Preferential Diffraction Spikes (6)

By: Steve Koehler
Last Updated: 1 August 2003

I'm finally having some good success simulating sharp diffraction spikes, Mauro's spider web phenomenon, and spiky in focus stars. My current model of boundary layer turbulence starts with a midpoint displacement fractal (as in Suiter), and filters out all but a fairly narrow range of frequencies to make a random lumpy pattern.

Note: my previous efforts had a mistake in the turbulence function that resulted in sharp ridges at the bottom. Thanks to Nils Olof for pointing out this mistake.

The table below shows a filtering value, a graph of the turbulence function, and star test images defocused -10 waves, -5 waves, and 0 waves for an aperture with -.04 waves RMS of high order spherical aberration. Please note the following things in these results:

• Mauro's bright spider web pattern shows up very clearly in the .04 line.
• The most realistic spikes show up in the .03, .04, and .05 lines, especially for -5 waves defocused.
• There are spikes in the in focus image in the .02 and .03 lines. At .04 and above, the in focus image becomes a bright spot surrounded by a swarm of dots.

I think I finally have a workable model for boundary layer turbulence. Now I can go back and explore in more depth how various aberrations affect how the diffraction spikes show up. Thanks to all who helped me to refine this mode by sharing their ideas.

Filter	Turbulence Function	-10 waves	-5 waves	in focus
.08
.06
.05
.04
.03
.02