About white noise, pink noise, and brown noise

[gopirate]

hey i'm newbie here and i have some questions about pink noise, white noise, and brown noise.
i don't know what they are.
is there anybody can tell this to me?
point of them, and how they work.
thank you

 

[JimmyS1969]

Um, try Google. There you will find a bounty of treasure. Arrrr!

 

[ecc83]

hey i'm newbie here and i have some questions about pink noise, white noise, and brown noise.
i don't know what they are.
is there anybody can tell this to me?
point of them, and how they work.
thank you

White noise is the signal produced by any electronic device that has real resistance (this means a "perfect" capacitor and inductor produce no noise). It has a flat power/frequency response and the nose extends from DC to infinite frequency!
There isn't a lot of it for small value resistors, if you terminate a mic amp with 150 Ohms that resistor will generate about 0.23microvolts (-130dBu) which IS tiny but sets the noise floor for everything else and shows why the dynamic range of 24bit recording of -144dB is way out of reach!

White noise was usually generated by a reverse biased transistor junction but is today done digitally (when it is strictly "quasi" white noise) White noise generators are generally made to supply a filter to produce Pink noise. This has a power that falls at 3dB per octave and simulates much closer the power spectrum of music. The company where I worked use pink noise to drive amplifiers for heat rise and longevity tests.

Noise is a handy signal for general signal tracing and is less annoying than tone but be wary of putting high noise levels, especially white noise, thru amplifiers and tweeters, the amp might not be as tough as the ones I tested!

Unless you have good grounding in the use of noise and spectrum analysers (I don't!) it is of little use to the noob recordist.

I now await a slap from the Top Chaps here!

Dave.

 

[RAMI]

Not directing this at the OP, because there's nothing wrong with knowledge.

But I have to say that, on a scale of 1 to 100, the imprtance of knowing any of this is about .5, to me. I've been recording on and off for about 25 years and I have to say, I've never needed to know the difference and I still don;t know what any of these different types of "noises" are and why they exist. Again, there's nothing wrong with knowing, but I'd say there are probably about 3, 587 things more important when it comes to knowing things about recording.

 

[ecc83]

Not directing this at the OP, because there's nothing wrong with knowledge.

But I have to say that, on a scale of 1 to 100, the imprtance of knowing any of this is about .5, to me. I've been recording on and off for about 25 years and I have to say, I've never needed to know the difference and I still don;t know what any of these different types of "noises" are and why they exist. Again, there's nothing wrong with knowing, but I'd say there are probably about 3, 587 things more important when it comes to knowing things about recording.

Could not agree more. They are essentially test tools and my job was testing stuff for the most part. (I have put together a "studio" for my son).

Noise sources are more useful to recording peeps than say a gamma ray source, but not a lot.

Dave.

 

[miroslav]

About the only time I use pink noise in the recording environment is during the initial setup and level calibration of the studio monitors. I've messed around comparing the results between using white or pink noise for monitor calibration, and there is a difference in the outcome.

I've maybe needed to do that a few times only over the years....like when making changes to the physical layout.
It's also used for the same purpose in live sound work, and as Dave said, for a variety of audio testing purposes.

So it's not of any critical importance on any daily basis as RAMI said, but if/when you do need to use it, it's good to understand what the differences is and why you would use one type of noise over the other when doing a calibration, especially in the studio, where a lot of importance is placed on the accuracy of the monitor system.

And yeah....just Google it....there's a ton of info already out there....

 

[c7sus]

Analog synths use noise as both a sound source and a modulation source.

Moog uses a hybrid white/pink noise on their Voyagers.

Brown nose is usually attributable to having much too close proximity to management ass.

 

[ecc83]

Ah! Monitor calibration! I had forgotten that Miroslav.

This is something the noob should investigate as soon as they have a stable hardware set up. A Sound Pressure Level meter can be had for $20 that is perfectly adequate for the task, get one with a C weighting if at all possible. Yes, the noise "colour" is going to have an effect but THE most important thing is that you monitor at consistent levels, day by day and not as your guts guide you depending upon the genre and whether or not you like the song.

Audacity can give you pink noise.

Dave.

 

[Bobbsy]

I agree there's no pressing need to know about various types of noise in order to record--but there's nothing wrong with a bit of curiosity. Similarly, if you Google "white noise" and "pink noise" and so on, many of the results are badly written or difficult to understand.

So...now's my chance to write something equally badly!

What is noise?

There are lots of things that can be considered noise in terms of an audio signal--hum, buzz, clicks and so on. However, what we're talking about here is purely random signal with all audio frequencies present (over time). To get complicated, it's generated by random thermal motion of electrons. When the noise is at a high enough level, we perceive it as a hiss--like we hear when a radio isn't tuned to a transmitting station.

White Noise

White noise is the simplest form of this random noise. Every frequency (Hz) is at exactly the same amount of energy. However, this is a bit of a problem if you want to use the white noise to calibrate things. The span from 50 Hz to 100 Hz is one octave; 100 Hz to 200 Hz is another octave and so on. At higher frequencies, 10,000 Hz to 20,000 Hz is also an octave. Since every frequency has identical energy, you can see that the energy level for an octave is higher each time the frequency goes up. I don't know if you've played with the maths of dB measurements but this means each successive octave is 3dB louder than the one beneath it--meaning that the noise will sound kind of harsh and "toppy". Displayed on a frequency graph, white noise will be a diagonal line running upwards from left to right.

Pink Noise

Pink noise starts off exactly the same as white noise but is passed through a filter (believe it or not, called a "pinking filter"). What this does is know the energy down by 3dB per octave so every successive octave has the same amount of energy. Displayed on a graph, this'll look like the kind of flat line you want when you check the specs of a mic or speaker. (Yeah, yeah, I know that some subtle peaks in the right place can make a mic or speaker sound better subjectively but you get my drift.)

Once you have the same energy per octave, the noise becomes useful for the alignment and calibration of gear, particularly speakers. When I'm setting up a live sound system, I'll play pink noise through the speakers and analyse the frequency response with a "real time analyser". This gives me a display of the frequency response coming from the speakers and any diversion from a straight, flat line can be balanced out using equalisation.

That said, the pink noise and RTA only give a starting point. What can look perfect on an RTA can still sound like rubbish--listening to real music and making is sound good is still the final stage of a set up.

Brown Noise

...just to confuse you, has nothing to do with the colour Brown. It's named after the scientist Robert Brown who discovered Brownian motion--and it's noise created by exactly that. It has no real application in audio work (at least not outside test labs) but, if you're interested, is noise where the energy REDUCES as the frequency goes up, this time by 12 dB per octave.

Dang. I can see why so many Google hits make this all seem complicated. It's hard to both keep it simple and keep it accurate at the same time.

Anyhow, hope it helps.

 

[ecc83]

I agree there's no pressing need to know about various types of noise in order to record--but there's nothing wrong with a bit of curiosity. Similarly, if you Google "white noise" and "pink noise" and so on, many of the results are badly written or difficult to understand.

So...now's my chance to write something equally badly!

What is noise?

There are lots of things that can be considered noise in terms of an audio signal--hum, buzz, clicks and so on. However, what we're talking about here is purely random signal with all audio frequencies present (over time). To get complicated, it's generated by random thermal motion of electrons. When the noise is at a high enough level, we perceive it as a hiss--like we hear when a radio isn't tuned to a transmitting station.

White Noise

White noise is the simplest form of this random noise. Every frequency (Hz) is at exactly the same amount of energy. However, this is a bit of a problem if you want to use the white noise to calibrate things. The span from 50 Hz to 100 Hz is one octave; 100 Hz to 200 Hz is another octave and so on. At higher frequencies, 10,000 Hz to 20,000 Hz is also an octave. Since every frequency has identical energy, you can see that the energy level for an octave is higher each time the frequency goes up. I don't know if you've played with the maths of dB measurements but this means each successive octave is 3dB louder than the one beneath it--meaning that the noise will sound kind of harsh and "toppy". Displayed on a frequency graph, white noise will be a diagonal line running upwards from left to right.

Pink Noise

Pink noise starts off exactly the same as white noise but is passed through a filter (believe it or not, called a "pinking filter"). What this does is know the energy down by 3dB per octave so every successive octave has the same amount of energy. Displayed on a graph, this'll look like the kind of flat line you want when you check the specs of a mic or speaker. (Yeah, yeah, I know that some subtle peaks in the right place can make a mic or speaker sound better subjectively but you get my drift.)

Once you have the same energy per octave, the noise becomes useful for the alignment and calibration of gear, particularly speakers. When I'm setting up a live sound system, I'll play pink noise through the speakers and analyse the frequency response with a "real time analyser". This gives me a display of the frequency response coming from the speakers and any diversion from a straight, flat line can be balanced out using equalisation.

That said, the pink noise and RTA only give a starting point. What can look perfect on an RTA can still sound like rubbish--listening to real music and making is sound good is still the final stage of a set up.

Brown Noise

...just to confuse you, has nothing to do with the colour Brown. It's named after the scientist Robert Brown who discovered Brownian motion--and it's noise created by exactly that. It has no real application in audio work (at least not outside test labs) but, if you're interested, is noise where the energy REDUCES as the frequency goes up, this time by 12 dB per octave.

Dang. I can see why so many Google hits make this all seem complicated. It's hard to both keep it simple and keep it accurate at the same time.

Anyhow, hope it helps.

As you say Bobbs', hard to K.I.S.S! As you might expect Douglas Self has some words on this and in addition to the other "colours" he lists Blue, Red, Violet, Grey and Green noise!

Note too that there is no such thing as a "3dB per octave" simple filter. The first order R+C is 6dB/octave and you have to build a more complex kludge to get the right shape. Duggy of course shows we how.

Dave.

 

[Bobbsy]

Happily I don't have to build my own pinking filter. I just push a button marked "pink noise" and it works as if by magic!

Frankly, I don't think I've ever actually used white or pink noise in my home studio--it's a live sound tool for me and, as noted, even there human ears get the final say.

However, you can have some fun with it when building science fiction-y sound effects by adding all sorts of effects to it!

 

[wtfuhz]

I thought the brown noise was a frequency that made you shit your pants

 

[wannaknow]

I agree there's no pressing need to know about various types of noise in order to record--but there's nothing wrong with a bit of curiosity. Similarly, if you Google "white noise" and "pink noise" and so on, many of the results are badly written or difficult to understand.

So...now's my chance to write something equally badly!

What is noise?

There are lots of things that can be considered noise in terms of an audio signal--hum, buzz, clicks and so on. However, what we're talking about here is purely random signal with all audio frequencies present (over time). To get complicated, it's generated by random thermal motion of electrons. When the noise is at a high enough level, we perceive it as a hiss--like we hear when a radio isn't tuned to a transmitting station.

White Noise

White noise is the simplest form of this random noise. Every frequency (Hz) is at exactly the same amount of energy. However, this is a bit of a problem if you want to use the white noise to calibrate things. The span from 50 Hz to 100 Hz is one octave; 100 Hz to 200 Hz is another octave and so on. At higher frequencies, 10,000 Hz to 20,000 Hz is also an octave. Since every frequency has identical energy, you can see that the energy level for an octave is higher each time the frequency goes up. I don't know if you've played with the maths of dB measurements but this means each successive octave is 3dB louder than the one beneath it--meaning that the noise will sound kind of harsh and "toppy". Displayed on a frequency graph, white noise will be a diagonal line running upwards from left to right.

Pink Noise

Pink noise starts off exactly the same as white noise but is passed through a filter (believe it or not, called a "pinking filter"). What this does is know the energy down by 3dB per octave so every successive octave has the same amount of energy. Displayed on a graph, this'll look like the kind of flat line you want when you check the specs of a mic or speaker. (Yeah, yeah, I know that some subtle peaks in the right place can make a mic or speaker sound better subjectively but you get my drift.)

Once you have the same energy per octave, the noise becomes useful for the alignment and calibration of gear, particularly speakers. When I'm setting up a live sound system, I'll play pink noise through the speakers and analyse the frequency response with a "real time analyser". This gives me a display of the frequency response coming from the speakers and any diversion from a straight, flat line can be balanced out using equalisation.

That said, the pink noise and RTA only give a starting point. What can look perfect on an RTA can still sound like rubbish--listening to real music and making is sound good is still the final stage of a set up.

Brown Noise

...just to confuse you, has nothing to do with the colour Brown. It's named after the scientist Robert Brown who discovered Brownian motion--and it's noise created by exactly that. It has no real application in audio work (at least not outside test labs) but, if you're interested, is noise where the energy REDUCES as the frequency goes up, this time by 12 dB per octave.

Dang. I can see why so many Google hits make this all seem complicated. It's hard to both keep it simple and keep it accurate at the same time.

Anyhow, hope it helps.

Thanks Bobbsy, I've wondered about the noise thing for a while. All the suff I've read on google about the subject made my one remaining functional brain cell go onto shock.

 

[Tom Hicks]

the guitar player in my brother's band has built a control room for his studio, and i'm going over there today with some pink noise .wav files, a ECM8000 measurement mic and several RTA programs to see how well his sound treatment has done in evening out the room response.

as mentioned, pink noise has the same amount of energy per octave, so its a great test tone for measuring room response curves. the omni measurement mic will be placed in the listening position pointing up at ear height. we can then adjust the speaker angles, move furniture etc to see what looks flattest on the RTA. i have Room EQ Wizard and Voxengo Span.

 

[grimtraveller]

Shiver me timbers !!

hey i'm newbie here and i have some questions about pink noise, white noise, and brown noise.
i don't know what they are.
is there anybody can tell this to me?
point of them, and how they work.

White noise ~ punk rock, heavy metal rock
Pink Noise ~ the strange sound effects on "The piper at the gates of dawn" and "Saucerful of secrets"
Brown noise ~ Jimi Hendrix's live guitar jams, free jazz by the likes of Albert Ayler, Pharoah Sanders and Sun Ra
They work by stimulating the senses and many would argue that they have no point.

thank you

And thank you too, me hearty.

 

[Tom Hicks]

the measurement session in the control room went great. we used pink noise wave file imported into reaper on a hackintosh. the ECM8000 omni measurement mic was on a stand in the listening position (pointing up at ear height) running through an X32 mixer firewired into the computer.

while the reaper track with the wavefile played the pink noise through the monitor speakers in the control room, the measurement mic captured that room sound onto another reaper track, which had the plugin Voxengo Span instantiated.

on playback of the room track, we hit hold to freeze the waveform and scrolled the mouse over SPAN to identify 5 bumps or room nodes in the 20 Hz - 1KHz range. the room already has bass traps and wall panels, so this is really fine tuning the remaining low rumble. he will use further room treatment to smooth out the rest of the low end room reflections we found.