Dear Bernhard,
Have a look at the Java applet "Diffraction and Fourier Transform"
located here: http://lcr.epfl.ch/page37304.html
You can draw an image, periodic or not periodic, the application will
compute the FFT. You can then select either phases or magnitudes or
Re/Im components and compute the FFT-1. You can also combine one set of
phases with another set of amplitudes etc. The Fourier spectrum can be
masked to simulate resolution limits, truncation errors etc. Using the
"Advanced" option, you can read in an image file (e.g. jpeg) and then
perform the FFT.
You need to have Java running on you computer.
There are other interesting applets on that page, e.g. "diffractOgram",
which illustrates the Ewald construction. I use many of these applets
for teaching purposes.
best regards
Marc Schiltz
Bernhard Rupp wrote:
> Dear All:
>
> I am trying to make FFTs of images of assemblies of
> spheres and other shapes to explain diffraction, the
> usual thing. So far I do this through cumbersome
> cludges, and I bet there are better ways, and I
> am looking for free or cheap software to do this.
>
> I load the image into basic mathcad, dump it as a grayscale
> array, SFT it with a F90 kludge (s=slow), and load the
> transform back into MC and convert it into an image. However, what
> I also want to do is generate a false color diffraction
> pattern where the color is the phase. I believe this
> is quite similar to the Fourier duck transforms Kevin Cowtan
> made. I also recall a textbook were two crystallographers
> were phase exchanged, so there got to be packages
> that to what I want.
>
> The mathcad image processing module is too expensive for my taste.
>
> Any suggestions for software - maybe there is some software
> out already to simulate diffraction from one, 2, 3 an array of
> objects?
>
> Cheers, br
> --------------------------------------------
> Bernhard Rupp
> www.ruppweb.org
> --------------------------------------------
>
--
Marc SCHILTZ http://lcr.epfl.ch
|