Driven to Diffraction

Ever since I got really interested in Landscape Photography, rather than just pointing my camera at a nice view, I’ve been reading about Diffraction. As some of you will know, Diffraction, in photographic terms, refers to the way that the rays of light disperse having passed thorough the aperture of the lens and before hitting the sensor, or indeed film. The effect of this appears as a softening of images, particularly notable on edges The smaller the aperture you use, the greater the effect will be, and more noticeable on the resultant image. For a full explanation of the subject, I can do no better than recommend the excellent article on the Luminous Landscape website here

Now a lot of landscape photography, particularly the ‘big vista’ genre needs a small aperture to get a decent depth of field, as you want both the foreground and background to be in focus. The downside to this is of course that the small aperture results in increased diffraction and thus a softer image, a bit of a catch 22.
As a result, I’ve recently being reluctant to step down my lens below f11 for fear of getting the much diffraction and soft images. I can’t recall why I chose f11, I probably read it in a magazine of the “Grab Lots of Photo’s Monthly” type, or on some website. It certainly wasn’t scientifically arrived at.

Recently I’ve been concerned that some of my landscapes which should have had good depth of field were a bit lacking, and wondered if stopping down the lens really would make that much difference, so I decided to do some tests.

In order to do this I needed some images which had defined edges, shot with each of my three lenses at different apertures. I could then compare these and see at what point the softness became unacceptable.

The brick wall of my garage looked a likely target, so much to the neighbours’ bemusement I took a series of 18 images, 6 on each lens, at a range of full stop apertures.

Because of the different lenses, the largest and smallest apertures differ, but this doesn’t really matter, as it’s the ‘threshold’ of each lens I wanted to find, not a comparison between them. The lenses and apertures tested were:
  • Canon EF-S 18-55mm III set t 35mm: f4.5, f5.6, f8, f 1,f16,f22
  • Canon EF-S 55-250: f4 at 55mm: f5.6, f8, f11, f16, f22
  • Sigma 10-20mm at 20mm: f5.6, f8, f11, f16, f22 ,f32.

All of these were taken on my Canon 600D, at ISO100, with the camera locked down on the tripod and fired using a remote cable to reduce camera shake.

Once into Lightroom, each set of images were sharpened by the same amount, to remove the natural softness you get when shooting in RAW

You can see all of the resultant images here, but in order to get a better Idea I compared them at 100% on screen.
So what did I find?. Well interestingly it does indeed appear I was being unnecessarily pessimistic. As expected all 3 of the lenses show an increase in softness with decreasing aperture. However, in my opinion with both of the Canon lenses the softness is still acceptable at f16, a full stop smaller than my artificial barriers, whilst the Sigma is still acceptable at F22, two stops smaller.

So, next time I’ m out in the field taking ‘big vistas’, I’ll be stopping the lens down and hopefully seeing the benefits. Of course I’m sure I could have found a website which would have told me all this, but it was much more interesting finding out for myself, and it entertained the neighbours!