1/f noise (was: RE: super cables)

[Paul Perry]

At 02:53 PM 16/07/99 -0400, Terry Bowman, KA4HJH
>"The Mac Doctor" wrote:

>1/f should be a "random walk", and if that's whats happening you can
>imagine why there would be problems with measurement.
>
It isn't a 'random wark' in the generally accepted sense.
A 'random walk' is usually generated by taking a fixed value, and adding 
+x or -x to it randomly. 

For the fractal or 1/f equivalent of this, you add or subtract a random
variable with the property that the size of the step is inversely 
proportional to its size. So you have many little steps, and few large steps.

The two sound completely different.

A sequence of notes with frequency following a 1/f distribution sounds errily
like 'music', except that in traditional music there is a tendency for 
large jumps downward to be more common than large jumps upward.
Raw 1/f is of course symmetric in this regard.

paul perry Melbourne Australia