-
Noise 001
Complex documents about cacophony
This section is from the document '/email-lists/khoros/Khoros.010'.
From donohoe@chama.eece.unm.edu Thu Dec 2 11:49:36 1993
Status: RO
X-VM-v5-Data: ([nil nil nil nil nil nil nil nil nil]
[nil nil nil nil nil nil nil "donohoe@jemez.eece.unm.edu" "donohoe@jemez.eece.unm.edu" nil nil nil "^From:" nil nil nil nil nil nil nil]
nil)
Return-Path: donohoe@chama.eece.unm.edu
Received: from chama.eece.unm.edu (chama.eece.unm.edu [129.24.24.12]) by nx1.soils.umn.edu (8.6.4/8.6.4) with SMTP id LAA25045 for; Thu, 2 Dec 1993 11:49:32 -0600
Received: by chama.eece.unm.edu (4.1/EECE/Dec 1 1993)
id <9312021603.AA26414@chama.eece.unm.edu>; Thu, 2 Dec 93 09:03:34 MST
Received: from jemez.eece.unm.edu by chama.eece.unm.edu (4.1/EECE/Dec 1 1993)
id <9312021603.AA26410@chama.eece.unm.edu>; Thu, 2 Dec 93 09:03:29 MST
Received-Date: Thu, 2 Dec 93 09:03:29 MST
Received: by jemez.eece.unm.edu (4.1/VISION/14Apr88/C)
id AA11340; Thu, 2 Dec 93 09:23:47 MST
Posted-Date: Thu, 2 Dec 93 09:23:47 MST
Message-Id: <9312021623.AA11340@jemez.eece.unm.edu>
From: donohoe@jemez.eece.unm.edu
To: khoros@Soils.Umn.EDU, vprakash@cs.uml.edu
Subject: random uniform noise
Date: Thu, 2 Dec 93 09:23:47 MST
In "random uniform noise", Scott Wilson writes:
+
+try vggauss from the image processing library. Seems to work fine.
+
V.G. Prakashreplies:
>
> This generates a gaussian noise. Is this same as a 2dimensional random
> uniform noise (or white noise)?
Whether a random variable is "Gaussian" or "uniform" refers to the shape
of the distribution function. "Whiteness" refers to the shape of the
autocorrelation function: truly white noise should have a spike at
zero in the autocorrelation function, or, equivalently, a flat
Fourier spectrum.
Both uniform and Gaussian random variables, as produced by random
number generators, are usually uncorrelated (white). The one you
pick depends on physical process you want to model. Many random physical
processes are approximately Gaussian. With Gaussian RVs you need to
specify a mean and variance, however.
p(x) ^
|
|++++++++++++++++++
|
|
|
-----------------------> x
Uniform
p(x) ^
|
| +
| + +
| ++ ++
|++ ++
-----------------------> x
Gaussian
-- Greg Donohoe
=======================================================================
Gregory W. Donohoe, Assistant Professor
Electrical and Computer Engineering
University of New Mexico Tel: 505-277-6724
Albuquerque, NM 87131 Internet: donohoe@chama.eece.unm.edu
"This space intentionally left blank."
=======================================================================


