Olympus M. Zukio 12-50 f3.5-f6.3

The Olympus M. Zukio 12-50 f3.5-6.3 is one of the more controversial lenses. It has it’s fair share of defenders and detractors. It has not done particularly well on a number of bean counting sites. “The Olympus lens can't fully convince in terms of resolution,” gripes photozone. Lenstip shows, if anything, even more disdain for the lens:

It would be difficult to go into raptures here. The Olympus 12-50 mm has problems to pull level with the “pancakes” which are already a kind of compromise between a high image quality and small dimensions. Maximum results of less than 60 lpmm are certainly not impressive.

Thom Hogan’s review was not much less negative. “The kit lens does more than well enough, indeed, often better than [the 12-50], on most things for stills,” he opined. Around the web the lens is generally considered no better than the various 14-42 kit lenses and decidedly inferior to the Panasonic 14-45 f3.5-5.6. When the lens is praised at all, it is often in a rather condescending tone. Often one hears that the 12-50 is “good for a kit lens,” which is a bit of back-handed compliment.

I’ve had a chance to get use and get to know the lens. I’ve actually found it to be a better lens than its reputation would suggest. While I wouldn’t go quite so far as to call it a great lens, it’s actually better than would be expected, considering it’s internet reputation.. In any case, it doesn’t deserve much of the criticism that has been directed at it. I have found the lens to be better than the well-regarded Lumix 14-45. Granted, it’s not quite a good as pro-quality glass. It’s definitely a consumer grade zoom lens. But as consumer grade zoom lenses go, the 12-50 is a pretty good one.

Resolution. First, let me take on the myth that the lens is not sharp. While it may not be “tack” sharp, it’s still plenty sharp for most practical purposes. Indeed, I would go further. In terms of practical output, I don’t think you’re going to notice much of a difference in the resolution attained by the 12-50 and the resolution attained by many of the outstanding primes lenses available for micro-four-third platform. Yes, the prime lenses are sharper; and if you pixel peep, this will be evident. But short of pixel peeping, the extra resolution is not likely to be noticed. In large prints, for example, you’re only likely to notice if you have real good eyesight and look close. At web resolutions, you won’t be able to detect a difference in sharpness. If there’s an area where the 12-50 might come a bit short in terms of high class lenses, it’s in lens contrast and color rendition. Pro quality glass will feature better contrast and brighter, richer, and/or more vivid colors, leading to images with more “bite,” “pop,” or “snap.”

Now let’s look at some crops. One word of warning, however: these shots are not meant to be “tests.” They are only meant to illustrate what I have found through using the lens in actual photographic situations. Rigorous tests have their place; but, unfortunately, such tests tend to be of subjects which most photographers, or at least most serious photographers, eschew, such as MTF charts, brick walls, patio furniture, stuff animals, books, newspaper print, etc. The following images are meant to be illustrations of what the 12-50 is capable of accomplishing when used for “real” photography.

Unless otherwise noted, the following set of images are all shot in raw with the OM-D EM-5, at f7.1. Since I rarely shoot zoom lenses wide open, I can’t tell you how well the lens performs wide open. Most reviewers of the lens seem to believe the 12-50 tends to be sharpest wide open. Perhaps they are right.

Let’s begin with an image shot at 12mm.

Here’s the 100% center crop:

Here’s the 100% corner crop:

We plainly see in this example the chief weakness of the 12-50: wide open the edge to edge, corner to corner resolution is not great. But this happens to be true of most zooms that go fairly wide. Indeed, most lenses, prime or zoom, tend to be sharper in the center than in the corners.

Here’s another 12mm shot:

The 100% center crop:

The 100% corner crop:

In this image the corner seems to exhibit better resolution. The reason for that is that the point of focus is much closer. At closer distances the lens seems to resolve better edge to edge at wide focal lengths. I’m not sure if this is due to field curvature issues or to perspective issues relating to greater distances between optimal focus point in center and optimal focus point along edges when photographing distance objects.

Here’s one more 12mm shot:

Here’s the 100% center crop:

Here’s one 100% corner crop:

Here’s the other, far corner, also at 100%:

Perhaps it’s my imagination, but it does seem to me that the latter corner crop is a bit sharper than the former. This relates to the issue of distance to the corner, which is often exaggerated at wide angle focal lengths. The latter crop is farther from the lens than the former crop, and hence, is closer to the optimal point of focus. This is an issue that must be kept in mind when evaluating resolution of corners achieved by wide angle lenses.

Here’s an image shot at 14mm:

The 100% center crop:

The 100% corner crop:

Even at 14mm, we’re already beginning to see improvements in edge to edge, corner to corner sharpness. This improvement will steadily continue until the 25mm to 30mm range.

Here’s an image shot at 18mm:

Center 100% crop:

Edge 100% crop:

Here’s image shot at 26mm:

The 100% center crop:

Corner 100% crop:

Looking at these crops, one wonders why anyone would complain about the resolution of this lens. (More on this below.) This is actually a pretty sharp lens. Other than the less than stellar results in the corners and along the edges at the wide end of the lens, it’s really a very nice lens.

Here’s a shot at 25mm:

The 100% center crop:

And the 100% edge crop:

If these crops don’t look as sharp as the previous 26mm set, that’s because of the distance of the subject. Distant subjects will always lose at least a little bit of resolution due to atmospheric conditions. You shooting through more air; and that leads to a loss of resolution.

Here’s a shot at 35mm:

The 100% center crop:

The 100% corner crop:

Here’s a shot at 39mm:

The 100% center crop:

The 100% corner crop:

Here’s a 43mm shot, taken at f6:

The 100% crop:

A corner crop here would be superfluous. This image is taken in the 12-50’s special macro mode. It works only at 43mm. To get even a remote idea of the corner to corner sharpness in the macro mode of the lens would require a rigorous formal test. I personally won’t be doing any macro shots where corner to corner sharpness would be important; so I’ll be undertaking no such test.

Here’s a shot at 50mm:

And the 100% center crop:

The 100% corner crop:

One more 50mm shot:

The 100% center crop:

And the 100% corner crop:

The general consensus is that the 12-50 loses some resolution at the long end of the zoom. While I suppose that’s true, I’m seeing very little loss or resolution in my copy of the lens; certainly not enough to get uptight about. While I do wish the lens were sharper corner to corner at the wide end, I am otherwise entirely satisfied with its performance, in terms of resolution.

To further refute the claims that the M. Zukio 12-50 is a “soft” or “mediocre” lens, I’m going to do some comparisons with other lenses. Again, these are not rigorous, formal tests: they’re merely illustrations of what I have found through real world use of these lenses.

Here’s a 100% crop from Panny 14-45, taken at 20mm focal length, f7.1:

And a similar 100% crop from the 12-50, taken at 19mm, f7.1:

These previous crops were shot with EM-5. The following are some crops shot with Lumix GH1. These are jpegs sharpened by the camera.

100% center crop from 12-50, shot at 19mm, f7.1:

100% center crop from 14-45, 18mm, f7.1:

100% corner crop of the 12-50, 19mm, f7.1:

100% corner crop of the 14-45, 18mm, f7.1:

At 18mm to 19mm, I’m unable to detect any significant difference in resolution between these two lenses. The notion that the Lumix 14-45 is better than the 12-50 is an internet myth. As a matter of fact, I would consider the 12-50 to be the better lens, not merely because of the greater focal length of the zoom, the macro mode, and the weather sealing, but because of the superior color rendition of the Olympus zoom. Part of the reason why I bought the 12-50 was due to my dissatisfaction with the color rendition of the 14-45. Originally, I had purchase the 14-45 because of the widespread disparagement of the 12-50, particularly in comparison with the 14-45. But after using the 14-45 for eight months, I became increasingly frustrated with it. I was spending too much time trying to fix color issues related to the lens. In particularly, I found the colors produced by the lens a tad flat. This suggests that inferior coatings were used in the lens, leading to a loss of color contrast. This is an issue difficult to fix in post. My style of photography requires bright, vibrant color; and I was just not achieving that with the Lumix 14-45. Meanwhile, I was noticing that, despite all the negativity directed online against the 12-50, the actual images I saw produced by the lens on the internet had the colors I was looking for. This is why I wound up with both lenses. (And I will be keeping the 14-45 for infrared work: it’s considerably better for that kind of photography than the 12-50.)

How about some comparisons between the 12-50 and the M. Zukio 9-18? First, an uncropped 12-50 pic, shot at 15mm and f7.1:

And the 9-18 version, shot at 18mm, f7.1:

The light changed just a bit between the 12-50 and the 9-18 shot, so that has to be kept in mind when comparing the images. Nonetheless, I find them quite close. I suspect the 9-18 has just a bit more lens contrast, and better color rendition. Otherwise, there’s not much to choose between them. Let’s examine some crops.

100% center crop with the 12-50:

100% center crop with the 9-18:

100% center crop with 12-50:

100% corner crop of the 9-18:

These are very close. The 9-18 looks a tad sharper in the center, the 12-50 a tad sharper in the corners. In real world use, these differences would be insignificant. The 9-18 does feature better lens contrast, leading to better color rendition.

One other comparison, just to drive the point home. I don’t own any micro-four-thirds’ prime lenses, so I can’t compare the 12-50 to a native prime. But I can compare it to a prime lens of another system: namely, the celebrated Pentax DA 15 Limited. Let’s look at the crops.

100% crop from the 12-50, at 12mm, f7.1:

And the 100% center crop from DA 15, shot on a Pentax K-5 at f8:

Now the 12-50 corner crop:

And the DA 15 corner crop:

As can be ascertained, there’s not much difference between these crops, at least in terms of resolution. Now I’m not suggesting that the M. Zukio 12-50 is as sharp as the Pentax DA 15 Limited. The Pentax K-5 has a more aggressive AA filter than the Olympus OM-D EM-5. This leads to softer raw images at default Lightroom settings. These examples are merely to demonstrate that complaints about the resolution of the 12-50 are misguided, based on over-emphasizing MTF scores and the how images look at a pixel level. Images are rarely displayed for others to see at 100% resolution.

To sum up: other than edges at the wide end, the 12-50 has plenty of sharpness. It’s no less sharp than lenses that are usually praised for their resolution, such as the Lumix 14-45 and the Olympus 9-18. In terms of practical output, the differences between the 12-50 and high-end prime lenses will not prove all that significant, at least in regards to sharpness. (In other words, the number one reason to use a prime lens is not for added resolution, but for fast aperture, better lens contrast, and flare control.)

Chromatic Aberration. Thom Hogan issues the following complaint about the 12-50:

At 12mm we get a highly visible lateral chromatic aberration presence, and at most apertures you'd ever use. That's with the in-camera correction active, so raw files can be laden with CA. I did note variable correction at 12mm between different camera bodies, even with the latest lens firmware installed. The 12mm chromatic aberration in the corners may exceed what some of the cameras can deal with. At longer focal lengths, the chromatic aberration mitigates to the point where it can be mostly ignored in JPEGs, though it is still highly present. More so than any other m4/3 lens I've used recently, I find myself doing post processing chromatic aberration correction with images shot with the 12-50mm.

To me, the only issue with chromatic aberrations is whether they are easily fixed in post. As far as I can tell, the aberrations in the 12-50 are in fact easily fixed in post. Indeed, I have not found them to be especially worse than other m4/3rd lenses. I have never understood why some people are so uptight about chromatic aberrations. Since they are easily fixed in post, they are not that big a deal. If you’re worried about chromatic aberrations because you’re shooting jpegs, then you’re not being entirely consistent. If IQ really is important to you, shoot raw and do a bit of post processing. It’s not that difficult.
Size of the lens. Believe it or not, some have complained about the size of the 12-50. Although it’s a fairly narrow lens, using 52mm filters, it’s longer than the Lumix 14-45 and Olympus’ collapsing zooms. The reason, however, for the 12-50’s length is that it is an internal zooming lens. It’s doesn’t extend like the 14-45 or collapse like the 9-18 or the Olympus 14-42 kit lenses. The reason Olympus choose this non-collapsing design is that they wished to weather seal the lens, and it’s easier to weather seal an internal zooming and focusing lens than a collapsible lens. In other words, it’s a trade off. Sealing a collapsible lens would have been more difficult, and hence more costly.
Price of the lens. One argument essayed against the 12-50 is that it underperforms at its price point. The MSRP of the 12-50 was $500. Currently, the lens sells for around $430. If it is purchased with EM-5, it costs $300. Anyone who claims the lens too expensive isn’t paying attention. The Nikon 16-85 f3.5-5.6 sells for $630 at amazon.com. The Canon 15-85 f3.5-5.6 sells at amazon for $699. These two lenses are a third of a stop faster on the long end than the Olympus zoom, yet they cost at least 60% more. Neither Canikon lens features weather sealing or a special macro mode. Could the Canikon zooms be appreciably sharper? I doubt it. But if they are, it would constitute an admission of the advantages of larger lenses using a larger sensor.
A more fair comparison is with other micro-four-third lenses. The only real rival to the 12-50 is the Lumix 14-45. At amazon, the lowest price for a new copy is $368. Granted, only a few months ago, you could get a new copy for around $275. But that’s not a whole lot less than the 12-50. Moreover, keep in mind that the 12-50 enjoys a longer range, features a macro mode, and is weather sealed. It’s a better all around lens and is therefore more expensive.
Lenses cost money to produce. According to some internet sources, Olympus may have lost as much as $190 million dollars last year. It’s unlikely that the company is reaping huge profits on the 12-50. So why all the whining about the cost of the 12-50? Do we wish to drive Olympus out of business?
Darkness or slowness of the lens. This is probably the biggest complaint against the 12-50. In fact, I suspect that it’s the number one reason for all the negativity that has arisen over this lens. Some photographers dislike slow lenses. If you need a faster lens, get a prime or an f2.8 zoom. The 12-50 is not a lens designed for those who demand fast lenses. With the exception of pics taken in macro mode, I could probably count on one hand the times I have shot the 12-50 wide open. In landscape photography, fast apertures aren’t required. The 12-50 may be slow. It’s not the best lens for either attaining narrow depth of field or for hand-held use in low light. But as a landscape/architecture lens, it’s plenty fast — and more than adequate.
Possible issues with the 12-50. I have found the M. Zukio 12-50 to be a fine lens, capable of producing excellent images. But many others online have reached a very different opinion. How can we account for this? I have four hypotheses:
  1. The slowness of the lens
  2. Cost of the lens
  3. Sample variation
  4. Over-reliance on numerical tests
  5. Shutter shock issues
  6. Confirmation bias
I’ve already commented on the first two issues. Let’s examine the second two.
Sample variation. While sample variation among lenses is a very real phenomenon, its degree and extent tends to be exaggerated. Unless a lens is actually defective, differences in sample variation should be insignificant. Yes, they exist, and if you pixel peep, you might notice them. In practical use they won’t be noticed and are not important. The wide differences that appear in MTF scores between various internet review sites are probably not the result of sample variation, but are rather manifestations of tester error.
Although I’m generally dubious about claims that sample variation can explain widely diverging opinions about a lens, in the case of the 12-50 I’m not so sure. It’s possible that there are a number of bad copies floating around. I’ve seen some 100% crops taken either at fast shutter speeds or on a tripod that don’t come close to matching what I have gotten from this lens. If you’re not getting reasonably sharp shots from the 12-50 and you’re confident that it is not a result of poor shooting technique, you may have a bad copy of the lens.
Over-reliance on numerical tests. Many photographers exaggerate the importance of numerical tests, such as MTF scores. The quality of a lens cannot be determined by how well it performs on resolution tests. While perceptual sharpness tends to correlate with MTF scores, that correlation is not 100%. I have two lenses, the Pentax K 28 f3.5 and the Pentax M 28 3.5, that score about the same on resolution tests; but images from the K 28 f3.5 simply look better (sharper, better contrast, better colors) than images from the M 28 f2.8. The IQ of a lens is not determined by how well it scores on numerical tests, but on how good the images it produces look to human perception. Photography is, at bottom, an aesthetic discipline; and aesthetic values cannot be quantified.
Shutter shock issues. This issue has been raised here. It’s possible that it’s affecting the performance of the 12-50. I find that shots taken in between 1/125 and 1/60 second sometimes seem a little less sharp than shots taken at faster or slower shutter speeds (although the difference is small and probably not significant). This is true even when a tripod is used. The difference is not very great. But it’s possible that some people are getting less resolution out of the 12-50 dues to shutter shock. The cure, which works in most but not all of the time, is to set the Antischock to 1/8 of a second.
Confirmation bias. Cognitive scientists have found that human beings are strongly prone to confirmation bias. If we become convinced that a lens is “mediocre,” this can affect how we judge images produced by the lens.
An example of this kind of bias is produced by the Pentax M 85 f2. The lens scored very low on resolutions tests performed by Modern Photography. These tests were confirmed by later tests. Not surprisingly, Pentaxians on internet continued to confirm these tests. The lens was decried as mediocre. Years later, a strange thing occurred. Some people, entirely ignorant of the resolution tests, bought used copy of the lens for cheap on ebay. They put the lens on their Pentax DSLRs and looked at its results with fresh eyes. They found a lens that, far from being “soft” or “mediocre,” was actually pretty darn good. The price of the used copies of the lens has more than doubled in the last few years, and it’s on the way to becoming a cult classic.
Don’t let the negativity surrounding the 12-50 cloud your judgment of the lens’ merits. A good copy of the 12-50, while not quite on the level of pro glass, is nonetheless a decent performer. It’s a good lens for outdoor use: versatile, weather-sealed, and sharp, capable of producing very nice images.
Here’s more pics from the Olympus M. Zukio 12-50 f3.5-6.3: