The Diffraction aperture limit - Canon VS Nikon!

[Overread]

Yes its a Canon VS Nikon thread! :mrgreen:

However with some sanity I hope and a clear question. First off some ground info for those not aware and to bring everyone to a similar point of understanding regarding macro photography and canon + nikon.

First up aperture reporting - most true macro lenses are f2.8 lenses (a few exceptions exist but by and large most are f2.8); however (as far as I know) all these lenses stop down their actual aperture as they enter their minimum focusing distance and once at their closest focusing point most are around f5.6 in their actual aperture. I am also led to understand that the actual focal length of the lens also changes as part of this process. This process appears to be standard even in macro only lenses such as the canon MPE 65mm - with the actual aperture continuing to diminish as the magnification increases.


However with canon camera bodies this change in actual aperture is not reported to the user and the camera remains reporting an aperture of f2.8 - whilst with nikon the camera body does report the actual aperture change; thus the aperture reading will change to f5.6 (or other if the max aperture was not f2.8).

This brings in an interesting question regarding the diffraction aperture limit (the point at which the loss of sharpness due to diffraction is too great next to the depth of field gained) since the camera bodies are effectively reporting different apertures for both cameras and yet in most cases advice on aperture remains fairly constant between brands.

However I'd like to challenge this and find out if canon and nikon, when shooting macro, have similar or different recommended minimum apertures for macro work. Of course the only way to test this is to compare small aperture shots on similar level camera bodies between each other (ideally with the same macro lens model - eg a 3rd party Sigma, Tamron or Tokina).

I know with my own shooting on canon that f13 is the general point of limitation with f16 occasionally used. However since my actual apertures are 2 stops smaller (because I'm starting at f5.6 and not f2.8 as the camera says) does this also hold true for nikon - and thus mean that a nikon shooter can go down as far as f22 and still get similar image quality as the canon at f13 (yes I know f13 is a 1/3 stop and that it should translate to a 1/3rd stop less than f22, but the wiki article only shows the range as far as f22 )


So have we got any Nikon shooters willing to cast their views on this - maybe even some Nikon + Canon shooters willing/able to make the test?


PS for the purposes of this test we are talking about 1.6 and 1.5 crop camera bodies and not f1.3 or fullframe versions.

 

[Helen B]

If you used a simple macro lens (no floating elements, no physical change in the diameter of the iris as you focus, neither retrofocus nor telephoto) that was set to f/5.6 (which is the aperture when focused at infinity) and left it set to f/5.6 as you focused to 1:1 magnification the effective aperture would now be f/11*. In terms of diffraction, depth of field and depth of focus, and exposure it is now an f/11 lens.

Most Nikon macro lenses correctly report this change (taking into account any changes in pupil magnification/effective focal length as they focus closer). Just because a camera does not report the change doesn't mean that it isn't happening. Once you take the correct effective aperture into account diffraction effects are generally consistent (other factors being the same, such as sensor resolution and format).

Best,
Helen

* Because the effective aperture is a function of the angle subtended by the exit pupil at the point where the lens axis passes through the image plane. The f-number is 1/(2 sin theta), where theta is half the angle subtended by the exit pupil.

 

[Overread]

Thanks Helen!
So that would confirm that if a canon lens can shoot sharp whilst reporting an aperture of f13 (when its actual is just 1/3rd less than f22) then a correctly reporting nikon camera body and lens should be able to set their aperture to f22 and get a similar sharpness (and of course depth of field) as the canon shooter.

Provided the lens retains the properties you list

 

[Overread]

For those after an interactive demonstration have a look at the article here:
Macro Camera Lenses

which has few calculators that give a basic idea of the different apertures and of the canon/nikon difference

 

[Stormchase]

Great topic, I wish i could contribute to it but I'm Canon ... and didnt know about any of this. That said I learned something today wich im happy about.

 

[NateS]

Another thought:

If the actual aperture is f5.6 and the camera reports f2.8......are you positive that anything is even changing when you change the camera setting from f2.8? Is it not possible that the aperture does not actuallyl change but merely catches up to the true aperture? This might be something you want to test for as well.


For me and my Nikon, I still get very sharp images at f/16 but I can tell a difference between f/13 and f/16 (f/13 being sharper). I do shoot f/22-f/25 every so often with also good results, but by f/22 I can see some noticeable drop-off in quality.

I would love to see a sharpness test comparing say a 40d/50d and a D90/D300 both with something like a Tamron 90mm mounted. The other option would be to shoot a DOF chart at each aperture at 1:1 and see how much DOF the Canon has versus the Nikon at each aperture.

 

[Overread]

The first part that you suggest about the aperture "catching up" was something I did wonder about at one point - however even my regular shooting has shown that the aperture us stopping down from wide open so the catchup idea I isn't occuring.

Also whilst its not that accurate you could compare your shots to these ones:
Triple Macro lens test - a set on Flickr
fullsizes (unsharpened aside from RAW default) are online and there are 3 lenses there so a little spread of data to work with. Be interesting to see where you feel your aperture values are in comparison to the sigmas and the canon.

 

[NateS]

The first part that you suggest about the aperture "catching up" was something I did wonder about at one point - however even my regular shooting has shown that the aperture us stopping down from wide open so the catchup idea I isn't occuring.

Also whilst its not that accurate you could compare your shots to these ones:
Triple Macro lens test - a set on Flickr
fullsizes (unsharpened aside from RAW default) are online and there are 3 lenses there so a little spread of data to work with. Be interesting to see where you feel your aperture values are in comparison to the sigmas and the canon.

I can't compare any of my recent shots...I rarely shoot at 1:1 anymore and am almost always beyond that. If I knew what size that coin was, I'd do a test to compare...you think that's like the size of a US penny/nickel/quarter?

 

[Overread]

I've honestly never used US currency to know, though according to a quick google it should be:

25.9 mm diameter, 2.03 mm thick

 

[NateS]

Well, our Quarter is 24.26mm diameter and 1.75mm thick so that's probably the closest I'll find. I will try to do similar shots of a quarter with my Tamron 180mm at various apertures to gauge the DOF. I will post results when I find time to do so.