Why am I not thrilled with the 17-35 f2.8?

Discussion in 'Nikon' started by steve_phillipps, Feb 27, 2010.

  1. I've had the AF-S 17-35 2.8 for a few years but never used it that much. The silent wave motor packed in (and will cost £430 to repair!!!) but that's not a big deal for a wide zoom.
    Thing is I've never been too thrilled by the performance. Have I got a bad sample, and is there anything obvious that could be out of alignment etc.?
    For example it's nowhere near as good as the 50 1.8af at either f2.8 or f4 - is this a surprise?
    Steve
     
  2. I've thoroughly enjoyed this lens on a DX format. The 50mm f/1.8 is one of the sharpest and most perfect color and contrast controlled lenses you can buy (though inexpensive). So the comparison between a sweet prime and an aging zoom is not apples to apples. But you possibly have a bad sample. I've experienced this with having owned many 60mm Micros before I got a good one, and the same with a 50mm f/1.2.
     
  3. ShunCheung

    ShunCheung Administrator

    I have only used one sample of the 17-35mm/f2.8 AF-S, mine. I think it is an excellent lens. I have pointed out a few times that I have tested 6 different copies of the 24-70mm/f2.8 AF-S; they are so consistent that I tend to dismiss those discussion about bad samples.
    I have no idea why Steve is not trilled with his lens. However,
    For example it's nowhere near as good as the 50 1.8af at either f2.8 or f4 - is this a surprise?​
    Comparing the 17-35 to a 50mm/f1.8 is totally meaningless. There isn't any overlap between their focal length range.
     
  4. I've used only one sample of the lens and it was beyond exceptional.
    What exactly is wrong with it's performance?
     
  5. Shun, I tested the 50 and the 17-35 with the subject covering the same amount of the frame (ie moving the camera further back for the 50mm shot). It's not exactly the same of course but it's a long way from meaningless - you can still judge relative sharpness.
    Peter, it's just not as sharp as I thought it should be considering all the amazing things I've heard about it (including in your post!)
    Here's the test shots.
    Steve
    00VsPR-224365584.jpg
     
  6. Here's the 50
    00VsPX-224365784.jpg
     
  7. And here's a real world example. Think it was 1/125 f4.
    Am I just expecting too much?
    Steve
    00VsPd-224367584.jpg
     
  8. What exactly is wrong with it's performance?
    If we were talking about cars, I'd suggest looking at the "nut" behind the wheel. What I mean to say is that technique is the limiting factor for the performance of any lens that hasn't been bounced down the stairs - careful composition and focus, a good tripod and smooth release.
    00VsPw-224371584.jpg
     
  9. It is one of my 'always carry' lenses it's so good.
    Conni
     
  10. Edward, it can't be bad technique when you've got a locked-off shot with 2 lenses and one looks OK but the other one doesn't!
    And a lens doesn't have to be "bounced down the stairs" to have issues. People have lenses brand new out of the box that are out of alignment etc. and not performing anything like as good as they should. Just trying to find out whether mine has some of problem or not. Any more useful suggestions to contribute?

    Steve
     
  11. ShunCheung

    ShunCheung Administrator

    Steve, on the image you posted at 3:37pm, the subject seems to be a DVD type box. At 35mm, my estimate is that you captured that image from about 2 feet away from the subject. The 17-35mm/f2.8 AF-S is not a macro lens. I wouldn't expect a wide angle to perform that well from such a close distance. When you used the 50mm, naturally you were farther away from the subject. That was why I said this type of comparison is unfair and meaningless.
    Your other image captured inside a gym seems to be with a D300S based on the EXIF data. The focus seems to be at/near infinity and the far away background looks quite good. However, the people holding the British flag were much closer to the camera and that is why they are somewhat out of focus.
    So far, there is no reason for me to believe that there is anything wrong with your lens.
     
  12. No, the focus was on area around the flag.
    The shots of the box were at about 3 and 4.5 feet, both what I'd call medium close, so a fair-ish test. But the reason I did it is that I wasn't that thrilled with it on more distant shots either.
    Steve
     
  13. With my FX body (D700), it always performs excellently and I love it and carry it all the time together with 24-70. Those two are, for me, "must" lenses for my FX body. Maybe your example needs a bit of AF Fine Tune adjustment?
     
  14. Thanks Ken, that has worked for my 85 1.4, but as I said the AF doesnt work on my 17-35!
    I've never been a massive zoom fan, was thinking of maybe getting a 35 prime, maybe the 1.4 or maybe the Zeiss f2. Also tempted by the 21mm f2.8 Zeiss, pricey though.
    Steve
     
  15. AF Fine Tune doesn't work perfectly for zoom lenses, you are right. If you so feel, 35mm 1.4 AIS is also a wonderful lens, I have one and I can vouch, as long as you won't miss wider than 35mm..... But then again, if you go that way, AF 35mm f/2 is much less pricey and you also can get 24 or 28 primes with the similar budget.
    As for me, I am happy with my several-year-old 17-35 and its convenience as a zoom is sometimes priceless on walkabout. YMMV.
     
  16. The lens clearly has something wrong with it in that it won't AF, even if we're all trying to figure out if it's soft or not. I'd send it in for repair and re-alignment, myself. Then if I still didn't like it, it would be an easier sell than if it were broken.
     
  17. I don't know, I use it on both Dx and Fx and it's been a great lens. Now with the 24-80 I keep that on the D700 and often use the 17-35 on the D200.
     
  18. ShunCheung

    ShunCheung Administrator

    Steve, I have a few questions and comments on your test images:
    1. The entire frame from a D3 should have 4256 x 2832 pixels. Somehow your first two samples have 4272 x 2828 pixels. Have those images been edited and do they represent the entire original frame out of the camera?
    2. The EXIF data indicated you used 1/10 sec and 1/15 sec without flash. That is a major no no for any lens test. 1/15 sec is the shutter speed that is most prone to vibration due to the mirror slap and shutter opening. Were you using a tripod with mirror lock up or exposure delay?
    3. For the indoor gym image, the problem is simply wrong focusing. The focus is definitely not on the flag since none of the girls on either side of the flag is in focus. However, all the lettering in the background looks quite good.
     
  19. When you have the repair done to the autofocus motor in the lens, the repair service can also check the element alignment and adjust the accuracy of the autofocus.
    The lens should be dead on when you get it back.
    Manual focus in the finder of a crop camera like the D300 is not easy to do with precision. I should think that you want to know both that the lens is back to factory spec on sharpness and element alignment and to have the convenience and efficiency of being able to autofocus with one of the best autofocus cameras on the planet earth.
    I'd say to fix it. Test it. Enjoy it.
     
  20. << I've had the AF-S 17-35 2.8 for a few years but never used it that much. >>
    I think the key is to just go and use it, instead of testing with made up scenarios. It is worth a fix, even if to sell it at the end.
    I used it a lot when I shot film. I always felt safe when I had this lens on the camera - because I trusted it enough to know that if the results were lacking, it would have to do with something other than the lens. Some of the images shot with this lens had won awards. I am not using it much these days only because the range overlaps with the 24-70 and other wide angle DX zooms.
     
  21. even on a tripod, it's just hard to do a test at 1/10 and 1/15 sec without flash. these variables result to vibration due to mirror slap as shun mentioned.
    in the real-world sample, the letters on the banner on the officials' table are sharp. even with huge misalignment of the elements of the lens, if the focus was on the flag, it wouldn't be that off.
    and i'm afraid to ask more simple focusing questions. but then again, i'm biased to nikon lenses not having that big of a problem....................but you know also sometimes it's hard to train the eyes of a purist prime person. steve did say he likes and prefer primes :)
    btw, i still have mine and still happy with it. i bought it a long time ago for my D70 when i first ventured into shooting weddings and other events. proved to be short so it's a studio lens now. i hope you get the problem solved. it's a great lens.
     
  22. 430 pounds, $656 ...??? Thats ridiculous. Are they installing a solid gold motor or something ? I assume you are in England as you have quoted in UK pounds. I would investigate other repair centres abroad, or even back to Japan if I were you. At that price (is the VAT included or is that more on top ?) is it worth it ? As you are not happy with the lens anyways, you are better off selling it 'as is', with the faulty motor and getting something for it. You can buy another good second hand one for about the price of the repair, and you'll be better off pocket-wise because you got some return on the faulty one.
     
  23. Thanks for the replies, much appreciated.
    Shun, I've no idea why they are not showing all the pixels. They haven't been cropped, just exported from Aperture. As for focus, why would I have focussed on the background? I do know a bit about focussing, I've been a professional wildlife cameraman for BBC and others for the last 15 years using lenses equivalent to 2500mm on a daily basis, all manual focus, so I'm kind of OK at it!
    As for the test shots and shutter speed/vibration, that's a red herring. If it was a single shot with one lens then yes it could be a factor, but they are side by sides and only one lens is soft and yet the shutter speed is the same, and this happened every time - it would be a massive coincidence if all the shots on one lens were soft due to shutter speed and all the others were OK!
    David, Peter, I could have it repaired but at £430 it seems a bit insane. Peter, I'm using it mostly on the D3.
    Mary, I'm only testing it 'cos I wasn't that thrilled with it in real world use.
    Steve
     
  24. I think you are being quite unfair here. This lens is a "hall of famer" from Nikon. Mine is as good as it gets for a lens. I would say the issue is the silent wave motor which could quite possibly be the issue giving you soft focus. Did you buy it new? If not it may have been dropped or knocked prior to you owning or even while you owned it.
    Please dont denigrate lenses or cameras, especially if they are on the way to the repair shop. BTW I agree with Shun, there is no comparison with 50mm 1.8, different focal lengths. In my experience, the 17-35 is as sharp as it gets, MTF charts rate this as the sharpest of the Nikon WA zooms.
     
  25. ShunCheung

    ShunCheung Administrator

    Steve, as long as your images represent the entire frame, that is fine.

    Now, I have captured the attached image to simulate your test shot. I taped a DVD box onto the wall. I picked this particular DVD box because it is all cardboard; there is no transparent protection plastic on top that might create glare. My box occupies roughly the same area in the frame as yours. My image is indeed 2832 x 4256, which is what is expected from the D3/D700 sensor.

    I put my 17-35mm/f2.8 AF-S on my D700, which is has the same sensor and electronics as the D3. I set the lens to 35mm, f4 also, and I used live view on the D700 to carefully focus manually around the area with all the words, e.g. around the name "Thomas Hampson" a bit to the lower right of the center. Since my sensor plane wasn't entirely parallel to the subject, the rest of the frame might not be as sharp. To line that all up from such a close distance would have taken a lot of time.

    My lens indicated that the subject is a tiny bit more than 2 feet away, but certainly less than 0.7 meter. That is why I think your claim that your subject distance being 4, 5 feet is way off. 2 feet is too close to expect great results from a regular lens that is not optimized for macro. Still, I think my image is very sharp around the area I focused to.

    Again, my settings are also 35mm, f4. However, I focused manually with live view. My primary source of light was an SB-900 and my shutter speed was 1/250 sec with little ambient light. The flash essentially freezes all vibration, eliminating that concern.

    I think even at a very unfair distance of 2 feet, my 17-35mm/f2.8 AF-S still generates very fine results at 35mm, f4. If you cannot achieve such results with yours, either there is something wrong with your lens or there is something wrong with your test methodology. Unfortunately, it is known that AF-S motors will fail over time. If you need to get the AF motor fixed anyway, you might as well let the technician calibrate your lens.
    00VseF-224519584.jpg
     
  26. FWIW, I had the 17-35, and as far as zoom lenses go, it was good. Take it for what it is...a fast zoom lens with good build quality. If you really want to pixel peep, then get primes.
     
  27. Peter, you're quite possibly right, there may be a problem with it, that's why I was asking about it. I'm trying to denigrate anything, and it's precisely because it's a "hall of famer" that I was wondering if I should be more blown away by it than I am. The 200 f2 is another hall of famer and to me lives upto the reputation, it's the best lens I've ever owned quite probably.
    Shun, thanks for taking the time to do the test, don't knwo about you but I do think yours looks sharper and clearer than mine. Might see if I can get hold of a 35mm prime to check it against.
    Incidentally it seems to look quite a bit better at 17mm.
    I realise I'm not testing quite methodically and extensively enough, just interested to get some thoughts, so thanks again.
    Steve
    00Vsf1-224527584.jpg
     
  28. ShunCheung

    ShunCheung Administrator

    Steve, when I had that set up, I also tested with my 35mm/f1.4 AI-S at f4. Of course that too was from 2 feet away, which is highly unfair for that lens also. I would only test a lens from that close if it is a macro. I also tried it with my 28-70mm/f2.8 zoom at 50mm. With a 50mm, the subject is considerably farther away at 3+ feet. That is why I think your testes between the 35 and 50mm is hgihly unfair.
    To me, the 35mm/f1.4 is a little better than the 17-35 from 2 feet away,
    I did spend a fair amount of time to simulate your test environment. No offense, Steve, but I find a lot of problems in your test methodology, mainly in terms of subject distance, shutter speed, and ISO. All of my tests were done at the base ISO 200. I would suggest retest your lens with the right methodology, and if it still shows softness, get that fixed. If you need to get the AF motor replaced anyway, they can take care of any other issue.
    00VsfG-224529584.jpg
     
  29. Thanks Shun. To me the 35 1.4 looks a whole lot better than the 17-35.
    Thing is I still maintain that the testing methodology is not a problem as long as it's the same for both lenses, in that way any shortcomings will be applicable to both.
    Steve
     
  30. ShunCheung

    ShunCheung Administrator

    Steve, part of my problem is that my feet accidentally moved my tripod a few times when I was changing lenses. My DVD box also fell from the wall a few times that I had to tape it back on. When your camrea is that close to the subject, it is very difficult to align the sensor/film plane to the subject. For my images, I would only evaluate the area I highlight below. There was where I manually focused on. The rest of the image may be in or out of focus due to subject alignment at different times and the shallow depth of field. For the area in focus, the difference between the 17-35mm and the 35mm/f1.4 both at f4 is small. And even so, that is an inappropriate subject distance to evaluate those lenses. One lens may perform better at 2 feet, but that means nothing if it is supposed to be used from 20 feet to infinity.
    To be blunt, Steve, IMO your testing methodology is completely wrong. Your subject distance was very different between the two test images (3 feet is 50% farther away than 2 feet) and inappropriate for the lenses involved. Your shutter speed was also different. When your shutter speed is that slow, the impact can easily be different from frame to frame. That is why to test any lens in the macro range, you have to use a flash as the sole source of light to eliminate any vibration concerns. And if you did not use mirror lock up or exposure delay as I did, it completely invalids your results.
    00Vsg4-224537584.jpg
     
  31. ShunCheung

    ShunCheung Administrator

    And Steve's landscape image looks perfectly fine. Apparently it was focused to near the center of the frame. It looks very sharp there as I highlight below. Both the extreme foreground and the background are out of focus, pretty much as expected.
    That is why I still see no reason to believe that there is anything optically wrong with Steve's lens.
    Incidentally, while the 17-35mm/f2.8 AF-S is very fine wide zoom, in the range they overlap, the 14-24mm/f2.8 AF-S is now slightly superior. I have both lenses and have carefully tested them side by side. On the 24MP D3X, some of the flaws from both lenses become more obvious, more so for the 17-35. I am looking forward to seeing how the new 16-35mm/f4 AF-S VR stacks up against them.
    00Vsga-224547884.jpg
     
  32. Like Steve, I've never been too thrilled by the performance of the AF-S 17-35 2.8 either. And I am not alone, I know three other guys who bought the lens and seldom use it.
    I have the AF-S 17-35 2.8 , the AF-S Nikkor 24-70 1:2.8G ED and the AF-S Nikkor 70-200 F2.8 G ED on FX format (D700).
    The 17-35 is the weakest of the pack. The 24-70 and the 70-200 are pretty equal in image quality. All three are wonderful lenses, but there's a difference. I don't know exactly what it is. But the lenses you use the most are normally the better ones. And indeed, the 17-35 is nothing more than a fast zoom lens with good build quality. You cannot compare it with a prime.
     
  33. I think Steve's lens may have issues beyond the AF motor not working. In the gym photo, the lettering on the basket is clearer than the lettering on the shirts of the gals holding the flag. The box truss in the ceiling of the gym over the far corner of the court is quite sharp as well, as if the lens were severely back-focusing. It's possible there are internal misalignments in the moving elements, etc, not allowing the AF to work, or that caused the AF to stop functioning.
    While it's not the same lens or AF system, I had some internals let loose on an 80-400 VR zoom, and prior to 'the big failure' day when the zoom locked up, over weeks, the AF looked progressively soft, and MF didn't help. Mine was fixed under warranty and shoots sharp as new (it's quite sharp for a tele zoom). I'd hazard a guess that if Steve gets his lens repaired, it'll be fine. If not, as mentioned, it'll be fully functioning for sale.
     
  34. Just curious, but the photos Steve has posted all appear to be smeared with some sort of heavy noise reduction or JPEG artifact removal processing, kind of like what you get when you use Blow-Up or something similar? How can anyone make a fair assessment of a lens when the image file is not pure?
     
  35. Agreed Nathan, they look edited.
     
  36. ShunCheung

    ShunCheung Administrator

    The 17-35 is the weakest of the pack. The 24-70 and the 70-200 are pretty equal in image quality. All three are wonderful lenses, but there's a difference. I don't know exactly what it is. But the lenses you use the most are normally the better ones. And indeed, the 17-35 is nothing more than a fast zoom lens with good build quality. You cannot compare it with a prime.​
    Cees, I wish you could provide some more concrete information than this. At least for me, the lenses I use the most are the ones with the focal length range and aperture that meet my needs. I tend not to own any really bad lenses; I have made one or two mistakes before but I sold off those lenses quickly.

    I have demonstrated in this forum several times that the 17-35mm/f2.8 easily beats "primes" such as the 24mm/f2.8 AF-D, although those primes are very old designs by now. It'll be interesting to see how the 17-35 and 14-24 stacks up against newer lenses such as the 24mm/f1.4, etc.

    As I pointed out to Steve earlier, a RAW file from the D3/D700/D3S has 4256 x 2832 pixels. His files all have 4272 x 2828 pixels. I can understand that perhaps a few pixels are cropped from the short dimension. But why the long dimension has more pixels than what the camera provides is puzzling. His files are certainly edited.
     
  37. 'fraid not guys, no editing or post processing on the test shots. There was some on the scenic I think.
    Shoot on D3 in RAW lossless compressed, imported to Aperture, exported as "jpeg of original size". Beats me why there are any irregularities, not my area of expertise (yet!).
    Steve
     
  38. Steve, I always felt my 17-35mm was not very good, esp. at f2.8. Lots of CA, soft, etc. 24-70mm was much better. I only ever tried one and I guess maybe mine was not the best sample. I sent it to Nikon and they said it was at factory standard. I basically stopped using it in favor of 24-70. Now bought the new 16-35 and will have to see how I like it, but so far it seems only a touch worse than 24-70, but generally same feel to it. I always wanted to try another sample, but never had the chance to.
    Dmitry
     
  39. I don't see a lens problem based on the landscape shot.
    I think, in order to make a credible claim that this lens (not just this sample) is an under-performer for images such as this landscape shot, one should compare apple to apple. One needs to use the various "better" lenses on the same scene with the same composition, on a tripod with remote trigger, shot the same way - i.e., f/8, 1/60s, ISO 400, focussing dead center, and under less-than-interesting light.
     
  40. So, why not try a few other 17-35s???
    Borrow someone else's. Rent one. Borrow another. Get a loaner from your favorite photo store.
    You will have then done an empirical experiment where the changed variable is substitution of others that do not have broken down internals.
    Ashamed I did not post this before. Apologies if this has been suggested already. I could not bring myself to read the last of it all.
     
  41. ShunCheung

    ShunCheung Administrator

    So, why not try a few other 17-35s???​

    I think that is unnecessary.

    Your need to establish that your current copy has problems first. Steve's own landscape image looks perfectly normal. The area in focus is extremely sharp, sharpness is even on the right and left sides, and the entire image is free from any chromatic aberration. Therefore, that lens is fine at least at that focual length, which is 24mm and about in the middle of the zoom range. I would repeat the test at the long end and the wide end. But for the most part, Steve's image already discredits those claims that this lens is unsharp.

    Otherwise, you can get 10 perfectly fine copies of the 17-35, but if you keep shooting test images from 2 feet away at 1/10 sec, you'll see "problems" over and over.

    The remaining question is why Steven's gym image is out of focus. If, for example, one of the 51 AF points that happens to cover the background was actually in use for a brief moment, the camera would simply follow command. Or perhaps there are indeed AF issues on his camera and/or lens. Only Steven himself can figure that out by running more tests.
     
  42. Or perhaps there are indeed AF issues on his camera and/or lens
    Shun, I think he said that his AF motor is goosed, kaput ! So his AF points are null and void. So he has been manual focusing it. Maybe it comes down to an eyesight issue.
     
  43. Andrew, I hope it's not an eyesight issue, as I explained I'm a professional wildlife cameraman, all manual focus with HD cameras and lenses like the Canon HJ40 that's equivalent to 2500mm+. No complaints from BBC or anyone else so far!
    Steve
     
  44. Just another quick thing on focus. This shot was on a D3 at 17mm f4. Even if I'd focussed on something 100' away there'd still be depth of field from about 7' to infinity. The flag and all the people are further away than that.
    Steve
     
  45. ShunCheung

    ShunCheung Administrator

    Steve, the image with the flag was captured on a D300S, not D3, according to your EXIF data. With the crop factor, your lens was not nearly as wide and those people were likely much farther away than you thought. But somehow that still did not give you sufficient depth of field.
    In any case, if your 17-35mm needs to get its AF motor replaced anyway, you might as well send the lens in and ask them to caliberate it. Otherwise, if you really want to know whether there are an optical issues with it, I suggest you go out and retest it with a sun-lit subject (i.e. you'll use a fast shutter speed to eliminate all vibration) similar to your landscape image @17mm, 24mm, and 35mm. For each focal length at f2.8, f5.6 and f11. You can pick to focus at say 10 feet. With a continuous foreground to some distant background, it should be very easy to determine whether there are any issues or not.
    Otherwise, without further images, I am afraid that we have already squeezed as much out of the existing test samples as we can.
     
  46. Yes, I'm sure you're right. If I get around to doing any more meaningful tests I'll post them.
    Thanks,
    Steve
     
  47. it is possible that the perimeter focus points (even just one) were used briefly and the camera thought the shooter wanted to use them and locked on. that is why the banner in the officials' table was sharp and not the intended focus point(s)
    i had a very short stint with ABC as a contract photographer in the far east. with no malice or offense intended, we, photographers never allowed our professional cameramen in the group even touch our equipment. they have their field of specialization, we have ours. in a critical project, this is almost a law not to be broken, in the old days. it may be different now.
    steve might be a professional cameraman handling thousands of dollars/pounds/euros worth of equipment but that doesn't mean an equivalent in the knowledge of the behaviour of equipment for stills.
    a very simple change of test methodology would have answered most, if not all, of the concerns and questions here about the lens. please remember that i am not a very active professional and come from the old school. these are just my thoughts.
     
  48. Ramon, I've been taking stills with SLRs for 30 years. A lot of the disciplines are similar.
    Steve
     
  49. Ahh, Steve never underestimate the eyesight issue. In my fifties and wearing glasses, I could never use a manual focus again. I could never be sure that what I see in focus is correct or simply according to my eyes, with or without glasses. Thank goodness for AF ! Not saying that you have a problem, but the older we get.......
     
  50. ShunCheung

    ShunCheung Administrator

    Eyesight aside, modern SLRs' viewfinders are not designed for manual focusing.
    I have been using Nikon cameras since 1977, the year AI was introduced, so I am not that young any more. However, with glasses, I still have 20/15 eyesight, better than average. But even back in the 1970's when I was a teenager, I had to have split-image focus assist for precise manual focusing. A few years ago I ran into the same problem with my Contax 645 and added their focusing screen with split-image focus assist (similar to the KatzEye) for manual focusing.
    Today, when I use the likes of the D3, D700, D300 ..., live view is the way to go for precise manual focusing. You can magnify a small section from any part of the frame and fine tune your focusing.
    Another issue is test procedure. I never realized how difficult it was to line up my camera with a flat, macro subject from a foot or two away. A few years ago I taped a 10 Euro bill on the wall for testing my 200mm macro. After a long long time I wasn't able to get all four corners line up in focus; the camera can easily be a bit too high or too low or slightly tilted to the left or right. Even at f11 or so, depth of field is still very shallow in the macro range. Essentially I have to use a mirror on the subject plane until I could see the camera straight on thru the viewfinder to line it up.
     
  51. Another consideration about manual focus (Live View) against camera autofocus.
    I have been testing and comparing my latest lens(es) resolution power. I took many photos from a software`s box at a distance of 6-8 feet, using Live View (tripod mode), w and w/o using a Rodenstock 4x loupe for better focusing. I repeated the test many boring times.
    Result: very inconsistent results. Some pics were perfectly focused, others almost perfectly focused.
    I decided to try with the camera`s AF:
    Result: very consistent, almost all the pics (if not all) were in perfect focus.
    Conclusion: with some lenses, a very slight focus error can be produced even using Live View, and will result in a non-perfectly focused pic at 100% on the screen.
     
  52. ShunCheung

    ShunCheung Administrator

    Jose, that has not been my experience. Manual focus with live view has always worked very well for me. You are focusing on the same sensor which captures the final image. There is no mirror or anything in the middle to mess up focusing.
    Another plus with live view is that you not restricted to some 51 or whatever number of AF points. You can use any part of the frame to focus.
    Of course, you have to use a tripod and it is a very slow process. Great for landscape, macro, etc., including testing lenses, and not practical for many situations.
     
  53. I think I don`t expressed it correctly. I wanted to mean that given the... roughness? and... delay? of the screen when using Live View, sometimes could be a bit difficult to be sure if focus is optimal (I suspect it depends on the lens or magnification used, for macro shots I don`t use to have any problem).
    Even so, I think it`s an extremely useful tool (that I like to use for almost everything).
     
  54. ShunCheung

    ShunCheung Administrator

    Jose, I assume you use the +- buttons to the left of the back LCD (or the command dial in the case of the D3 family) to magnify a small section of the frame to fine tune focusing.
     
  55. Here a couple more tests. This time outside, distance of 10m or so, 400ISO.
    The first is with the 50mm 1.8AF at f2.8. Admittedly this is stopped down 1.5 stops and the 17-35 at 2.8 is wide open. Even so, I expected the 17-35 to be superb even wide open, and as you'll see it's not as sharp as the 50 by any means.
    Steve
    00VuKv-225645584.jpg
     
  56. This is the 17-35 at 35mm. I know, I know it's zoom vs a prime, and it's wide open vs stopped down. But I'd just heard so many great things about the 17-35 I expected better.
    Steve
    00VuL3-225647584.jpg
     
  57. Shun, I do... now I`m wondering if my eyesight is even worst than I thought... 8-(
     
  58. Just for the fun of it I downloaded both of your images and gave them my normal work-flow treatment which includes light sharpening. They look great to me, after light sharpening which all my images need (unless I crank in-camera sharpening way way up). Just something you have to do with digital. In fact, after sharpening the 17mm looks superb (IMHO) for a wide open f2.8 shot. I presume your focus point was on the fence. Your 50mm shot needed sharpening too.
    While certainly not an expert in all things digital, I have learned there is a right amount of sharpening and then overdoing it. Try to avoid oversharpening, it looks like crap.
     
  59. Steve-
    Sorry, I got your images reversed in my little experiment so that changes my conclusions. Your 50mm shot, indeed, looks quite sharp (after sharpening per above). In looking closely at the 17mm shot it appears your lens is focusing IN FRONT of the fence by a couple of feet or so. If you look at the chairs on the left and then follow that focus plane to the right you can see that there is a line of focus in front of the fence. Adjustment or fine-tuning in order?
     
  60. I have the Nikon 20-35 f2.8 AF D, the predecessor to your 17-35. It is soft from within 7 or 8 feet - the closer the focusing distance gets, the softer the performance. At infinity, however, its is unsurpassably sharp.
     
  61. I too was never particularly impressed with my 17-35, even though Bjorn Rorslett and others rave about it's sharpness. Personally, I found my Nikon 35-70 2.8 to be sharper at 35mm that the 17-35.
    Lisa
     
  62. ShunCheung

    ShunCheung Administrator

    I think the 17-35mm/f2.8 is a very sharp lens. However, you must focus correctly and you can't expect too much wide open at f2.8. That was why I suggested to Steve to try at least 3 different apertures: f2.8, f5.6 and f11.
    Having said that, the 17-35 is now superseded by the 14-24 in terms of sharpness. I am waiting for tests of the new 16-35mm/f4 also. Additionally, on the 24MP D3X, the 17-35 is showing some edge perofrmance issues even at f8 or so.
     
  63. << ... the 17-35 is now superseded by the 14-24 in terms of sharpness. >>
    I read that the 14-24 is sharper when shot wide open, but not so when stopped down. The 17-35 is basically a landscape lens, and landscape images are usually shot stopped down for maximum depth of field.
    I am skeptical of "tests" that are performed without consideration of all the variables that may make a difference in the results.
     
  64. But if you've got 2 lenses in the test and the variables are all the same then you've got a valid test.
    And I wouldn't say the 17-35 is just a landscape lens, it is used in all sorts of applications. Sports for example, wide angle impact shots of dunking in basketball or from inside an ice hockey net for instance. These venues will have poor light and need wide apertures.
    Steve
     
  65. And you were testing the 17-35mm for its suitability for sports photography? Next I would test the 50mm for wildlife. LOL!
    Oh by the way, the "valid test" you performed did not have the same variables except for the apertures and ISO: The differences between the 50mm and 17-35mm test, respectively, were: 50mm vs 35mm; 180/s vs 125/s. The points of focus was different as well (as someone pointed out). It appears our test adds more confusion to whatever claim you made.
     
  66. Looking at your test shots, your 17-35 looks fine to me. At f2.8 its sharper than the Nikkor 24mm f2.8 I've been using for 30 years or so. Even so, f 2.8 is handy for indoor candid shots, and mine has done many a memorable image at f 2.8. Indoors, hand held at slower shutter speeds, at 2-10 feet, sharpness is not the biggest issue; aperture is more important. Shooting a landscape, you'd stop down anyway, and hopefully use a tripod. If you are really unhappy with the lens, I would try another sample or go for a different lens.
     
  67. Mary, I think you'll find many people use wide zooms for sports photography, Dave Black for one - do a little investigation.
    I never said that it was an equal test, how could it be, as you point out the focal lengths were different for a start, but the sort of variables that might adversely affect each lens were the same. Shutter speed was high enough that it would never be an issue, close enough to subject that atmospheric haze was not relevant. Don't understand why the point of focus was different as it looked right to my eye and was right according to the D3's focus indicator.
    I have just got rid of it, picked up the 24-70 instead and it's amazing! Sharper than the 24 f2.8 by a country mile, and sharper than the Sigma 50 1.4 at f2.8 even though the Sigma is then stopped down 2 stops. Say what you like about my testing but I know a well-performing lens when I see it, and the 24-70 is it, my 17-35, for whatever reason, was not.
    Steve
     
  68. << I have just got rid of it >>
    Great, it's about time you do that. LOL!
    << picked up the 24-70 instead and it's amazing!>>
    Sorry again, I use this one now. It's nice, but I don't see a huge difference in results.
     
  69. Well there you go Mary, if your 17-35 and 24-70 were similar and mine were not that reinforces the idea that my 17-35 had some sort of issue. Thanks for the confirmation.
    Steve
     
  70. The 17-35 is a great street lens too. That's what I've used if for straight candid shooting and also events. Works very well on the D200 and 700. I shot a street event with it, and I felt I wanted a little more reach, so I got a 24-70, but I'm not sure I like the IQ any better yet...still getting used to it.
     
  71. Man I must've gotten a great copy of my 17-35. I shoot film and only film and my shots are brilliant in color and tack (almost ridiculously) sharp. I know when I have used this lens against other lenses. It's got a look of it's own...
     

Share This Page