Zuiko Digital 11-22 mm F/2.8-3.5

My other articles related to the Olympus E-System cameras.

This is a user report of the mid-price, wide-angle zoom lens for the Four Thirds system SLRs (mostly, but not only, by Olympus). While basically finished, some corrections and additions may be introduced to it in the next month or two; maybe also a few more image samples.


For the last few years I've heard only good things about this lens from people whose opinions I trust. Now I was able to try it out for two weeks, having bought it for a friend.

This is one of the Olympus mid-range Pro, or High Grade lenses, priced at below $700 in the States. (Olympus is still not sure how to refer to various tiers of their lens offer; while the bottom shelf is called Standard, the top one — sometimes Top Pro, and sometimes Super High Grade.)

This is currently the sole very wide, yet not exotic (like the 7-14 ZD), zoom lens for the Four Thirds system, and it has been available from day one of the Olympus SLR E-System, back in 2003. No wonder the lens is widely used in the Four Thirds community by the people who demand high optical quality, and for whom 14 mm is not wide enough.

The 11-22 mm, F/2.8-3.5 ZD lens, mounted on an E-500 body, shown in the longest (left) and widest (right) zoom setting (the lens is most retracted at about 19 mm).
Shot with my favorite tabletop outfit: the C-5060WZ, lens at 15.4 mm (77 mm EFL), aperture priority (-0.3 EV): 3/4 s at F/7.1, ISO 80. Tripod, two fluorescent panels.

Specifications and what they mean

These are the lens specifications, with my comments added.

  • Focal length: 11-22 mm.

    This is equivalent to (i.e., provides the same field of view as) a 22-44 mm lens mounted on a 35-mm film camera. In the last years of my film era, 20 mm became my favorite focal length for travel scenics (a lifesaver in the Maya city of Tikal, Guatemala) and architecture. This is why I would prefer this lens to have the short focal length of 10 mm — and, believe me, this one extra millimeter is a lot!

    Have a look for yourself: this is the difference between 14 and 11 mm; each pair of pictures was taken from a tripod, so that the camera's position remains unchanged.

14-54 mm F/2.8-3.5 ZD at 14 mm 11-22 mm F/2.8-3.5 ZD at 11 mm
14-54 mm F/2.8-3.5 ZD at 14 mm 11-22 mm F/2.8-3.5 ZD at 11 mm
  • Maximum aperture: F/2.8 at the short end, F/3.5 at the long one.

    For a 22 mm EFL, F/2.8 is a large aperture; a larger one (smaller F-number) would be considered quite exotic (and costly) glass.

  • Minimum aperture: F/22.

    I wouldn't recommend using apertures beyond F/11, where the lens performance fall-off begins, due to diffraction effects.

  • Minimum focus distance: 28 cm from the image plane.

    While a wide-angle lens would rather not be used for making close-ups, the close-focusing capability may be useful for including a dramatic foreground.

  • Maximum magnification: 0.13× at F=22 mm.

    This value corresponds to a magnification of 0.26× on a 35-mm film camera, and this has to be taken into account when making comparisons. It corresponds to a field of view about 14 cm (5.5") across.

  • Construction: 12 elements in 10 groups; no special-dispersion elements.

    Actually, at this focal length range this seems like a streamlined design; most probably chosen to minimize internal reflections (perceived as glare). The lens has internal focusing: done with moving a group of elements internally so that the lens front does not move or rotate; zooming changes the overall length by about one centimeter.

  • Filter size: 72 mm.

    Large (read: expensive filters), but unavoidable at this wide angle. Besides, the only filter you will probably want to use is an UV lens protector, as a polarizer may be not too useful: the sky hue will no longer be uniform due to large differences in angle. And if you want to shoot infrared with this lens, this may become quite costly!

  • Dimensions: 75 mm diameter, 93 mm length.

    This is quite big, but this is common with wide zooms! Actually, the lens is larger than the 14-54 mm, F/3.8-3.5 ZD, see the picture below.

  • Weight: 485 g (17.1 oz).

    Quite heavy, too. This lens weighs more than an E-500 or E-510 body (not to mention the E-410).

  • Drip/dust-proofing: yes.

    While some people need this feature, most don't. Still, this is a standard within the Olympus Pro lens group.

Make, feel, and operation

As we've grown to expect from Olympus Pro (and not only) lenses, there is nothing to complain about in the make-and-finish department. The lens is beautifully made.

The manual focus scale behind an arched, transparent window, is just a decoration — not useful for anything beyond a "sanity check" of the focused distance. Don't even think of scale focusing: the scale is simply too short and imprecise. (I have the same complaint about other ZD Pro lenses I've tried.)

The manual focus ring (coupled electronically to the motor which actually changes the focus) does not provide a precise and positive action. I found manual focusing with this ring an unpleasant experience, much worse than with most of the many mechanically coupled legacy lenses I've tried on my Olympus SLR bodies.

Ironically, the only Olympus ZD lenses about which I do not have this complaint are the "new" kit ones: 14-42/2.8-3.5 and 40-150/4.9-5.6, bundled with the E-410 and E-510. The manual focusing action in those is vastly improved, hard to tell apart from that in smoothly-working mechanical focusing system.

Do not let this deter you: most users, even serious ones, would rarely, if ever, focus ZD lenses manually. Still.

The zooming ring action is just right, perhaps a tad better than in the 14-54 ZD.

Did I mention the lens is quite big? Here it is (at the left), compared to the 14-54 mm F/2.8-3.5 ZD, so far my favorite Four Thirds zoom lens. Both lenses are in the shortest position.

Shot with the C-5060WZ; technical details as above.

In the field

I've used this lens on two bodies: E-500 (mostly) and E-300. Photographers trying to shoot one-handed, in a camcorder style, will find the combination quite front-heavy. On the other hand, this is generally a very bad way to handle a camera, with any lens.

With my left hand providing lens/body support and zooming action, both combinations felt right. (The E-500 one was my preference, in spite of a lighter body, perhaps because of the different body shape, more to my liking.)

The autofocus action was as expected: good if not spectacular on both cameras, but this depends largely on the camera's AF system, not just the lens. I've experienced some slight hesitation and delays at low light levels (one-second exposures at full aperture), but nothing I couldn't live with.

After the kit lenses, the larger maximum aperture of this one makes using the viewfinder, especially in low light, a better experience.

Most of the pleasure of using this lens comes, however, from the wider image angle. This is why most of the time (outside of shooting samples), I've been using it at or near to the short end: between 11 and 14 mm. Among others, I took it for a day trip to the Calvert Cliffs State Park in Maryland, and it allowed me to get some swamp shots which otherwise would be not so great.

At the first glance, the difference between 14 mm (provided by the "standard" ZD zooms) and 11 mm may not look that great, but it really is quite significant.

Anyway, I enjoyed using this lens a lot.

Performance: the MTF data

While I could easily shoot some targets and run the results through the Imatest software, I see no compelling reason to do that. If you would like to see such a test, have a look at the PhotoZone review. performed on the E-300 (all resolution test done with the procedure they're using measure the combined body/lens performance).

The MTF curves for the lens fully open are available at the Olympus Japan site; here they are (11 mm at the left, 22 mm at the right). (For an introduction to MTF, see here; also notice the difference between LP and LW in the specs.)

The left graph is for 11 mm, the right one — 22 mm. Blue lines correspond to a frequency of 10 LP/mm (this can be loosely interpreted as a measure of contrast), while red ones — 30 LP/mm (sharpness). Then, solid lines are for sagital, and dotted for tangential patterns.

Image © by Olympus Corporation

The graphs indicate clearly that, at the wide end, the central resolution is at par with the 14-54 ZD, and the sharpness fall-off is not significant up to 9 or so millimeters away from the image center; this covers the whole frame width (17 mm); beyond that (frame corners are 10.8 mm from the center) MTF values at 30 LP/mm drop significantly at the widest angle.

Keep in mind that such data, especially when provided by the lens maker, may not reflect the actual resolution of the lenses sold in stores; the quality and uniformity of the manufacturing process may affect the real performance. (Most of the large camera manufacturers, like Canon, Nikon, Pentax, or Olympus, have good quality control, so significant deviations are rather unlikely.)

Distortion

Wide-angle lenses, especially zooms, tend to exhibit an optical flaw called geometric distortion. Because the image magnification changes with the distance from the axis (at wide angles: usually increases), straight lines not crossing the image center will be bent (here: outwards, hence the term barrel distortion). The effect is quite hard to correct and therefore quite common, especially at the widest angle.

Here is how our lens performs at the shortest focal length: my local library in Crofton, MD; a thumbnail of the full frame at the right, and a full-width horizontal strip, reduced in size, below.

This is a good performance for a wide zoom: comparable to "kit" zooms at twice the equivalent focal length. Still, after having heard so much good about this lens, I was expecting something better; maybe I have been spoiled by wide primes I've used in the past?

I estimated the distortion at about 1.5%, which is just a tad above the 1.3% reported by PhotoZone.

Note: this passage has been updated since the original posting, as I have switched to the SMIA definition, also used by the Imatest program. The real results did not change; just the definition itself.

At 14 mm the barrel distortion drops considerably (my quick check shows a factor of two, The PhotoZone report even slightly more), to become unnoticeable beyond that. The effect at 14 mm is comparable to that I've experienced in the 14-54 ZD.

Once again, a solid, if not breathtaking, performance — even more so for a wide zoom.

First samples: 11 mm

Note of 2008: Keep in mind that all samples shown here were shot with an Olympus E-500, therefore they are not directly comparable against those from newer models (E-510, E-3, etc.). On one hand, the 8 MP images from the E-500 are less demanding from the lens, but on the other — the newer models seem to have less intrusive antialiasing filters and allow to reduce the degree of noise filtering, both effects resulting in more resolution.

This is the moment of truth, whatever the MTF numbers say. In order to get some idea about this lens' performance, I took about 300 pictures with it, 90% of them from a tripod; in addition to that, another 80 frames with the 14-54 mm, for the sake of comparison.

Obviously, I'm not going to bother you with all these samples; this would be boring, wasteful, and pointless (in particular, this time I'll save you my usual samples of the Crofton lake, of which I took two series). Still, I went methodically through all series shot, analyzing them carefully on my 1600×1200 screen, most of the time "flipping" between pairs or triplets of images in full pixel scale, with the zoom and pan of my browser locked — an amazingly efficient technique of spotting minor differences between almost-identical samples. Therefore I'm fairly confident about my conclusions.

As always, the 1:1 pixel scale fragments are, except for cropping, not adjusted in any way, while the full-frame thumbnails have been re-sharpened after size reduction. All pictures were saved as SHQ (1:2.7) JPEGs, and my E-500 was set to the Normal image mode, with contrast additionally reduced by one notch, and sharpening — by two.

I believe that pictures saved this way in-camera submit better to further postprocessing; on the other hand, the first impression may be less sharp than if they were saved with the default (I believe, somewhat excessive) settings. Sharpening increases the steepness of contour transitions (making them look more "sharp"), but it does not preserve, or generate, any more detail. Photographers who do not know that, should not be allowed to view lens evaluation samples. I hope I'm lucky, having Readers who do not need this reminder.

All this said, let's start from a few casual samples, the only ones shot handheld.

11 mm; aperture priority (0 EV): 1/400 s at F/5.6, ISO 100; WB at 5300 K.
At F/5.6, the sharpness is exemplary near the center, and slightly worse, but still good, in the edge areas, where also a moderate amount of chromatic aberration can be seen: purple away from the lens axis, and green towards it (counting from the highlights, from where the effect spills into darker areas). This I found somewhat disappointing, after having heard so much raving about this lens.
11 mm; aperture priority (-0.7 EV): 1/250 s at F/8, ISO 100; WB at 5300 K.
Interesting: not just the center, but also corners are very sharp here. A great show, indeed. Is it possible that in the previous sample some resolution was lost to local glare? (Disregard the red flower: it seems to be out of the depth-of-field range, being just about eight inches from the lens.)
11 mm; aperture priority (-0.7 EV): 1/320 s at F/8, ISO 100; WB at 5300 K.
This one is a no-brainer, and the lens took it on a stride. Actually, my other samples led me to believe, that at F/5.6 the performance is a bit better than at F/8, — but it is excellent anyway.

For a better impression (see my remark above), a moderate amount of sharpening (anything but the crude unsharp mask!) would make the characters in the blackboard really stand out. Note the detail in the trailer company logo above it.

The next image was chosen to show the depth of field you can get at the 11 mm focal length.
11 mm; aperture priority (0 EV): 1/80 s at F/8, ISO 100; WB at 5300 K.

At F/8 I'm getting both the foreground and the background acceptably sharp — and that's in spite of less-than-ideal focusing: right on the post, instead of slightly beyond. This wastes the DoF in front of it. For more critical uses it would be nice to have a "depth-of-field priority" autoexposure mode, like the one in the Canon 400D.

If you have to know: the camera was focused at 1.165 m (according to the EXIF data). At F/8 and F=11 mm, the hyperfocal distance is, according to my tables, 1.02 m. When focused at that distance, the DoF covers the range from half that (0.51 m) to infinity. This is computed for the commonly used circle of confusion (CoC) of 0.015 mm (or 1/1440 of frame diagonal, or about three pixels) across. Please, don't get me started...

Samples at various focal lengths and apertures

To see how the lens performs at various apertures and focal lengths, I have shot three different series, iterating through all combinations of these two factors. While I've analyzed them all, let me show you just one, with 1:1 samples chosen from the center of the frame. To save space, I'm including only three apertures for each focal length: the full opening, the mid-range F/5.6, and diffraction-impaired F/16.

F = 11 mm

Obviously, this is the focal length of most interest to us: the main reason to buy and use this lens.


11-22 mm at 11 mm, full frame

11-22 mm at 11 mm, F/2.8

11-22 mm at 11 mm, F/5.6

11-22 mm at 11 mm, F/16

Surprisingly (and I could see that in the other series), the sharpness at the center does not really change between F/2.8 and F/8 (or maybe a bit less); diffraction effects start being visible at around F/11, getting worse at F/16, and quite severe at F/22.

Maybe some changes could be detected with the proper measurement technique — but if I cannot tell the difference flipping between images from an 8 MP camera in a 2:1 pixel scale, then I really do not care. Neither should you.

Now, corner samples at 11 mm, to check for chromatic aberration, which should be most visible at this focal length.

Again, some chromatic aberration in extreme corners is visible; this may be the reason for the MTF fall-off above 9 mm away from the axis I've discussed before. This improves at F/4, but then stays on the same level, as far as I can tell, with the affected width of about 1-2 pixels.


11-22 mm at 11 mm, F/2.8

11-22 mm at 11 mm, F/5.6

11-22 mm at 11 mm, F/16

F = 14 mm

This is also the shortest focal length offered by the "standard" ZD zooms, including the 14-54 mm I like so much. Let's have a look.


11-22 mm at 14 mm, full frame

11-22 mm at 14 mm, F/3.0

11-22 mm at 14 mm, F/5.6

11-22 mm at 14 mm, F/16

At first I thought I had a focusing problem — but this was not the case: the results are quite consistent on that. At 14 mm the center is at its best at the full aperture (F/3.0), getting a bit worse at F/4, and staying at the same level all the way until, at F/11, going downhill again. While not common, this is not impossible (no, you do not want to listen to all details on that).

Now the corner samples again. The chromatic aberration at full aperture seems now to be better corrected, although some hints are still visible.

These samples are not a proof of that: comparing different frames is not safe here; the edge contrast may be different; this is rather an impression I've got from all relevant images. As unscientific as it may be, I'm quite confident in it.

The unsharpness in the leaves can be attributed to limited depth of field, which was extending from four meters or so up.

Anyway, the chromatic aberration at 14 mm seems to be well controlled.


11-22 mm at 14 mm, F/3.0

11-22 mm at 14 mm, F/5.6

11-22 mm at 14 mm, F/16

F = 22 mm

Let me skip the next marked focal length, 18 mm, going straight to the long end of 22 mm. Here are the samples.


11-22 mm at 22 mm, full frame

11-22 mm at 22 mm, F/3.5

11-22 mm at 22 mm, F/5.6

11-22 mm at 22 mm, F/16

This is becoming predictable: results hard to tell apart for apertures up to F/8, then degrading due to diffraction (although not as strong as for the shorter focal lengths). I'm not even showing the corners: the hint of chromatic aberration is a bit less than at 14 mm.

Comparison with the 14-54 mm ZD at 14 mm

During the same session I took a series of almost-identical pictures with the 14-54 mm F/2.8-3.5 ZD zoom lens. Let us have a quick side-by-side comparison at various apertures, showing the frame center as before.


11-22 mm at 14 mm, F/3.0

14-54 mm at 14 mm, F/2.8

11-22 mm at 14 mm, F/5.6

14-54 mm at 14 mm, F/5.6

11-22 mm at 14 mm, F/16

14-54 mm at 14 mm, F/16

The apparent differences are quite subtle: at the full aperture, the wide-angle zoom seems to have a slight edge; the situation is reversed at F/5.6. I could see this in two out of three such comparative series I've shot; in the third one the match was too close to call. Anyway, we are hairsplitting here. These are not differences you would see in a 30×40 cm (12×16") print.

So, what do I think?

This is a very good, if not outstanding, lens. Trying to combine a short focal length with large maximum aperture had to induce some compromises in the distortion and aberration department, especially if the lens were to stay in the non-exotic price range.

While generally offering a solid optical performance (in addition to top-notch construction quality), it suffers a bit at the shortest focal length of 11 mm, losing some resolution in the corners and showing some chromatic aberration there. While these flaws are unlikely to show enlarged up to 30×40 cm (12×16"), I was slightly (just slightly!) disappointed after having heard to what most users had to say about it — especially after my previous acquisition of the 50 mm Macro ZD.

Update of 2008: Now, a year after this article was originally written, I have to rewrite the conclusions, so the rest of this section is brand new, to reflect my current thoughts.

First of all, originally (just a year ago) I wrote:

One lens causing a stir in the Four Thirds community is an upcoming 12-60 mm ZD, expected later in 2007. Its wide-angle capabilities will be close (if not the same; 1 mm makes a visible difference here!) to those of the 11-22 mm, and the long end of 60 mm makes it ideal for travel photography and other applications of general nature. I do not expect that lens to match the 11-22 mm optically, especially at the short end, but the lens may be "good enough" for what most of us are doing.

Now, my expectations were wrong. The 12-60 mm ZD turned out to be as close to perfection as possible. True, the one-millimeter difference at the short end can be noticed, but it is not that significant, and what you get with the new lens is the whole range up to 60.mm, making it an excellent all-around shooter.

Most importantly, optically the 12-60 mm ZD is at least as good, perhaps better than the 11-22 mm ZD, something I would hard to believe in a 5× zoom. This is against my preconceptions, but I've used both lenses a lot, and that's what I can see. (A direct, strict comparison is not practical, as the 11-22 ZD samples were shot with the E-500, while the 12-60 ZD ones — with the E-510 and E-3, see my note above).

The 12-60 mm ZD is just $250 or so more expensive, which makes it a very strong contender in this context (for me: a clear winner).

Then, another alternative: the 7-14 mm F/4.0 ZD. True, it is one F-stop darker (at the wide end) and costs more that twice as much. So what? You get what you pay for: its resolution is outstanding, and the wideness of the short end has yet to be matched in any rectilinear lens for a Four Thirds (or APS-C) camera. If you really need a super-wide angle, that's the only solution.

In 2008 Sigma released the 10-20 mm F/4.0-5.6 EX DC HSM (the "DC" is important here: this is not a re-labeled 35-mm "full frame" lens!). I haven't tried it yet, and I haven't seen any meaningful samples, but at $500 it is aiming for a share of the same market, even if it is one F-stop darker.

Last but not least, later this year we are expecting the 9-18 mm F/4.0-5.6 ZD ED by Olympus. All right, I do not expect it optically to be a match for the 11-22 mm, and, again, it has a smaller aperture, but 9 mm (as compared to 11) is nothing to sniff at, and many people who do not want to spend $1500 for the 7-14 mm ZD (which is worth every penny of that price!), may see it as an attractive budget alternative.

With all these choices, I no longer consider the 11-22 mm ZD a very attractive offering, at least not without a significant price cut. Feel free to disagree with me on this subject: after all, your needs and preferences may differ.

Actually, what I think the ZD line needs is a mid-priced, wide-angle lens with a fixed focal length of 10 mm and an aperture of F/2.8; I would be the first to stand in line to get it, especially if I can afford it.


My other articles related to the Olympus E-System cameras.

Olympus® is a registered trademark of Olympus Corporation.
This page is not sponsored or endorsed by Olympus (or anyone else) and presents solely the views of the author.

Home: wrotniak.net | Search this site | Change font size

Photo Tidbits | The Gallery


Posted 2007/07/09; last updated 2008/06/28 Copyright © 2007-2008 by J. Andrzej Wrotniak