(Tuan - Thanks for sending the link, I appreciate it.)
I usually avoid these sorts of discussions because: a) I approach photography primarily from an empirical standpoint, and b) they can lead to an affliction which Jeff Cooper has termed P.I.I. (Preoccupation with Inconsequential Increments). However, in this instance I think there may be some practical reasons to discuss what seems like a purely theoretical circumstance.
Yes, the standard formula which N. Dhananjay was probably referring to (r=0.00061 f/number) does indeed seem to indicate that f-stop is all that matters. There are some caveats, however. The first is that the formula commonly thought of as "the diffraction formula" does not apply to the entire diffraction pattern, only the most significant part (the radius of the Airy disk in the center of the pattern). There are other components involved here - such as the rings around the Airy disk - which are usually ignored due to the fact that they are 12 times (3 1/2 stops) less bright. (* see pg. 68 of "Image Clarity - High Resolution Photography" by John B. Williams.) Nevertheless they are there and contribute to image degradation.
In a more practical vein, it may be more accurate to look at the correlation between diffraction and f-number as something analogous to a film curve. The majority of the graph is a straight line, with a constant correlation between image degradation and f-number. However, there is a definite "toe" to the curve, and - I posit - a "shoulder". For the toe you can refer to the thought experiment I described above (miniature camera such that even at f/2 the absolute size of the aperture is smaller than the wavelength of visible light and thus suffers from extreme image degradation due to diffraction.) For a more practical example of this, see the illustrations (Figures 1-3 [A] and [B]) on page 4 of Ansel's "The Camera". It shows a photo made with a 1/64" aperture (pinhole), then the same image made with a 1/100" aperture. This is only a 1-stop (approx.) reduction in aperture size, yet there is a HUGE loss of resolution with the smaller aperture, much more than you would typically see with "normal" aperture sizes. So, obviously at some point the absolute physical size of the aperture has an effect on diffraction, beyond simply the relative dimensions of aperture and focal length.
At the other end of the scale, I believe it is also important to look at the system as a whole (i.e. absolute physical size of aperture and image) and not just the ratio of opening to focal length. As the format increases (i.e. physical aperture and image size increase) diffraction becomes much less of an issue than many people would think by just looking at the standard diffraction-induced formula. Let's hear from someone far more knowledgeable than I on this subject: Leslie Stroebel. In Sec. 3 (Image Formation) of "View Camera Technique" (5th Ed.) he talks about avoiding apertures which can result in "...an objectionable loss of resolution". More interestingly, he talks about comparable f-numbers between formats which would result in similar diffraction-induced loss of resolution. He starts with something we are all familiar with; f/16 on 35mm. To get to even this point with 4x5, you have to stop down to f/64. On 8x10 it takes f/128. And to bring this whole thing full circle to what started it all (diffraction worries - and my cavalier lack thereof - on 12x20), Stroebel states that with 16x20 (close enough for me) you would have to stop down to f/256 to have a problem. 'Nuff said.
Cheers!
Bookmarks