Daniel Boyd Fox wrote:
> Okay, I'm going to put forward a crazy idea and why I think the idea is not so crazy....
... Doppler shifted white noise still sounds the same, but there may
> well be another nonsinusoidal radio wave that will stand out in the noise and not be effected by doppler shift.
> Anybody care to dig into the idea?
> Daniel Fox
I don't know if this works specifically to your "crazy idea" but it may
apply as another form of filtering white noise. Bear with me...
I analyze vibration from electric motors and determine their condition
based partly on a historical trend but mostly with pattern recognition.
To do this I have several averaging modes with which I am able to
The first is called Synchronous time averaging. This essentially
filters out all vibration or white noise that is not an integer multiple
of a specific frequency. In vibration analysis this frequency is
typically the turning speed of the motor but it could also be set to the
hydrogen line, 0.1 of the hydrogen line or whatever.
The second averaging mode I have been thinking of is called Negative
Linear Averaging. This guy collects 2 sets of data (in this example)
from the same point and simply subtracts 1 from the other. An example
would be if an amplitude of 1 exists at frequency x in the first set and
an amplitude of 1.1 exists in the second at the same frequency x then
the result is 0.1 (absoloute) at frequency x. If something is present
in the white noise that was'nt there a minute ago then it will stand
I like your idea Daniel with the fractals. It sounds sorta like 1/2 or
1/3 harmonics-is that right?. please post or forward me more info on
them or a source where I can get it from.