang for the buck”. If you need a fast standard zoom, I don’t think you can go wrong with either one.

Where to Buy’s partners have both the Canon 17-55/2.8 and the Sigma 17-50/2.8 lenses available. Their prices are fair and you help to support

  • canon_17-55
  • sigma_17-50

Image Samples

11673693 Canon EOS 40D, Sigma 17-50/2.8 OS @ 30mm, f/4, 1/160s, ISO 100 This shot was taken just before sunset, hence the warm color balance. Distortion at 30mm focal length is very low. This shot has the same field of view as a standard 50mm lens would have on a full frame camera
11673691 Canon EOS 40D, Canon EF-S 17-55/2.8 IS @35mm f/4, 1/320s , ISO 100 The building on the left of the frame appear to lean inwards slightly due to the camera being tilted upwards for this shot. They are not the result of lens distortion.
11673690 Canon EOS 40D, Canon EF-S 17-55/2.8 IS at 17mm, f/2.8, 1/200s, ISO 100 The distortion of the Canon lens is very low at 17mm making it useful for a architectural work without the need for digital correction
11673694 Canon EOS 40D, Sigma 17-50/2.8 OS @ 50mm, f/5.6, 1/250s, ISO 100 A 17-50mm lens is generally useful as a travel lens, spanning everything from landscapes (or cityscapes) to closer views of interesting subjects.

Sign in or Sign up to post response

    • I bought the Sigma lens as a fast zoom for wedding photography on my Nikon D300. I have been very pleased with the lens. I was surprised when you zoom the lens it is in the opposite direction of the Nikon Lens.  (Nikon goes clockwise from Wide to Tele.) It is true it comes with a bayonet lens hood but the bayonet hoods have a tendency to move or fall off, particularly at a hectic wedding. I had to solve this problem with tape.

      In the end, because of its close up ability and vibration reduction, it has proved to be a good general lens as well. The price of the Sigma certainly makes it a good value. I would recommend it.

    • To post a reply Sign In
    • So the Sigma seems really good value...How about compared to the Tamron 17-50 2.8?

    • To post a reply Sign In
    • Thanks for this comparison.  I've been looking for a lens in this range, and the Sigma definitely caught my eye.  From the sample photos and your comments, it really seems like a good way to go.

      The main reason I would take the Sigma over the Tamron is the faster/quieter auto-focus of the Sigma.  Though, it is a shame that full-time manual isn't available on either of them.  

    • To post a reply Sign In
    • I've no idea about the Tamron SP 17-50mm F/2.8 VC since I've never had the chance to use one. It does have stabilization and it's a few dollars cheaper than the Sigma, so it's clearly a contender.

    • To post a reply Sign In
    • I have used Sigma lenses on several different Canon cameras, and for the price I like it.  But like with all brand v/s off brand, brand always has a slightly better look and feel.  I would buy this lense if I had $600 laying around.

    • To post a reply Sign In
    • I did a comparison between Tamron 18-50 F2.8, Sigma 17-50 F2.8  and the Sigma 17-70 F2.8-4 In my limited test in a camera showroom Both Sigmas had less wide angle distortion and sharpness wide open were noticeably sharper than the Tamron.

      I went  in the store expecting to buy the Tamron or Sigma 17-50. but ended up with the Sigma 17-70 as the extra range was more important than the constant aperture and the 17-70 focuses closer I believe. I could see no significant difference in IQ between the 2 Sigmas.

      I'd highly recommend either.

    • To post a reply Sign In
    • Forgive me if you did cover this, i searched for the word noise in case I missed it .. what is the noise of the motors on the sigma like compared to the canon. I have had some sigma lens in the past whose motor sounded like a dentist drill. 

    • To post a reply Sign In
    • Sigmas HSM motors are virtually silent, but there is a very slight clunk when switched on.

    • To post a reply Sign In
    • Hi Bob,
      didn't you review Tamron 17-50 VC in this very site?

      I have tried briefly my friend's Tamron 17-50 VC and was not satisfied with the results. My non-VC version apparently gives sharper images.

    • To post a reply Sign In
    • I bought the Sigma lens 12 days ago and i am very happy the way it performed. I agree Bob, performance wise, it has more or less equal performance with the Canon lens and it is cheaper. For me, optical performance is more important than its feature. Good job Sigma! this is my first Sigma lens, i used to have Canon lenses.

    • To post a reply Sign In
    • I have the Canon 17-55 2.8. It's a great lens from image quality perspective, but too bulky for carrying around for non-professional or "serious" use. I am seriously thinking in selling it and buying the Sigma instead. Charts show that Sigma's sharpeness @f/2.8 is rather worst in comparison, specially in the edges, but @f/4 are equalled or even has a slight advante over the Canon.

    • To post a reply Sign In

Sign in or Sign up to post response