gdanmitchell
Online
• • • • • • •
Upload & Sell: Off
|
 p.3 #4 · Photozone Review: Canon EF 16-35mm f/4L IS | |
Roger, I think there is a lot of interesting and important stuff in your post.
Your reminder that no lens is perfect is an important one for everyone to keep in mind. We want lenses that are as good as possible for our purposes, in the range of ways that "goodness" can be measured, but it isn't useful to hold out for "perfect" lenses.
I got my copy of this lens nearly two weeks ago and finally had a chance to "field test" (otherwise known as "use it to make photographs") late last week. (I posted a report on that experience here. Keep in mind that it is a "report," and not a "test.") My impression of the lens in use limited to the way I will typically use the lens in my photography is that the purchase was a good one. As you (Roger) note, it seems to be better than the existing Canon ultra wide zoom alternatives.
Regarding copy variation, a story. A photographer friend of mine who often shoots subjects similar to many that I photograph got his copy at almost the same time I did. Before our lenses arrived he had made a joke about wondering which one of us would get the "good copy." It looks like I did. We have corresponded since he received his, and he has shared some photos he made while testing the lens, and he definitely has one very fuzzy corner on his copy. Think "softer than what you would expect from a 17-40 at f/8" and "with the same sudden drop off in sharpness very near the corner that you might see on the old EF 35mm f/2."
Since he wondered how mine performed, I went to the trouble of setting the camera/lens on a tripod in front of a grasscloth wall, lining everything up reasonably well, and making some informal test photographs at a range of apertures and focal lengths. The tests confirmed that mine is a "good copy," but also showed that, as expected, things are not uniform at all possible settings. As many have written, it is remarkably good all the way open at f/4. In 100% magnification crops I can see that the far corners on one side of the frame are a bit softer than the other, but I have to go looking for this. I can also see that in all situations the center seems at least as sharp as my 17-40 and the corners are noticeably better.
So, it will be interesting to see what you eventually figure out about these issues, Roger:
- the range and nature of variations among copies of the 16-35mm f/4 L IS
- how that variation compares to variations among copies of the 16-35mm f/2.8 L and the 17-40mm f/4 L
- how the variation among Canon ultra wide zooms as a breed compares to that among similar lenses from other manufacturers.
Two final things.
1. I'm very happy with my copy of the 16-35mm f/4 L IS. It is better at a wider range of apertures than my 17-40mm f/4.
2. Your comment about getting Canon to recenter lenses ("As to Canon recentering them, I wouldn't put a lot of hope in that. And I speak from over 100 copies of the 16-35 and 17-40 sent in to Canon for recentering.") brings to mind the story of a person I know, whose ultra wide zoom lens Canon would most likely want to bring into excellent adjustment (for a range of reasons that I won't enumerate, since the person prefers anonymity), who eventually came to the conclusion that this wasn't possible with the ultrawides...
Dan
RCicala wrote:
I haven't had enough stock to tear one apart yet, although I've been testing them for a couple of weeks now.
Bottom line on testing was the copy variation was greater than I expected: not that any were awful but each and every one had a weak point. So after testing 10 of them, I decided I needed to repeat the same tests with 17-40s and 16-35 f/2.8 II because we've never tested wide zooms on the optical bench before. And then repeat them all on Imatest because that's what we usually test on (although I hadn't planned on it, since with Imatest we're testing so close to MFD at 16mm).
It's been a complex problem, but the bottom line is the 16-35 f/4 IS is better than either the 17-40 or the 16-35 f/2.8 II shot at f/4 - at least once you're off center. The difference is more appreciable at infinity than at very close Imatest distances (which is what I've thought would be the case all along).
But the amount of copy to copy variation is noticeable. Not just with this lens but with the 17-40 and 16-35 f/2.8 II when I test to these standars. So far out of 10 of each there is none, repeat none, that don't have at least one softer corner at one end of the zoom range. I guess it's just the limitations of designing a zoom that wide.
The good news is the soft corner of a 16-35 f/4 IS is sharper than the sharp corner of a 17-40 or 16-35 f/2.8.
As to Canon recentering them, I wouldn't put a lot of hope in that. And I speak from over 100 copies of the 16-35 and 17-40 sent in to Canon for recentering. There's a reason we spent a small fortune on optical testing equipment (other than that I just wanted to) and spent a few hundred hours learning how to do it ourselves. It's not a money saver -- the investment in equipment would take about 4,000 repairs to pay for itself. It's just the only way we can get it done right. ...Show more →
Edited on Jul 17, 2014 at 03:24 PM · View previous versions
|
