An experiment with log normal density distributions (simple, pretty obvious) and a test of including ipython notebooks in blog posts. To do this, I created a gist (https://gist.github.com/3750781) and used the ipython notebook viewer (http://nbviewer.ipython.org/3750781/) to render the notebook in HTML. It worked pretty well; all I had to do was c&p the source and then change 2 lines of the style (background color -> black, foreground color -> white for the markup... maybe I should do it for the rest too).

Comparison of the derived power spectra and density PDFs for lognormal power distributions. Obviously (if you think about it), the power spectra are always flat for randomly distributed signal. The PDFs change significantly.