Diffraction effects by f stop on D800

Discussion in 'Nikon' started by rconey, Jun 26, 2013.

  1. [​IMG]I was taking pictures of some mushrooms when a fly walked up, giving me the chance to look at diffraction effects on a D800 as I varied the f stop. I was pleasantly surprised at the usability up to f16 or even f22. This was on a tripod. The lens is a favorite, although not one o the sharpest- the Series E 75-150 f3.5 MF lens on an extension tube for close focus. Approximately 135 mm. The f22 shot is a little darker, but the hairs on the fly are only somewhat lost. The depth of field is much greater, which would be the reason to use such a small aperture. There is a larger version of this on my photo.net page in the contributions section. The larger image is the f22 version.
     
  2. Sometimes what you lose in 'sharpness' is made up by increased depth-of-field, after all is said and done.
     
  3. Interesting, Robert - thanks for sharing. On a related note, I recently read this on the subject of sharpening and the ability to recover diffraction-softened images. It's not going to stop me feeling worried about going below f/6.3 on my D800E (I paid for the "E", I want my money's worth!) but it's good to know how bad things get. To be honest, I get more worried about dust on the sensor at that aperture - one good thing about shooting at f/2 the whole time (on my D700 - I stop down for sharpness on the D800) was that it hid sensor spots.
     
  4. Long ago with slide film I did a similar sort of experiment using a Nikkor 105 /2.5 lens (because it goes up to F32 and has few, if any, other vices), with the camera on a tripod, shooting burdocks on a bright snowy day. I found that the diffraction is detectable all right, if there's sufficient magnification, but that for normal viewing of an uncropped photograph it's hard to spot, and its extent varies with light and contrast, most noticeable in edge lighting. My somewhat unscholarly conclusion was that while there's a good enough reason to favor middle stops when you can, it does not pay to sacrifice needed depth of field for fear of diffraction.

    I'm glad to see someone doing this with digital equipment, and have often wondered why so few people avail themselves of the free shots to experiment this way.
     
  5. [​IMG]
    f32 was very soft so I did not include it above. here it is.
     
  6. I shoot at f/11 frequently. Don't worry about it. It still blows 24 MP cameras away.
     
  7. I recently did a similar run (sorry, images not handy at the moment) while using a 60/2.8 Micro on the D600 with its paltry, embarassing 24MP sensor. I was genuinely surprised at how little impact diffraction appeared to be having (we were shooting jewelry) even up past f/22. In practical terms, a good exposure and only casual skills with contemporary post-production tools can mitigate that sin far better than I'd have guessed.
     
  8. Diffraction (or lack of it) is just one more reason to choose full-frame over DX. If you do the same test with a DX body you'll see that diffraction becomes noticeable about a stop wider, all other things being equal.
    Anyway, I think that being able to resolve the hairs on a fly at that magnification is plenty good enough, whether they're crisply defined or otherwise.
     
  9. paltry, embarassing 24MP sensor.​
    And yet I somehow clearly remember when the Nikon digital shooters were saying that 6MP were more than enough!
    Not that they weren't mostly right. :eek:
     
  10. "Diffraction (or lack of it) is just one more reason to choose full-frame over DX. If you do the same test with a DX body you'll see that diffraction becomes noticeable about a stop wider, all other things being equal."
    But you also gain about a stop's worth of DoF using DX, so it would seem it evens out.
    Kent in SD
     
  11. "But you also gain about a stop's worth of DoF using DX" - That's not a good thing if you actually want a shallow depth-of-field Kent. And I did qualify it with "all other things being equal" by which I meant not swapping the lens for a shorter focal length to get extra D-o-F.
     
  12. My takeaway message was simple. When these machines came out there was a lot of talk about the need to limit to f5.6 or lower to avoid diffraction effects. What I see is that this is not true. The diffraction effects are at a very minute detail level. If the depth of field is needed (as in landscapes) the loss of minute detail will not be seen until very large prints are produced. As in film from the old days, use the f stop you need for the depth of field.
     
  13. In general, if everything is done correctly, and there's no camera or subject movement to worry about, details on the plane of focus in a photo taken at f/5.6 will look sharper than details on the same plane taken at f/11. But, it's not a big enough difference that you shouldn't ever take a photo at f/11 or even higher. If you need to see detail in front of an behind the POF, a smaller aperture will be critical.
    In certain instances, focus stacking will be an option, but only when nothing in the scene moves (including light). Obviously, you can't focus stack portraits or wildlife or street scenes or a bride throwing a bouquet.
     
  14. Diffraction becomes more of an issue with shorter focal length lenses as the actual diameter of the "pupil" of the aperture blades becomes smaller and smaller. It is less noticeable in longer focal length lenses.
     
  15. If you look at graphs such as these..
    http://www.lenstip.com/upload3/3525_roz.jpg for the Nikon 40mm macro or
    http://www.lenstip.com/upload3/2563_Nik105mm_res.jpg for the Nikon 105mm VR
    There is a very noticable fall-off in res @ small apertures.
    Those images posted above appear to make these graphs questionable in the way I (we?) interpret them.
     
  16. Hmmm, variables that are of interest: lens focal length -wide angle to telephoto, distance of subject to camera (these are closeups of about 10 inch). Any others? I suppose print size. sharpening techniques. Iso settings.
     
  17. Shooting technique is very important for extreme crops/post sized prints, as is post processing software and technique.
     
  18. I don't think there are really any surprises here; it is just that the degree of magnification is not enough to show up the diffraction effects.
    By my measurements the fly on the original above is around 8 mm wide while on the enlargement it is around 25 mm wide, a magnification of 3. The original picture is 20 cm wide so the enlargement is part of a picture 60 cm or 23.6 inches wide. Most monitors resolve about 100 dpi so the little images are part of a picture 2360 pixels wide or a total of 1573 * 2360 = 3.7 Mp. 3.7 Mp is way below the capability of a D800. Anyone care to check my analysis?
     
  19. There are no surprises in that on my monitor with the original raw images I see the same detail out to f16. There is a gradual softening that would require large prints to see. It is ok to use fstops up to f 16 or f22 if depth of field requires it. The loss of very fine detail is only apparent at very large print size. When I read the Digital Photography (dpreview.com) review of the D800/D800E again it is clear that they reached the same conclusion. With careful sharpening even that loss of detail can be lessened.
     

Share This Page