Notes from Under Sky

Diffraction Vanes

What is the critical parameter for diffraction—perimeter or area?

I did a little experiment back in September 2003. In an effort to determine, at a coarse level, the effect of spider vanes on the image quality, I made two "vanes": one from human hair, and one from a thick piece of wire (about 1.5 mm thick). I first put the hair on the front of the scope, east-west. (The scope is my C5+, equatorially mounted.) I aimed it at Fomalhaut, and then at Mars. No diffraction spikes were visible on either target.

I next took the wire vane, and attached it to the scope, north-south. Faint diffraction spikes now appeared on Fomalhaut, aligned east-west. The spikes on Mars were much more evident. When I removed the wire vane, the spikes went away. At no time could I detect any north-south spikes which would presumably come from the hair vane.

Finally, I tried to assess the effect on the detail visible on Mars, resulting from both vanes. First, with both vanes on, the image was still quite detailed. I could see the pointy tip of Syrtis Major, and the ragged edge of Sinus Sabaeus, leading over to Sinus Meridiani. There appeared to be a cleft in the South Polar Cap, at about longitude 320, but that might have been just a contrast effect. (It turns out it wasn't. There was in fact a cleft exactly where I saw it.) There was plenty of other detail that I sketched briefly, but won't repeat here.

When I took the vanes off, and looked again, I found that there was no detail that I could see now that I could not see before. I couldn't tell whether the detail that was there now was easier to see than before, but there was no new detail.

The limiting magnitude in the area of Fomalhaut and Mars was about magnitude 4.0, plus or minus 0.2. The seeing was about 1 arcsecond and fast, but still bulk—not quite blur. It still made it difficult to see detail more than half of the time.

Because I may be biased—I am sufficiently aware of optical theory to know the predicted effect of obstructions—I suggest that others try this same (or similar) experiment to see if they can judge the effect on visible detail. However, I remain convinced that whatever the relative effect of central obstruction and spider vanes, the crucial parameter is not edge length, as is so frequently claimed. If it were so, the hair vane would have diffracted just as much as the wire vane. And that conclusion is in accordance with diffraction theory, of course.

Copyright (c) 2004 Brian Tung