20 f/2.8 AF-D Sharpness Quality Issue?? Advice needed.

Discussion in 'Nikon' started by lee_vgg, Jun 23, 2010.

  1. Note: I've posted this as a response on an old thread, but was hoping to get some quick response in case I need to return this lens soon.
    I just picked up a used (Excellent Condition) 20 f/2.8 AF-D from Adorama and am considering returning it. I'm concerned it isn't as sharp as it should be.

    I would appreciate it if someone could take a look at the comparison photos to a 17-55 posted on my Flickr site and provide an opinion?
    Below are two examples. I can post more if necessary. 1st is roughly the center of the image and focus point at 5.6. 2nd is at f/9 in the bottom right corner. It really helps to view the rest of the shots at the various apertures. In general it does not meet my expectations from reviews online, and at worst it seems blurry.
    I've compared the 20 to the 17-55 at 20 and included each shot's full resolution file as well as print-screens of the comparison in LR3 Beta 2. They were all taken hand-held, available light, overcast skies, with no processing in LR3 Beta 2. All shot at ISO 400, same speeds at various apertures with a D300s. B&W MRC UV filter on the 17-55, nothing on the 20. All auto focused. The first set (labeled A) focus at the bottom left corner of the back windshield. Second set (labeled B) focus on the middle of the back, red edge of the seat.
    I noticed that the 20 seems to have out-performed the 17-55 for the close-up shot of the bike in the corners at f/9 (see f9 B Corner 1 and 2). Strange, because it absolutely failed when focusing on the car at all apertures. Maybe I shook the 17-55 a bit at 1/80th, not sure.
    Could this be a bad lens? Or is this the expected performance? I wouldn't mind the corner sharpness or having to stop down for full sharpness, but I want to use this for quality landscapes while traveling and if the center isn't super sharp at f/9 or f/11, it's really a deal breaker.
    Thoughts are much appreciated. Thanks.
  2. The AF 20mm lens was designed for film camera use. If you want a DX lens that is sharp to your requirements, you may not want to keep the AF 20mm lens. Nikon has spent some time engineering lens for the DX format....
  3. I understand there may be sharper lenses than the 20 mm, and the 17-55 is for sure one of them. However, I'm more interested in knowing whether this is what I should expect from the lens on a DX body (which should actually improve performance compared to FX). Many people appear to be using this on DX and FX digitals with great success in terms of sharpness.
    It seems like the center sharpness at f/9 isn't great. If it's what I should expect, then so be it, but it isn't then I'm going to return it and wait for another chance to purchase the lens.
    I would think any prime on any body should meet my requirements of "super sharp" in the center at f/9 or f/11. If not, what's the point?
  4. Nikon designs a DX lens to have crisp center area sharpness. If you get a FX body and take a image with a DX lens, you can see that the area of focus is not covering the whole sensor.
    The AF 20mm lens you have (used?) may have had a rough life and it may not behave as a new lens would. Only the clerk at Adorama knows who had the lens before....
  5. Jerry,
    Thanks for taking the time to respond, but I understand that it is a used lens and may have issues. My exact question is very specific - whether someone who has experience with the 20 mm can tell me whether the sharpness shown in these photos is typical of the lens or whether it is under-performing.
  6. " B&W MRC UV filter on the 17-55, nothing on the 20." - not fair comparison, - take off the filter of the multi group multi element zoom lens since is not needed there - except for protection, and put if on the old less coated lens that would benefit more from the UV filter, and try again.
    Your choice of subject for the comparison, the shiny glass and shiny car, is not the best to compare for the sharpness. Also compare sharpness in the flat surface, to eliminate any out of focus influence, and use a tripod, to eliminate incidental motion blur.
    You know better, but seems that the pictures are much different from both lenses.
    Try something less relfective with more defined details, and you could see better the difference. Typically, they compare using large newsppaper on a flat suface like a wall or a barn doors.
    Do your test also from a longer distance.
    The 20/2.8 new Nikkor I have is sharp all over the field, definitely af 5.6, 8, etc. but is more prone to reflections, and does not reduce the haze as well as your zoom.
    For one 17-55/2.8 perhaps you can get 2 or 3 older used 20/2.8 lenses?. The size and weight of the 20 mm makes it more suitable for walking around the city or in wilderness.
  7. Frank,
    Thanks for the helpful comment. I don't have a B&W for the size of the 20 mm, otherwise I would've used it. And I didn't think about the 17-55 w/ B&W until after the fact. I wouldn't think the 20 mm "needs" a haze filter.
    You make a good point that the scene has a lot of reflections. I thought this would be a good test for sharpness... I'll try another scene or two.
    Tripod - yeah, didn't have it with me. Tried to keep speeds above 1/80. Will take some with the tripod.
    Your last comment really makes me consider sending this guy back. I just wish there were new copies available....
  8. Lee sharpness is a visual illusion and harder to test for than resolution and contrast.
    So to test for sharpness is better done using a resolution test target.
    Luckily the difference between the newer zoom and the 20mm lens is so obvious that a test under ideal conditions is not really needed.
    Yes the zoom is better in sharpness than the zoom in the near distance where the zoom is at best. I rarely used both lenses stopped down a lot so it is a bit surprising (but not totally off) that the difference is so large at f9. But you may want to test if this is symmetrical. Some of the 20mm AFD lenses I saw were not perfectly centered.
    I would only prefer this prime for size and price over the excellent zoom (for DX).
    If you are after better performance shooting against the light the older AIS 20mm f3.5 is better.
    Hope this helps.
  9. Lee,
    I do not own these lenses, but looking at the charts/verdicts on Photozone.de I would expect noticable differences between the two of them.
    Whereas the 17-55 center resolution is judged as "Stellar", the 20mm reaches only "good",
    Also the 20 gets comments like : "The Nikkor AF 20mm f/2.8 D had a very good reputation during the film era but the D200 didn't really like the lens"
    This does not say that the 20 is a bad lens, but it would imply to me that it isn't the best wide lens available either. ..
  10. I try to remember that time ago I compared my primes to the 17-55 on a D200... It was a fast dirty test where I found that the only prime that outperformed the zoom was a 55 Micro (I had the whole range in different versions, but sadly not the AFD20/2.8). The 20/2.8AiS was clearly the worst.
    Then I`d not be surprised if the 17-55 outperforms any prime... as it was my experience.
    Another issue could be the right focus;
    • Are your pics 100% in focus? You need to know this to compare sharpness.
    • Are you 100% confident in your camera&lens AF ability? If so, nothing to say.
    If you have Live View, you should use it to achieve an almost perfect focus.
  11. FWIW, I recently did a quick & dirty comparison of my 20/2.8 AFD to my new 16-35/f4 VR on a d700. The results were similar to yours: The 20 prime was clearly not as good as the newer zoom under any conditions, eg, center, corners, stopped down, etc., but, by no stretch could it be called "bad".
    For many applications, it's far more than "adequate" and I still use it for walking around or other situations when I don't want to look like I have a grenade launcher attached to the front of my camera. :)
    Attached is an image taken a couple of years ago with my 20/2.8 AFD on a d200 body, handheld. This version is obviously processed, but as you can see, one can certainly get a good sharp, contrasty image out of it, pixel peeping criticisms not withstanding. From your comments, I suspect your lens is probably in ok shape.
    Tom M.
    PS - If you haven't already done so, shoot a brick wall and compare the sharpness in all 4 corners. If they are the same, the lens probably hasn't taken a hit, was assembled well, etc.
  12. Could this be a bad lens? Or is this the expected performance?​
    I tried three different 20 2.8 on DX when moving to digital, and none was up to even the 18-70 AFS DX that came kit with the D200. It was a big disappointment since I loved it on film, but I would say yes, it is the expected performance. That lens is just not good on digital.
  13. For what it's worth (playing today out back)...
    20mm 2.8 AFD versus 35mm 1.4 AiS.
    Both at f9 (supposedly beyond the 35mm's sweet spot) to keep it balanced.
    Nikon D200, fine, large jpg. Tripod. I tried to keep the same field of view, not distance.
    The 35 and a couple of others are my sharpest (I lack any modern zooms except the 18-35).
    The 35mm @ about F4/5.6 has produced some stunning larger prints when my aged eyes correctly focus the darn beastie.
    Good luck.
  14. To beleaguer the topic, the 35 1.4 at F4 ... note the street sign. Amazing to me.
    (D200, same sort of conditions)
  15. Before returning it you should test it again, this time focus manually instead of autofocus. If your camera has live view use it with the magnification to manual focus. Also test it shooting a flat surface parallel to the sensor plane like a brick wall, on a tripod.
    It's entirely possible you have a lens that is front-focusing. If you have a D300/D700/D3 you can fine tune the AF for that particular lens, problem solved. If not then you may need to swap it for another lens that hopefully will AF better on your camera. I know the 17-55 is a great lens but I doubt it would produce significantly better (if any better) sharpness than the prime.
  16. I think your focus is off. Could be user error or bad lens. In the first photo the street (behind the car) is in focus and looks sharp compared to the zoom. I have the 20mm and love it; sharp with little distortion. Try a few more controlled tests then contact Adorama. Their support is excellent.
  17. Thanks everyone for the comments.
    I mainly want this lens for traveling and will send back the 17-55 if it does well enough. I'll try a few more unbiased shots and re-post.
    May also try fine tuning it to the D300s...
  18. I bought a 20/2.8 and was very disappointed. I eventually sold it.
  19. I've had two copies of this lens and was hugely disappointed on my film cameras. One of the worst Nikon lenses I've ever used, it wasn't even sharp dead center and I know my focusing wasn't the problem. The 17-55mm is outstanding and the 12-24mm is right behind it. I'd say return the 20/2.8 and try something else, maybe a Zeiss ZF 21.
  20. I sent back a 17-55 DX because sharpness was very mediocre for such an expensive lens, and because it had a high level of barrell distortion.
  21. I also had a Nikon 20mm and sold it since I had very disappointing sharpness on digital camera's. In my opinion you will never be satisfied with it.
  22. Count me in as another former owner of a 20mm 2.8 AF who thought its resolution was below par on a DX, and sold it.
  23. I hate to be the guy that points out the elephant in the room, but has it occured to you that you're not comparing apples to apples? The fact that one lens is DX and one is not is not important only because of how much of the sensor they cover, but also the level of distortion. Your 20mm lens distorts (and has softness issues) like a 20mm lens. Your 17-55 is equivalent to around 28mm, and distorts more like a 28mm lens. If anything, the 28mm 2.8D is much more of a fair comparison, as both lenses SHOULD distort more similarly.
    If you really wanted a fair comparison, you'd need to test it against the 12-24 DX, as 13-14 in DX works out to about 20mm on camera.
    Personally I've never used the 20 2.8 enough to make a judgement call, but a friend is very fond of it, and I've had excellent results from the 24 and 28 2.8s. At the very least, the D lenses work on almost every Nikon ever built, while the DX lenses do not.
  24. My conclusions were reached on an FX camera-- the D700.
  25. http://www.naturfotograf.com/lens_wide.html
    I owned both the the MF and AF 20mm 2.8 and like the review from the link I found the 20mm AF was not as sharp as the MF. shooting a D2X I ended up shooting with a 28mm f2 MF as my wide lens. It's not as wide as the 20 but it is a sharp lens even wide open
  26. Zack,
    "because of how much of the sensor they cover, but also the level of distortion"
    Really? I thought it just cropped / zoomed in on the image (in super simple terms). Anyone care to comment whether it would adjust distortion as well?
    I just ordered the 12-24 DX and will probably go with it. I'll do some more tests vs the 17-55 this weekend, but will probably send both back (Adorama is being kind and letting me return the 17-55 after the 15 day period). I'm looking for something smaller and liter than the 17-55. The 12-24 isn't as small and lite as I would like, but it's closer.
    I did try the Voigtlander 20 f/3.5 at one point and love the size and quality but am just too scared to really solely on a MF for wide angle. I could manage the landscape shots, but I don't think I could manage the street / action shots. That's probably lame of me, but there it is.... Otherwise I'd look at the old 20 f/4.
  27. Lee
    I had my 20mm AFD 2.8 Nikkor overhauled by Authorized Photo Service after experiencing stuck diaphragm blades, they adjusted and fine tuned the lens to factory specs.
    It works equally well on D700 or D200. I like the results I get, realizing after all, it is a an ultra wide.
    This is a grab shot from a month ago, spinning around, aiming the camera and pushing the button.
  28. Lee
    The above should be the full frame, and it was sharpened very minimally in NX2 to correct any AA filter loss of sharpness.
    I would go out and shoot with the lens and see how you like it on a number of subjects. I generally leave the shooting of bricks and walls for others, after all there are other qualities that make a good photograph beside ultra sharpness.
  29. Thanks everyone for your responses. I've Auto Focus Fine Tuned each lens on the D300s. It made a huge difference, but surprisingly for the 17-55 instead of the 20! The 17-55 was clearly focusing too far in the distance. The 20 mm was also focusing in the distance, but because of a slight blue haze apparent in the image and poorer sharpness, the adjustments didn't make as obvious a difference. I settled on -7 for each lens.
    Here's a comparison of each after fine tuning (full view of the scene is further down in the post). Clearly the 17-55 wins. This book's binding is normal glossy, not extreme but not matte either. Natural light was entering from the right, and two lamps were in the back corners of the room.
    I was curious if I could get the 20 mm to the same sharpness of the 17-55 mm in post processing. So I sharpened (+40 in LR3) and added detail (+50 +/- in LR3). No matter what I tried I couldn't get rid of the blue haze, which is the real culprit in making the 20 mm look blurry.
    Here is the 20 mm RAW vs sharpened:
    Here is the 20 mm sharpened vs the 17-55 mm RAW:
    However, once you sharpen the 17-55 mm it again takes the lead:
    Here's the full image, with a comparison of the 20 mm sharpened and the 17-55 mm RAW:
    Well I am still on the fence about the 20 mm. By the time I thought of placing the B&W haze filter on the 20 mm to try and filter out what I refer to as the "blue haze", the sun had set. I went ahead and took a shot, and found that the "blue haze" was reduced. This helped the sharpness but it still was not as sharp. Here's the comparison (17-55 mm image is from previously). I would say this is acceptable, but I'm going to be shooting a lot of sunrise / sunsets with water and sky. These were all taken with soft evening light through a window. That's just about optimal conditions, except that the binding of the book is a bit glossy.
    I'll test the 20 mm on the same scene and time Sunday, but with the B&W filter. If the B&W reduces the "blue haze", then I may keep it.
    Thoughts and comments are appreciated.... Thanks.
  30. Well I'm sure some is reading this and thinking "but he's comparing those at f/2.8!" Well yeah, that's not really good is it. I was using f/2.8 to see the AF Fine Tune effects the best, but probably should have done some additional comparisons at f/4.5, f/8, f/11.
    Well I just did and got some interesting results....
    How about that focusing now? Well, the first shot at f/4.5 focused too close. Second shot at f/4.5 focused correctly. I don't suppose this fine of error really matters in the field, unless I'm at f/2.8-4.5 and trying to close-focus on a subject where this error would matter. If it does, I just gotta be sure to check the LCD and re-focus if appropriate.
    Here's the comparison:
    Here's the comparison of the 20 vs 17-55 at f/4.5 (which the good focus of the 20). Seems the 17-55 focused a bit too far back in this shot. Geez, you can't win huh??
    Here's the 20 vs 17-55 at f/8.
    Here's the 20 vs 17-55 at f/11.
    Here's two other areas of the image at f/11. Not bad little 20 mm!
    This 20 mm is now doing a pretty decent job against the 17-55. We'll see how it does tomorrow in the daylight..... I am hoping the B&W haze filter plus stopped down a bit will address that "blue haze" issue.
    I also noticed that the 20 mm seems to let in a bit more light. Maybe because of the 17-55 ED coating or zoom nature?
    Thanks again for everyone's comments. It's about 11 pm here so enough with this for now....
  31. While this thread may take the award for the ultimate pixel-peeping thread of the month, the latest results seem to exonerate us old fogies who, for the sake of nostalgia, weight, or unobtrusive size, occasionally mount one of our old primes instead of the newest glass and somehow manage to get decent results. ;-)
    Thanks for the interesting series of tests, Lee.
    Tom M.
    PS - There are quite a few older primes that, IMHO, simply can't hold their own against modern glass, e.g., the Nikon 28/2.8 AFD, the 35/f2 AFD (in the corners wide open), etc. The bottom line is that one has to know the limitations of one's equipment and use it appropriately.
  32. I few more tests today outside and in low-light and I think I'm convinced the 20 mm is alright for my needs. It is obviously not as sharp as the 17-55 at 2.8 and next to it seems blurry. At 4.5 it's noticeably less sharp but you have to do a double take to catch it. At 8 they are almost the same, though the 17-55 still wins just barely. With a little sharpening they're the same. At 11, they are pretty much equal, though I would still give a slight edge to the 17-55.
    I'll mainly be worried about sharpness for landscapes, which will typically be around f/8 or f/11. If I'm shooting at f/2.8 and absolutely need a wide angle, it will do decent enough. Most shots of that style will be for me personally with very little chance of making large prints.
    One interesting thing is that even though the 17-55 is sharper at f/2.8, that's assuming you focus correctly. I took a few hand-held test shots of my wife in a dark room with one lamp silhouetting her. I think I was at ISO 1600, f/2.8, and 1/125 or 1/200. There were more in-focus sharp images with the 20 mm than the 17-55! So when you add human error, even if you have an incredibly sharp lens, you may not get sharp images. Two things I do know for sure though: 1) The 20 mm was $400 and the 17-55 was $1300, and 2) The 17-55 is a pound heavier and massive next to the 20 mm. I think I'll be sticking with the 20 mm.
    Thanks all for your comments!
  33. I've compared my 20/2.8AF-D and 16-85AFS on my D300, and yes, sure at f/2.8 the corners are not great, but at equivalent apertures and 20mm focal length my 20/2.8 compares very favorably with my 16-85 and in fact has better (I.e., more) depth of field than the 16-85 at the larger apertures.
    Side Note: I've become pretty convinced that these optically complex zooms have significantly less depth of field than the simpler optical designs AND less than theory predicts.
    It is just as sharp in the center as my 16-85 (which is pretty darned good on the D300) and if you clean up the CA in the corners it's very good at higher f-stops.
    Like my 24/2.8 it's not the greatest in the corners wide open, but for landscapes I would not use it wide open (unless shooting star trails at night) and for people/candid shots in low light the center sharpness is more than good enough.
  34. I just bought this lens on d300s with the aim of moving up to FF later on. Took a few indoor facial shots at f3.2 and they came out soft. Autofocus just seems to miss even though the point was directly on/around the eyes.
    Set it at 5.6 and still the image quality wasn't particularly good for a prime. My 17-55 2.8 definitely outperforms this.
    Hmm...not too sure about this now but it seemed perfect to use as a street lens on my DX sensor and then great for travel when I pick up my FF.
    I guess I had my hopes up thinking I could get around the 'weight' issues of my current lens (Tokina 11-16. So sharp and produces excellent images) but no.
    Sending it back to the shop to check after they saw a few image samples.

Share This Page