PDA

View Full Version : What Exactly is "White Noise"?


EagleEye
10-25-2001, 09:50 PM
The title pretty much well covers it. My grandma had one of those "Soothing noise makers to help you sleep" deals. If I remember right, white noise is kinda like "chrchrchrchrchrchrchrchr" evenly mixed together. What makes it?

Duck Duck Goose
10-25-2001, 09:59 PM
The technical definition (http://www.its.bldrdoc.gov/fs-1037/dir-040/_5873.htm) of white noise is:

<ahem>

Noise having a frequency spectrum that is continuous and uniform over a specified frequency band. (188) Note: White noise has equal power per hertz over the specified frequency band. Synonym: additive white gaussian noise.

However, any $17.50 Wal-Mart box fan will make lovely additive white gaussian noise, and it's considerably cheaper than one of those "ocean sounds" generators.

Muad'Dib
10-25-2001, 10:00 PM
Well, I have heard Kid Rock describe himself as such...

But if I remember right it is all frequencies at equal volume, sounds like "very fine" static. Maybe rushing wind is a better description.

scr4
10-25-2001, 10:09 PM
The "static" from an AM radio when it isn't tuned to any station is a pretty decent white noise. It's called "white" because white light contains equal amount of all colors (frequencies) of light, more or less. Similarly, white noise contains all frequencies of sound more or less equally.

In electronics people talk about "pink" noise which is like white noise, but has more low-frequency ("red") than high-frequency stuff.

scotth
10-25-2001, 10:21 PM
pink noise is defined as having the same about of energy in each octave.

Since lower octaves have a narrower band, each freqency in that band must carry more energy. To the ear, pink noise "sounds more even" than white noise.

CalMeacham
10-25-2001, 10:36 PM
The "static" from an AM radio when it isn't tuned to any station is a pretty decent white noise. It's called "white" because white light contains equal amount of all colors (frequencies) of light, more or less. Similarly, white noise contains all frequencies of sound more or less equally.



As one of my optics professors pointed out, it's not trye that "whitelight" has all frequencies represented equally -- no matter what definition of "white light" you use. Ordinary daylight is pretty close to a blackbody spectrum, in which the peak occurs in the infrared, and isn't tolerably flat even over the visible regime alone. Light from fluorescent sources are characterized by a series of peaks. I don't know of any illumination circumstance under which you get equal contributions from ll wavelengths -- it would be interesting to try it out on people and see what they think. In any event, none of the CIE standard illuminants are flat across the visible. When they do color perception studies, they're doin it with "weighted" light.

"White Noise" in sound or electronics, however, is flat across a specfied band. Some sound or electronics engineer was probably making a false analogy.

justwannano
10-25-2001, 11:31 PM
[quote]
The "static" from an AM radio when it isn't tuned to any station is a pretty decent white noise. It's called "white" because white light contains equal amount of all colors (frequencies) of light, more or less. Similarly, white noise contains all frequencies of sound more or less equally.[/quote

I believe that should be FM.The no signal sound is a constant SSSHHHHHHHHHHHHH

vanilla
10-26-2001, 08:28 AM
white noise?
I believe theres an album of it: Metal machine Music by Lou Reed.

RM Mentock
10-26-2001, 11:23 AM
Originally posted by CalMeacham
As one of my optics professors pointed out, it's not trye that "whitelight" has all frequencies represented equally -- no matter what definition of "white light" you use.
Unless of course you use that as your definition of white light! :)
Originally posted by CalMeacham
[B]Ordinary daylight is pretty close to a blackbody spectrum, in which the peak occurs in the infrared, and isn't tolerably flat even over the visible regime alone.
This site (http://micro.magnet.fsu.edu/primer/lightandcolor/sources.html) shows a graph (http://micro.magnet.fsu.edu/ primer/images/lightsources/sources.jpg) that makes it seem like ordinary sunlight is reasonibly flat across the spectrum.

CalMeacham
10-26-2001, 01:27 PM
This site shows a graph that makes it seem like ordinary sunlight is reasonibly flat across the spectrum.

I couldn't get your site to load. But all of my plots (including my RCA Electro-Optics Handbook, open in front of me, show quite a bit of variation -- almost a factor of two in the Spectral irradiance from the Sun or a Blackbody (although the atmospheric absorption cuts that down quite a bit at sea level).

OxyMoron
10-26-2001, 01:57 PM
Lawrence Welk.

RM Mentock
10-27-2001, 02:23 AM
Originally posted by CalMeacham
I couldn't get your site to load. But all of my plots (including my RCA Electro-Optics Handbook, open in front of me, show quite a bit of variation -- almost a factor of two in the Spectral irradiance from the Sun or a Blackbody (although the atmospheric absorption cuts that down quite a bit at sea level).

Dang. Somehow I ended up with a space in the graph URL (http://micro.magnet.fsu.edu/primer/images/lightsources/sources.jpg). Anyway, just eyeballing that graph shows that the two references could be the same. The energy at 400nm and at 700nm is about half that between 500 and 600 where it is fairly flat. Still, if you were to fit it with a horizontal line, the values would be +/- 33%. That's not bad.

I wonder what color we'd say it was, if it were perfectly flat.

August West
10-27-2001, 11:58 AM
Pssst. Over here. Wanna know a secret?

At one of my company's buildings, we generate white noise and pipe it throughout the building. It is meant to sound like the climate control system pushing air through the ducts, but in fact it is electronically created sound.

Why do we do it? I haven't a clue, but part of my job is maintaining the system and have been told to keep it under my hat. Which, of course, leads to:
Random cow-orker: "What are you working on?"
Me: "Nothing..go away..ummm, I'm not here."

But seriously, does anyone know why my company does this?

Napier
10-27-2001, 04:18 PM
Since you ask what "exactly" white noise is, I think it is better to modify some of the above explanations by specifying that by definition white noise would have equal energy content per delta frequency over all frequencies (implying infinite energy of which almost all is above any arbitrary frequency), and that practical imitations of white noise would meet this description over some finite frequency band. It would be easier to define true white noise as that signal whose fourier transform was a constant nonzero value, and practical white noise as that signal whose fourier transform was a constant nonzero value from zero to the highest frequency of interest.
White noise isn't necessarily sound, and in many circumstances where someone cares about it, it is electronic. A modem emits a pretty fair rendition of white noise over the telephone frequency spectrum when it is transmitting compressed information (any predictability in the signal would represent a less-than-optimal compression scheme). Thus there's actually an intuitive reason why modems sound like rockets or wind or rushing water when they are in the middle of something.

Derleth
10-27-2001, 10:23 PM
Originally posted by August West
Pssst. Over here. Wanna know a secret?

At one of my company's buildings, we generate white noise and pipe it throughout the building. It is meant to sound like the climate control system pushing air through the ducts, but in fact it is electronically created sound.

Why do we do it? I haven't a clue, but part of my job is maintaining the system and have been told to keep it under my hat. Which, of course, leads to:
Random cow-orker: "What are you working on?"
Me: "Nothing..go away..ummm, I'm not here."

But seriously, does anyone know why my company does this?

Budgetary concerns and applied psychology.

If your cow-orkers think a climate control system is in operation, they're less likely to seriously complain about the lack of heat or cooling. And white noise machines are cheaper than actual climate control, especially for big buildings. It's the placebo effect.

Now, if your company does something like replacing the candy in the vending machines with little pellets, and reorganizing the cubicles into a little mazes that lead to the vending machines, maybe you should consider looking for the black helicopters. Or the men in white lab coats.

:D

saudade
10-27-2001, 11:39 PM
On "I Want You (She's So Heavy)", on Abbey Road, the Beatles used a "white noise generator" in the long outro.

To me it sounds a like the static on an old monaural radio, nothing special. Great song though, but I wonder they must have used for white noise.

Anthrax made an album called "The Sound of White Noise" a few years ago, with John Bush instead of Joey Belladonna as vocalist. No actual "white noise" on that one, as far as I know.

Is there "red noise"? I thought that was what occured in guitar feedback, or with a theramin...I'm talking out of my ass here.

erislover
10-27-2001, 11:52 PM
Pink noise generators are often installed in businesses in order to keep privacy even in open areas. How does it work? As mentioned previously pink noise has equal power over the spectrum (on average because remember, its supposed to be random), and so when you are in an area with a good pink noise generator it keeps a constant pressure on your eardrums, drowning out sounds from further away.

They work remarkably well when installed properly (used to install them!) but a generator alone isn't enough for perfect effect. The rooms themselves should have some sound-dampening properties (like plush carpets and heavy-fiber cubicles) for maximum effect. A perfect system won't allow you to hear either the pink noise itself, nor any coworkers who aren't directly talking to you (or hwo are more than say 5 feet away).

White noise generators are useful for general and directional frequency response testing. We always carried white noise generators around when we installed systems to tweak the equalizers in the final product (used to install all sorts of sound equipment).

Rysdad
10-27-2001, 11:56 PM
"Ice, Ice, Baby"

CalMeacham
10-27-2001, 11:57 PM
I wonder what color we'd say it was, if it were perfectly flat.


This is actually pretty straightforward to work out. Get a copy of Warren J. Smith's book Modern Optical Engineering and go to the chapter on color, and he gives full instructions on how to calculate the CIE chromaticity coordinates of a given spectrum, with examples (One example everyone does is to figure the color of the sky, assuming only Rayleigh scattering. It's blue.). If I were clever enough I could probably tell you this off the top of my head. Since I'm not clever enough, it'll have to wait until Monday.