| Articles |


Lens Comparison: Konica Minolta 18-70mm f/3.5-5.6 DT vs.17-35mm f/2.8-4 D

I bought the Konica Minolta 5D and 18-70mm kit lens from an online vendor in February of 2006 and almost immediately discovered that my lens was a much better performer than I expected it to be. Soon afterwards I visited a camera shop and informally compared the 18-70mm to an earlier generation KM lens - the 17-35mm f/2.8-4. At the time, the 18-70mm cost me about $130 over the price of the 5D body alone, and I recall that the 17-35mm was selling for something like $450. I took just two shots of the interior of the camera shop at f/5 and at the widest setting for each lens.

When I got home and reviewed the results, it was easy to see a difference between the two. While both images looked good in the center of the frame, there was a pronounced difference in corner performance. The 18-70mm was dramatically sharper there, even though it also showed somewhat heavier CA (chromatic aberrations). To my eye, the 18-70mm easily beat the 17-35mm, at least at this aperture and focal length. I assumed that the mediocre performance of the 17-35mm was due to the fact that it was originally designed for full-frame film cameras, not DSLRs. I wasn't really motivated to pursue the comparison any further, but was content concluding that the humble kit lens was actually an amazing bargain.

 

Here is the full scene taken in by the 17-35mm. Below that are 100% center and corner crops from both lenses.

18-70mm (18mm) at f/5:

17-35mm (17mm) at f/5:

18-70mm (18mm) at f/5:

17-35mm (17mm) at f/5:

 As the next several months passed, I read numerous reports from other 18-70mm owners. Many of them were as impressed as I was, but others were unhappy with the lens, citing poor performance. Eventually the consensus emerged that sample variation must be to blame: properly built specimens of the lens were quite good, but a significant percentage of others were substandard.

A New Test

Sixteen months later, not much has changed. I'm still happy with the lens and I continue to recommend it as a great value. At the same time, others out there still complain about unsatisfactory performance. I decided that I needed to conduct another test against the 17-35mm to be sure of my position. Why the 17-35mm? Because it's currently the least expensive alternative from KM/Sony that provides real wideangle coverage (less than 20mm) on a DSLR. The other options are the KM 17-35mm G, the KM or Sony 18-200mm, and the Sony/Zeiss 16-80mm, all of which I regard as too pricey for what they offer. Also, the 17-35mm is almost unique in that it is full-frame capable. The G version of the lens is full-frame as well, but that one typically costs a heck of a lot more: about $1,600. (Incidentally, there's a very interesting article comparing the 18-70mm and that expensive 17-35mm G lens at Michael Hohner's site.)

Anyway, I recently obtained a like-new 17-35mm at an attractive price and this time I did much more exhaustive testing.

(Sneak peek at the conclusion: I was right the first time.)

Here's a scene with main subject distances of less than 20 feet. This is similar to the distances involved in the original camera shop photo. I only tested the widest setting of each lens because to me, the wide end is the most important and useful aspect of these lenses - especially for near subject work, as in interior shots.

These are 100% crop comparisons showing center performance. The distance from the camera to the white ceramic figure (where the lens was autofocused) is 16 feet. The 17-35mm at f/2.8 appears by itself at the right. It looks good. And all of the other shots with both lenses at all apertures look good as well. In fact, I can't distinguish any differences among them (except for the slightly greater angle of coverage with the 17-35mm, which is of course normal). If I didn't already know, I would never be able to identify which apertures were used. Suffice it to say that the center performance of both lenses is impressive.

You might also notice a difference in color rendition. The 18-70mm tends toward blue while the 17-35mm tends toward green. The same color tendencies appear in all three test environments on this page. I have found that both of my Tamron-made KM lenses (the 17-35mm and the 11-18mm) share the same greenish color cast, while my other earlier design KM lenses (50mm macro and 75-300mm) match the color of the 18-70mm.

17-35mm (17mm) at f/2.8:

18-70mm (18mm) at f/3.5:

17-35mm (17mm) at f/3.5:

18-70mm (18mm) at f/5.6:

17-35mm (17mm) at f/5.6:

18-70mm (18mm) at f/8:

17-35mm (17mm) at f/8:

 

Now let's look at 100% crop comparisons showing corner performance. The distance from the camera to the backboard is 16 feet, the same subject distance as in the examples above. Once again, the 17-35mm at f/2.8 appears by itself at the right. But this time the story is different - it's quite smeary at maximum aperture. It gets better when closed down to f/3.5, but still not as good as the 18-70mm is wide open. Study the repeating pattern on the backboard to see the difference. At f/5.6 and f/8 both lenses start looking about equal in overall performance. To characterize the differences, the 17-35mm shows poorer sharpness, while the 18-70mm shows more extreme CA and purple fringing. In practice, the latter problems are usually easier to correct via software than the former is, so the 18-70mm can be considered to have the advantage here.

By the way, I use and highly recommend the inexpensive PTLens program for easily correcting CA, distortion, and other issues with a wide range of lenses.

17-35mm (17mm) at f/2.8:

18-70mm (18mm) at f/3.5:

17-35mm (17mm) at f/3.5:

18-70mm (18mm) at f/5.6:

17-35mm (17mm) at f/5.6:

18-70mm (18mm) at f/8:

17-35mm (17mm) at f/8:

 Another Test Environment

I decided I should do another round of testing in a different environment at distances approaching infinity. This one also includes tests at three different focal lengths. I tested each lens at the maximum common aperture and at f/8.

The full scene taken in by the 17-35mm, followed by 100% center crops.

18-70mm (18mm) at f/3.5:

17-35mm (17mm) at f/3.5:

18-70mm (18mm) at f/8:

17-35mm (17mm) at f/8:

18-70mm (24mm) at f/4.5:

17-35mm (24mm) at f/4.5:

18-70mm (24mm) at f/8:

17-35mm (24mm) at f/8:

18-70mm (35mm) at f/5.6:

17-35mm (35mm) at f/5.6:

18-70mm (35mm) at f/8:

17-35mm (35mm) at f/8:

Once again, all of those center crops at all apertures and focal lengths look fine.

And now, the 100% corner crops:

18-70mm (18mm) at f/3.5:

17-35mm (17mm) at f/3.5:

18-70mm (18mm) at f/8:

17-35mm (17mm) at f/8:

18-70mm (24mm) at f/4.5:

17-35mm (24mm) at f/4.5:

18-70mm (24mm) at f/8:

17-35mm (24mm) at f/8:

18-70mm (35mm) at f/5.6:

17-35mm (35mm) at f/5.6:

18-70mm (35mm) at f/8:

17-35mm (35mm) at f/8:

Again the 17-35mm goes smeary in the corners at large apertures, particularly near the wide end. Stopping down to f/8 is the only way to clean up those corners.

Conclusion

If you've followed along this far, you can see why I say that the 18-70mm generally equals or beats the 17-35mm at any aperture or focal length on APS-C format. And let's not forget that the lens also goes all the way to 70mm, though I ignored that capability in these tests.

I should also comment on distortion, vignetting, and flare control. Although I don't show specific examples of it, rectilinear distortion is essentially the same with both lenses - it's visible enough to be objectionable at the wide end, but it virtually disappears by 35mm. (You can see some bending of the basketball goal post and door frame in the full backyard shot above.) Such distortion is easily correctable using PTLens, which includes both lenses in its database.

Vignetting is not noticeable in either lens.

There is a visible difference in flare characteristics, which you can see in the examples at the right. When the sun is in the frame, the 18-70mm usually shows a small number of large, diffused spots while the 17-35mm typically produces a higher number of tiny, well-defined spots. I happen to prefer the way the 17-35mm looks; but since this is a subjective matter it's difficult to declare a winner.

18-70mm (18mm) at f/8:

17-35mm (17mm) at f/8:

To summarize:

18-70mm advantages:

17-35mm advantages:

1. Costs hundreds of dollars less (approximately $70 vs. $270 in today's used market)
2. Same or better image quality at all focal lengths and apertures
3. Twice the zoom range (although the long end is not as good as the wide end)
4. Greater macro magnification with focal lengths beyond 35mm (the maximum magnification is approximately 1:3.3 vs. 1:5)
5. Smaller in size and weight (but plastic mount and parts)
6. Smaller filter size (55mm vs. 77mm)

1. Full-frame coverage (probably the most important reason to have this lens, since KM/Sony have never offered a less expensive option for a full-frame superwide zoom)
2. One-half to one stop faster apertures throughout its range (but at the expense of image quality)
3. Focusing distance scale
4. Non-rotating front element with internal focusing
5. Closer minimum focusing distance (30cm vs. 38cm)
6. Much nicer and more robust build (metal mount and parts)

Text and images 2007-2012 Ray Lemieux (except for the lens photos at the top of the page, which are all over the web and presumably courtesy of Konica Minolta)