17-40mm lens on Canon 5D mark II

[evphotography]

<p>Please give me any feedback about this, let me know what you think. <br>

I recently purchased a Canon 5D mark II and after all the different postings on forums and lens tests I have seen on other sites I was convinced that my 17-40mm f/4L wasn't going to cut it. Because the resolution on the 21MP sensor is so good that it brings out all imperfections of the lens. So I have spent hours searching and checking out alternative lenses to use with an adapter, but was concerned about problems sometime associated with using those as well. Then I came across this article by Ken Rockwell "lens sharpness" and it kind of opened up a hole new light on the subject. Here is some of what he said.</p>

<p>"In real-world photography, natural factors do more to screw up a picture than any lack of lens sharpness. Sharpness tests are unlike real photos because a test does everything possible to eliminate any source of unsharpness. In a lab, nothing moves and the target is usually flat."<br>

<br>

Here is a link to his article it is worth reading I believe.<br>

<a href="http://www.kenrockwell.com/tech/lens-sharpness.htm">http://www.kenrockwell.com/tech/lens-sharpness.htm</a></p>

<p>So I decided to test his theory and low and behold he was correct. I staged a scene shot the scene at 17mm, 21mm, 24mm & 28mm. Shot it at f/8, f/11, f/13 & f/16. For each focal length and f stop used I focused on shot on foreground, middle ground & background. Used mirror lock up with 2 sec timer. Manually focused using live view at 100% to assure proper focusing and so back focus issues couldn't play a factor. When I compared the results, what I found was that the plane of focus and depth of field at different f stops had much more impact on lenses sharpness from center to corners than the limitations of the lens itself. When I found the correct plane of focus and f stop I could get great results in all the corners of my images. Much better results then any of other lens test or reviews I have seen. Where they staged the scene, shot each lens using one plane of focus only at different f stops. I have seen several tests comparing the 17-40mm to Carl Zeiss lenses, Nikon lenses and so on. My lens had much better performance than any of theirs did. Maybe I am lucky and just got a good copy from Canon, because I do believe that not all lenses that come off the production line are created equal.<br>

<br>

Don't get me wrong here, I'm not trying to say that 17-40mm lens is as good as Carl Zeiss lens, or other really high end lens, but just that when you learn how to get the most of your lens you can get excellent results. Also remember that all these lens test you see comparing this one to that one, are displayed from 100% crops, pixel peeping as they call it. When I also compared my test shots at 50% viewing size the differences where even much less, corner softness was lot less and overall results looked a lot better. Nobody makes prints that are as large as an image at 100% viewing size on computer screen, that would be at 52"x 78" print from my 5D mark II.</p>

<p>So probably wondering what is the point of this posting, which is that when proper technique is used and you select correct plane of focus and f stop, you can get great results even in the corners. I found that my lens at f/11 - f/16 was Excellent across the entire frame. The corners were indeed softer then the center of the image at 100% viewing, but when viewed at print size the difference was so small that you could hardly see the difference at all. There are limitations to every lens, just need to test the lens and find what the limitations are and you can get great results. But if you are truly a pixel peeper and are mostly concerned about how good image looks at 100% viewing, then indeed you need to spend the extra thousands of dollars for those expensive high end optics.</p>

<p>This test saved me a lot of money and trouble now that I know how to get most out of my 17-40L and not to worry if the corners are little soft when viewing it at 100% on my screen. Because in the end the prints will look awesome and you would have to look at even large prints like 16-24" very very close to see any imperfections of the lens.</p>

 

[bueh]

<p>Although I don't usually agree with K.R., he is right about this. Absolute sharpness is way overrated and worrying about is a sure sign of a photographer newbie. User mistakes like poor exposure, camera shake and misfocusing are MUCH more likely to ruin an image than not getting the highest possible resolution. Spend some time with classic cameras and you will see that sharpness rarely is the decisive factor in image quality. If the focus is correctly placed a picture will look good regardless of # of lpmms or MTF scores. I find focus errors and distortion much more distracting.</p>

<p>And if you have seen old large format glass plates or even actual results from wet plate or metal plate processes, you know it is pointless to be bothered with trying for absolute sharpness with modern gear: It is just not possible...</p>

[John Crowe]

<p>21 MP is only beginning to approach what the best 35mm films could resolve. The best films being Kodachrome 25, Ektar 25 and a few 25 ASA B&W films. At 25 to 30 MP digital cameras will likely squeeze everything out of the best current lenses available. Just like in days of superfine films some lenses are going to perform better than others, but we already could see that with 6 MP sensors as well. The best lenses of today and even from 30 years ago will still perform brilliantly on DSLRs for at least a couple more years yet. Once the megapixel wars plateau at 30 MP perhaps the sensor designers will finally turn their attention to improving high ISO performance, latitude, and sensor bloom. </p>

<p>At some point it will become too expensive to provide sharper glass for 24x36mm sensors and medium format will provide the next step in high resolution with full frame 56mmx68mm sensors which will require less enlargement and which will start to push the envelope of resolution of current and old medium format lenses in the same way as 35mm DSLRs.</p>

[mbkissel]

<p>Eric, when I first started taking photographs, I had no concept of depth of field, sharpness, ISO, etc. My work needs required that I take photos of small subjects to make them easy to see with the unaided eye (macro).</p>

<p>I would see photographs of circuit boards taken at about a 45 degree angle where every resistor, diode, chip, etc. was in perfect focus. My first attempt at doing the same resulted in a very sharp image of only half of the resistor closest to me, the rest were horribly blurred. What? What on earth had I done wrong? I justified my error by thinking it was an equipment issue. Those guys must have been using that super deluxe XBR-G2 5000 bi-focal wonderbox for those images. Yeah, that's it. That's how they did it.</p>

<p>Slowly I began to learn about photography. Stumbling along on my own, reading when I got time, and listening to others. Imagine the delight on my face when I finally developed that first photo, from my same old crappy camera and lens, and those resistors, diodes, chips and everything else were in sharp focus. </p>

<p>I continue to learn, and improve my techniques. It's folks like you and the hundreds of others here that make that possible. Thanks for writing your post, I enjoyed reading it, and for posting the link to Ken Rockwell's article. </p>

[evphotography]

<p>Your welcome Mark, I was hoping this post would maybe open the eyes to some beginner photographers that read those MTF charts and lens reviews and think they are gospel. I was little mad at myself for getting caught up in that when I first purchased the 5D II when I have known for long time, getting great shots is about the photographers ability to use his tools to their fullest and learn the limitations of your tools to get excellent results.</p>

[g dan mitchell]

<p>One thing... the 17-40 <em>is "</em> high end glass." It just costs less, has a smaller maximum aperture... and really excels at certain types of shooting.</p>

[steve santikarn]

<p>good lights and good techniques will beat good camera and good lens any day.</p>

[marco_hidalgo]

<p>I think the 17-40 is a bad lens for it's price.<br>

I think the Canon 35 f2 is a great lens for it's price.<br>

There are good lenses and bad lenses. I own these two lenses and I rather use the 35 f2 because my editors will pick the sharper images this lens produces, over the not so sharp images the 17-40 can produce at the same f stop. ( yes I have done my own tests with three dimensional objects and I have compared the results side by side and the 35 f2 wins hands down! Even on a not so large 8 x 10 print)<br>

I agree with some of Ken Rockwell's thoughts, but not all of them.</p>

[g dan mitchell]

<p>Marco, I think you are off base in calling the 17-40 a "bad lens." It is a wonderful performer that excels at some things, but like all lenses it is not the "be all, end all" lens.</p>

<p>I happen to use both the EF 35mm f2 and the EF 17-40mm f/4 L. Each has it strengths, and I'll choose one or the other depending upon what I'm trying to do. The 17-40 obviously provides a range of focal lengths that go all the way into the ultra-wide range on a full-frame camera. Stopped down it is a very sharp lens that produces photographs that stand up to printing at large sizes - certainly much larger than the 8 x 10 size you mention.</p>

<p>The 35mm f/2 shot the same way - e.g. at small apertures - can be a bit sharper. If the composition of a shot is right at 35mm and I have time, I'll choose the prime over the zoom for that reason alone. It also performs better at larger apertures - I generally avoid shooting the 17-40 at apertures larger than f/8, but I'd go considerably larger with the prime.</p>

<p>Perhaps you got a poor copy of the 17-40, or perhaps something else is up...</p>

<p>Dan</p>

[bobatkins]

<blockquote>

<p>...Because the resolution on the 21MP sensor is so good that it brings out all imperfections of the lens...</p>

</blockquote>

<p>The 40D, and even more so the 50D, have higher resolution sensors than the 5D MkII. Thus by this logic they would need even sharper lenses.</p>

[canfred]

<p>Having used my 5D II for a few weeks and comparing it with my 40D my advise is use faster lenses. F4 lenses work better on the 40D seems to me they do focus faster than on the 5D II, I have not used the 17-40 I am sure its a fine lens but a lens with f2.8 would be soo much faster to use. I only own two Canon lenses one is the sweet ( underated ) 50mm f1.4 the other one is the superb 24mm f1.4 L my statement is based on using the kit lens again a very fine and well build lens but in comparising to focussing speed on my 40D its quite a bid slower.</p>

<p> </p>

[marco_hidalgo]

<p>Dan: I guess you are right. I am just a bit annoyed at the 17-40 not performing well enough at f4 and f5.6. I have to stop it down to get good enough quality from it and I was expecting a bit more from an L designated lens. At it's widest apertures the 17 - 24 used at f8 - f11 performs really well indeed. But then again the 35 f2 outperforms it at any given f stop.<br>

The Canon 20 f2.8 is better than the 17-40 when used from f2.8 to f5.6 ( of course 2.8 doesn't really count as a comparison ), But I consider the quality of the 20mm. quite usable wide open, but not the quality if the 17-40 wide open. Another lens that shows great quality wide open is the Leica Summilux M 35 f1.4 Asph. Which is indeed outstanding! I have never been disappointed about the Leica lens and I paid a lot more for it, than what I did for the Canon 17-40. That is why I do not like the 17-40. Not a good lens wide open.</p>

<p> </p>

[james_glucksman1]

<p>I am not an expert by any stretch of the imagination, but I used to have the 17-40 f/4L and used it on my old 20D and then on my 5D. I recently got a 5DMkII though and I started to be dissatisfied with the 17-40 so I traded it in towards a 16-35 f/2.8LII, and so far I am pleased with the results. I don't think there is really anything all that wrong with the 17-40, but I found that the photos it produced did not have the same satisfying contrast that I perceive with the 16-35. Perhaps this is my brain trying to convince my wallet that the expense was worthwhile, but that's my story and I'm sticking with it!</p>

[sbp]

<p>FWIW, i use the 17-40 f/4L on 40D, 1D3, and 1Ds3 bodies. All give good results. While i, too, don't always agree with Ken Rockwell, he is on point regarding sharpness. </p>

[robin_sibson1]

 

<blockquote>"The 40D, and even more so the 50D, have higher resolution sensors than the 5D MkII. Thus by this logic they would need even sharper lenses."</blockquote>

<p>I see no fault in the logic, but it is important to add that for a 1.6-factor sensor the requirement is limited to a smaller area.</p>

<p>The OP's (or KR's) point about sharpness is of course a good one when you are photographing a subject not far from the camera, which will almost certainly be "three-dimensional". The same point arises in a different way with close-up/macro work, where it can be worth using a very small aperture that sacrifices some resolution in the plane of focus (because of diffraction effects) in order to gain more depth of field. However, for landscape work, the main subject matter is in many cases all so far away that issues of "three dimensionality" don't arise in terms of focusing - you want everything "at infinity" to be sharp. For obvious reasons, simple lab tests of lenses are not well-suited for assessing sharpness at infinity, and the seemingly more "amateurish" process of setting the camera up on a tripod an photographing distant skyscrapers may be more informative. </p>

 

[evphotography]

<p>As a few have stated here, you are correct it doesn't perform nearly as good wide open, but that was one of the points I was trying to make. There are very few if any lenses out that do perform excellent from it's widest aperture to it's smallest. Learn what the limitations of the lens is and use it to it's strengths and you will get excellent results.</p>

<p>I don't think the 17-40 was designed to be shot wide open, it's not a very fast lenses and therefore not really designed for low light action type photography. Another thing I didn't really mention in my original post was just how well it performed at f/16. It performed almost as well at that aperture as it did f/11 (sweet spot) and better than f/8. Which for a lot of other lenses, including some expensive high end optics, diffraction at f/16 causes the entire image to be a lot softer compared to the 17-40L, making this a great landscape lens. I have also read a lot about how it doesn't perform nearly as well at long end 35-40mm, but again that is another limitation that once you are aware of, use it to it's strengths. Where as the 16-35L starts to really sing at the longer end, that is that lenses limitation, not as good on the wide end. I borrowed a friends 35mm f/2 and I am going out today and see how it stacks up against the 17-40mm at that focal length.</p>

[kerry_grim]

<p>I have the 40D, 17-40 and 35 f2. I've been pleased with these lenses judging by my photos, so I am NOT going to compare sharpness between the two lenses. The 35 may be sharper by my assumption only. I use the 17-40 on a tripod at middle apertures. To compare them and look for the worst from one of them would be a mistake and likely I would shelve it or sell one of them. Look for the ways to get the most out of a lens (accurate focus, tripod, etc.) and not worry about lack of sharpness is the way to go. The 17-40 is now my landscape lens while the 35 has become my indoors lens.</p>

[g dan mitchell]

<p>This thread highlights something that many have already figured out: lenses have "personalities" that make them more or less suited for specific types of usage. Let me use the EF 1-40 as an example of how interesting and important this sort of understanding can be.</p>

<p>My main photographic interests include natural and urban landscapes, but I also do some street photography and a few other things from time to time. I first used the 17-40 on a 8MP 1.6x cropped sensor body. Here it was OK for the landscape-type work, though there were issues. While it is very sharp in the center of the frame, it is soft in the corners at larger apertures. The issue here for landscape on the cropped sensor body is that I was somewhat stuck between the proverbial "rock and a hard place" - and more or less left with f/8 as the only nearly ideal aperture. Larger apertures increased the corner softness, and smaller apertures increased the diffraction blur. In the end I often shot the lens at f/8 but assumed that I might have to crop a bit on any shots in which the farthest corners needed to be sharp. (In many cases the corners are less important on very wide shots and I did not need to crop.)</p>

<p>Before long I moved to a 5D full frame body. Here the corner softness is a non-issue for landscape type shots since I can now shoot at f/11 or f/16 w/o concern about diffraction. (Yes, there is a very tiny bit of diffraction blur visible in f/16 shots if I compare them to f/11 shots side-by-side at 1005 magnification on the screen. But in prints? Not visible at all.) At these smaller apertures the image has good uniformity across the frame and this is a fine lens for this sort of shooting.</p>

<p>At f/4 (and to some extent at f/5.6) the corners are less sharp and vignetting is more visible. (By the way, vignetting is present on essentially all lenses including some very excellent lenses.) If I were shooting wide open all the time and needing the best possible corner performance I would think seriously about the EF 16-35mm f/2.8 L. On the other hand, I have used the 17-40 at f/4 for street photography in some very confined areas, and I've had some success. Here I'm frequently shooting "on the move" and my subjects are also in motion. Corner sharpness is not a concern, and the vignetting effect is often just fine. So, guess what, it works well as a street photography lens - perhaps to my surprise.</p>

<p>I could tell a similar story about almost any of my lenses: they are great at some things and less great at others. As long as they are great at what I need them to be great at, they are the "best" lenses for me.</p>

<p>The key is not to look for "the best lens on the market" but rather to look for the lenses whose characteristics best match your subjects and style of shooting.</p>

<p>Dan</p>

[felix_mizioznikov]

<p>I have the 17-40 and from playing with it on both the 50d and the 5d2 i'd give it a 7 of 10. It's the lens that spends the most amount of time on my camera than any other. I primarily shoot for stock and sharpness is of utmost importance and many images get rejected due to incorrect of lack of focus. Actually I believe that stock agencies are the only ones that peek at %100 percent and make determinations as to weather to accept or reject an image for sale so for me all the charts do matter and the sharper the lens with 0 sharpness set on the camera, the better! Also most of the pics I take place the subject away from the center i.e. rule of thirds to leave space for advertising so I need the edges to be as clear as the center. I dont think Ken has any regards to stock photography in his rant.</p>

[marco_hidalgo]

<p>"The key is not to look for "the best lens on the market" but rather to look for the lenses whose characteristics best match your subjects and style of shooting."<br>

Dan: You have put it very well, I guess I was having unrealistic expectations about the 17-40.<br>

<br /><br>

Felix: I also think K.R. didn't think about stock or editorial. I shoot for editorial and I always try to leave space for big headings and text, thus using a lot of the borders of the image. I never know when my editors might need a vertical shot with lots of space on the upper half or a double spread where the least important part of the image should be placed in the middle of the frame. I hate the gutter to interfere with any key element of the image. So, I do as well use the thirds rule and try to place the main subjects to the sides.</p>

[stephen_asprey2]

<p>There is a lot of wisdom in the gathering opinion that unless you are going to print a wall sized image, the sensor size is less important than lens quality. Look at the results you can get from a D700 for instance (or from the old 5D).<br>

One can consider this: As the major manufacturers go for bigger sensors, then the resale value of the superseded models go through the floor. Now that we are at almost affordable FX, all that beautiful glass that has underpinned Nikon and Canon (to a lesser extent) for decades can be used again. We are now seeing eBay prices for fast Nikon primes exceeding their original retail prices in some cases. Leica have always chuckled at this phenomena for years.<br>

I am putting my money on the best fast, quality primes in the focal ranges that I want, and at 10mp I can produce an A3 print that is almost perfect.<br>

So try not to get on the upgrade bandwagon that the big manufacturers want you to do.</p>

[hawkman]

<p>Eric,<br>

I used this lens for about 3 years as my main setup with 20D which has the same pixel density as 5D MKII, the center sharpness of this lens is very good at f/5.6 and beyond however it is is soft at f/4 and there is significant fringing and softness in the corners. 5DII will accentuate the problems since it is looking at the full image circle. Despite being f/4 this lens is only 130g lighter than the 16-35 f/2.8L MKII and exterior dimensions are identical except for larger 82mm thread for the 16-35. I would not recommend this lens for critical work on 5D MKII unless you only shoot at f/8 and beyond at all times, a 16-35 f/2.8 MKII is a better choice.</p>

[dennis_d._hardenburger]

<p>Art Wolfe always takes the 17-40L lens with him, I think this says a lot about the lens.<br>

Art Wolfe's main kit is the 17-40L, 70-200 f4 and the 500 f4<br>

I wish Nikon had an equivalent lens to the 17-40 plus a 70-200 f4</p>

[studiox]

<p>Who is art Wolfe?</p>

[dmanthree]

<p>Google is youf friend: <a href="http://www.artwolfe.com/">http://www.artwolfe.com/</a></p>