Simple tests: Sigma 30mm/1.4 vs Nikkors

Discussion in 'Nikon' started by arthuryeo, Jan 12, 2008.

  1. Basic tests to see for myself, using visual inspection, whether the Sigma 30mm/1.4 was truly what it claimed to be. These tests were done using D2X, on tripod, MC-30, MLU, untouched images, Nikon CaptureNX.
    It's pretty clear on test charts, when wide open, the Sigma 30mm/1.4 has
    • distinctly visible chromatic aberration under normal lighting
    • distinctly visible spherical aberrations (double edges)
      It appears that the aberrations and resolution did not really diminish until f/2.5 and beyond. So, is the f/1.4 really for show, you tell me. At f/2.3, the image is as good as it any Nikkors. I have included 2 other images shot with Nikkor 50mm/1.4 and 35mm/2 (all opened wide) for your comparison to see the difference in aberration levels.
      DISCLAIMER This test was done for my own personal fulfillment and shared on a friendly basis. Samples of this can vary so, yours may be different. Point source lighting were not done in this test, so coma aberration is not determined here.
      [​IMG] Sigma 30mm/1.4 @ f1.4
      [​IMG] Sigma 30mm/1.4 @ f1.6
      [​IMG] Sigma 30mm/1.4 @ f1.8
      [​IMG] Sigma 30mm/1.4 @ f2.0
      [​IMG] Sigma 30mm/1.4 @ f2.5
      [​IMG] Sigma 30mm/1.4 @ f3.2
      [​IMG] Nikkor 35mm/2 @f2.0
      [​IMG] Nikkor 50mm/1.4 @f1.4
      Arthur had the 35mm/f2 image twice. I deleted one of them. -- SC
     
  2. The higher-res and EXIF info can be found in this link

    http://www.flickr.com/photos/phlog/sets/72157603696697656/
     
  3. >It appears that the aberrations and resolution did not really diminish until f/2.5 and beyond.

    Sorry, I meant to say that the aberrations did not diminish & the resolution did not improve until f/2.5 and beyond.
     
  4. In actual low-light shooting situations these problems with the Sigma prove fairly inconsequential. I don't think any lens at f1.4 is going to be without some optical compromises. I reiterate that the main problem with this lens is the inaccuracy of the autofocus mechanism.
     
  5. Thanks for sharing, Arthur. The 35/2 wide open isn't without problems on the D2X, it seems.

    If you could, would you try it on the D3 (APS-C crop mode) and let us know how it fares there?

    I found a cool demo on Chromatic Aberrations here: http://www.youtube.com/watch?v=KZl9Cfhp-5o&NR=1

    Douglas, Auto focus problems- wouldn't that be camera/light dependent?

    I have had excellent attrition rates on a D40x compared to a D70, for example.
     
  6. I have the two Nikkors here but not the Sigma. I picked the 35mm f/2 over the sigma lens only because of weight considerations. Although seeing how much nicer the Nikkor performs at f/2 than the Sigma stopped down to f/2 does, I'm all the happier with my decision. It is rare that I want wider than f/2 and also specifically want a 35mm lens. If I want wider than f/2 I can usually happily switch to the 50 f/1.4.

    The difference between the 50 f/1.4 lens and the 30 f/1.4 lens at f/1.4 is simply outstanding. Maybe you have a poor sample of the Sigma 30mm lens?

    Also, you have two images labeled: "Nikkor 35mm/2 @f2.0". I assume the second one is f/2.5?
     
  7. >The difference between the 50 f/1.4 lens and the 30 f/1.4 lens at f/1.4 is simply outstanding. Maybe you have a poor sample of the Sigma 30mm lens?

    Yes, that's in my disclaimer. :) Based on my experience, Sigma lenses tend to have a wider variation among the samples. So, mine is, possibly, not representative of the average.
     
  8. Just to add, my copy of the Nikkor 50mm/1.4 was made in China. :)
     
  9. Arthur, Seriously, I am looking for 1 (or two) more samples of the Sigma 30/1.4.

    If you want to sell yours, please drop me a line. Thanks.
     
  10. ShunCheung

    ShunCheung Administrator

    I tend to agree that no f1.4 lens is going to be really excellent at f1.4. If you must use an f1.4 lens wide open, you have to accept some compromise.

    Arthur, how far was the camera from the subject in those tests?
     
  11. >Arthur, how far was the camera from the subject in those tests?

    They were about 5 ft away.
     
  12. >Arthur, Seriously, I am looking for 1 (or two) more samples of the Sigma 30/1.4.

    Why are you getting so many copies of the same lens?
     
  13. Oh, one reason, the paint on the polycarbonate housing comes off easily.

    I had a 105/2 DC Nikkor with similar polycarbonate shell but that paint is a strong Nikon paint and does not rub off.

    Another reason is that I would like to dissect a sample for a special project.
     
  14. >Another reason is that I would like to dissect a sample for a special project.

    Whoa ... baby ... that sounds painful ... ouch! :)
     
  15. The Sigma 30mm f1.4 seems to have front focusing troubles with the D200 and D300. I chatted with the woman who runs the tech support office at Sigma. Canon-mount lenses have troubles, as well. The focusing problems are (she claims) solvable, but sometimes they actually need the camera body in order to resolve the problem. This seems like a lot of trouble to me.

    And yes, Sigma remains notorious for quality control issues with lenses, which is too bad, considering that some of their lens designs are very good.
     
  16. The rear cap that comes with this lens is very bad and should be replaced with a genuine Nikon cap as well.
    00NzVx-40942984.jpg
     
  17. What portion of the frame are these crops from?
     
  18. >The Sigma 30mm f1.4 seems to have front focusing troubles with the D200 and D300.

    Doug,

    Focusing problems results in soft images, which in the case of the f/1.4 shot, it is soft but it does explain the purple fringing and doubles edges. Somehow, the Extra-low Dispersion (ELD) element is not good enough to correct the CA at f/1.4. At smaller aperture, the ELD worked fine.

    I also have another image (somewhere) which depicted very clearly how bad the coma issue was. The coma issue raised a big question about the aspherical element's ability.
     
  19. I found the image that shows coma ... notice that point light source are pretty round at the foreground but as we sweep our eyes to the rear, they becomes more and more butterfly shaped. Click on image to get details.
    [​IMG]
     
  20. Vivek, did you call Sigma and complain about the chipping paint? On my Sigma, to forestall
    this problem, I placed a rubber "friendship" bracelet around it....
     
  21. Zenon, Nah, that will perhaps distort the bracelet as well. I value my friendship bracelets.

    Arthur, Cool shot. I wonder how this would look like had it been shot wide open and with the added CA and fuzziness.
     
  22. Arthur, have you ever tried the Sigma 28mm f1.8 on the APS-C Nikon D200?
    This is a FF lens that should perform better on the cropped format since it is operating in the center sweet spot. At the time that Sigma came out with it was the best of its series of faster than normal wide angle lenses. The extra few degrees wider would make this lens effective focal length is 42mm a very useful street lens focal length.
     
  23. Here is a handheld wide open with CA and comas....
     
  24. Vivek, you're absolutely right on that rear-cap issue -- it will only go on if you line it up absolutely exactly with the 'open' label aligning with the lens release area. I have a sigma 150mm f/2.8 macro and the rear-cap is awful and has been replaced with a Nikon one. I also replaced the front cap as it was objectionable too (can't be removed with the hood attached).
     
  25. Nice test Arthur, thanks for sharing.

    BTW, how big is the test chart (or rather what distance to the camera)? Did you bracket the focus?

    Peter
     
  26. I guess I don't see what's so bad about the posted photos with "coma" (in the corners, mainly, as I view it). The truth is that any lens at f1.4 is going to struggle a bit. And to Arthur, I understand that you weren't referring to focusing issues in criticizing the Sigma lens. I was bringing up the front-focusing as the issue that I have seen most typically associated with this lens in combination (apparently) with certain camera bodies.

    And yes, Sigma should include pinch-type front caps with their lenses at this point.
     
  27. >BTW, how big is the test chart (or rather what distance to the camera)?

    The distance was about 5 ft. and the chart is a regular US Letter size print.

    >Did you bracket the focus?

    No,I did not.
     
  28. ShunCheung

    ShunCheung Administrator

    Arthur, US letter size is 8.5x11" (approximately 22x28cm). From 5 feet away, a 30mm lens should cover a lot more than that area on a DX sensor. In other words, you are showing some major crops from (the center of?) the original images.
     
  29. They are all focused on the (horizontal and vertical stripes) next to the large #4 on the left most column. I made sure that the focus reticle covered half of the horizontal and the half covered the vertical.

    Distance between the subject and the camera is adjusted accordingly to make sure the focus reticle cover the same amount of the stripes.

    So, yes, they are cropped. But, the original cropped files turned out to be almost the same size due to the effect of my adjusting the distance to get the same Angle of View.
     
  30. BTW, the cropped files were 1000 pixels, which you can see in full in the Flickr link above.
     
  31. Interesting test. I'm impressed by how nicely the Sigma sharpens up when stopped down just a little. There's a dramatic difference between f/1.4 and f/1.8, good enough to make that lens acceptable IMO.

    Glad to hear the paint flakes off easily. Should give me a nice bargaining chip when shopping for a used one. Ugly on the outside, pretty on the inside. ;>
     
  32. Sorry, to be clear, are these corner crops or center crops? I've found the Sigma to be weak at the edges and excellent in the center...
     
  33. >are these corner crops or center crops?

    The crops are pretty central.
     
  34. My copy doesn't seem to produce the purple fringing that Arthur's does. Here's one at f/1.4, shot hand-held with the D40X and bounce flash. It's a 100% crop of an 8.5x11 calendar about 5 feet away.

    http://farm3.static.flickr.com/2367/2194270862_c376ea1e3e_o.jpg
     
  35. Sometimes the things that make a lens look good (or bad) "on-paper" may not be of any value within its intended use. I think that if I want to shoot flat text, I would choose a slow lens with good edge-to-edge sharpness. The value of a lens with a speed in the f/1-point-something range is for those times when, based on the light, that speed can make or break a shot, or for when you can exploit the speed for depth of field considerations. Maybe a more valid test would be a 3D subject in an environment with foreground, middle ground and back ground and see how the lens renders the subject within that environment. As I found with my 35mm f/1.4 Nikkor, a slightly curved field makes for a nice selective focus effect for a person in a room, but it would be the last lens I grab to shoot a test chart.
    00O1gK-41010284.JPG
     
  36. I'm considering the Sigma 30mm/1.4 for my Nikon D40x. Yes I'm new to DSLR or I wouldn't have bought that camera, right? I'm finding the AF-S limitation pretty frustrating.

    These tests don't look good, but other reviews call it a pretty sharp lens. I bought the camera for doing portraits of show-dogs and children so I need the sharpest prime I can afford, and AF is essential.

    Anyone got a better idea?
     
  37. PS Albert, you made a good point I think. I don't want to make pictures of test sheets. You rescued my enthusiasm for getting a prime lens.
     
  38. Kathy, the Sigma is your only AF fast prime choice with the D40x, and I think it's a very good
    lens. I don't know about test charts, but my copy of the lens is crisp and exhibits none of the
    issues listed above. Just make sure to get a good copy. I bought two of these, one for
    myself and one for a friend. One copy was perfect out of the box, the other was a bit soft at
    infinity and Sigma adjusted it at no charge and it's perfect now. I use mine mostly for indoor
    shots of children.
     
  39. Kathy,

    >I bought the camera for doing portraits of show-dogs and children

    I think you probably need a longer lens so that your portraits will look closer to a typical portrait.
     
  40. And, to end this thread with what superlative means at the other end of the spectrum, here's a "similar" focal length:
    <p>
    <img src="http://farm4.static.flickr.com/3174/2370234408_d0ef75a22d.jpg">
    <p>
    Leica M8; 35mm Summicron-M Asph shot wide open at f2
     
  41. At about 7 times in price difference compared to the Nikkor 35mm/2, I cannot say it is worth the price, it depends on your application. I am just showing what test images looked like from a highly corrected lens.
     

Share This Page