[AH] Red noize

[Martin Czech]

:::The colors are taken as an analogy to light, white noise is like white
:::light, it contains all colors of the spectrum (=all wavelenghts)
:::Pink noise contains less HF: Power spectrum decreases 6 dB/oct i think
:::My guess is that red noise or whatever color is band pass filtered noise.

if you look at the white noise spectrum, you see that it is flat. It
should have equal power per Hertz bandwidth.

Pink has equal power per octave (e.g.) , i.e. -3dB/Octave characteristic,
this is not flat.

It is used as a measurement signal for power measurements that operate
with log frequency scales. Any decade will contain as much power then
it's neighbors. White noise would have more and more power in the upper
ranges, because a 10000-20000Hz Octave has 10000Hz bandwidth, a 10-20Hz
octave only 10Hz.

Since any pole will lead to 6dB/oct an approximation network is needed
to filter white noise into pink. That is the measurement world.

Besides from the very measurement needs the "pink raving" seems
not to make much sense, you can tailor your noise if you have a
decent parametric equalizer. So I guess a good white noise
source is what you really need to get things started,
the rest is filtering with filters you allready have.

red, blue, brown noises are not standard, i.e. there is no common
aggreement what it should be, except that blue emphasizes on higher,
red on lower frequencies, brown is a bit more unclear, power supply
experts mean transients as well as noise, it seems that brown
noise has correlated things in it, i.e. not perfectly random.
I'm not so sure about the brown stuff. Anyone?


Idea : (may be stupid)

Shift register noise has the advantage to be really white in the interesting
bandwidth, but it is not really random. How about using a junction avalanche
(that may not be white noise, depends on current, surface effects i.e. devices
etc.) to xor some of the shift registers inputs? I guess the shift register
will scramble (modulate) the avalanche pulses to something white,
but the sequence will be more random, there may be some correlation in it,
but it will never be the same, i.e. no cycles.

2nd Idea:

I've read about mathematical theory about poles with rational order,
or in general network transfer functions with non integer order.

There is no reason why math should prefer x**2 instead of x**2.18973,
we do this every day using the pocket calculator.
Has anything useable grown out of this roots since 1970??

m.c.

making noise again