View Full Version : Pink noise vs. White noise
kcarter
08-21-2001, 07:40 AM
I was always told that pink noise is the equal combination of all frequencies in the audio spectrum, and white noise is (theoretically) the equal combination of ALL frequencies ... in theory, from DC to light.
Ok, if this is the case, my system should sound the same with pink or white noise, because being an audio system, it cannot reproduce anything outside of the audio spectrum. So, what are the differences between the pink & white noise generators in the "Signal Generator" RTAS plug-in?
The pink sounds very bassy, and the white sounds very high. Which should I be using for EQing my room?
Thanks for help on this.
F Umminger
08-21-2001, 08:32 AM
White noise has a flat spectrum when using a linear frequency scale, with constant energy per Hertz. The spectrum of pink noise has a -3dB/octave slope, or constant energy per octave. It looks flat when using a logarithmic frequency scale.
Pink noise is usually used for equalizing rooms because it shows up as a flat line on standard 1/3 octave band analyzers.
Be warned that the pink noise in the Signal Generator is not very accurate.
kalle74
08-21-2001, 09:05 AM
This is how it goes.
Pink noise:
Random noise (hiss) that has equal energy in each octave
White noise:
Similar to pink noise, except that white noise contains equal energy at each frequency point
Why does white noise sound so "hissy"? Because high frequencies take less energy than lows to be audible, when the energy is equal at all bands, highs tend to overpower the lows.
As for tuning your room, use pink noise, for that is much closer to music in itīs frequency content.
Also, be aware that white noise at high levels is capable of frying your tweeters if you donīt have a protection circuit in your monitors ( most active monitors have it, though ).
kcarter
08-22-2001, 10:59 AM
Thanks Frederick for the info, and especially for the honesty regarding the accuracy problems with the pink noise part of the Spectrum Generator. -- Just curious, is the white noise accurate?
Kalle74 - thanks for the layman's interpretation, and the clear answers. (Have you ever considered teaching???)
I will EQ my room with the Signal Generator's pink noise, and check it with a hardware pink-noise generator I have. If they match, that is one less piece of too-big hardware to find a place to mount.
BTW - I plan to use the Waves Analyzer plug-in to check my room. Does anyone know if it has the -3dB/octave slope which I believe FUmminger was implying standard 1/3 octave band analyzers have?
Thanks for any guidance -- and for those interested, I plan on miking the room with an AKG C-1000. Hopefully flat enough to work, as it is the best mic I have.
F Umminger
08-22-2001, 03:14 PM
The white noise in the Signal Generator is accurate. It is not random enough for cryptography images/icons/smile.gif, but it is spectrally flat.
kalle74
08-22-2001, 05:33 PM
Originally posted by kcarter:
Kalle74 - thanks for the layman's interpretation, and the clear answers. (Have you ever considered teaching???)
Oh yes, I have. In fact, I used to be a sound engineering teacher a few years ago...
Glad I could help.
Eric Heil
08-23-2001, 04:14 PM
It's very interesting that this topic should come up right now.
I am having a major problem with my reference monitors. I am using Alesis .7's and my budget is too tight to get something better right now. I found out, much after the fact that the .7's were more aptly suited for Video Post and consequently they are a little light in the bottom which either causes me to under compensate the bottom or over compensate the top.
I can live with a "slightly" inacurate signal generator if it get's me within the ballpark!
Here are my questions:
1. Do I need to purchase/rent a spectrum analyser (about $550.00 from Beringher and about $300.00/day to rent a pro analyser) or can I do this with my rig somehow?
2. If I can do it with my rig, what Mic should I use (much cheaper to rent a good Mic)?
3. If I can do it on my rig, what's the procedure?
Thanks,
Eric Heil images/icons/confused.gif
KamaSutra77
08-23-2001, 05:55 PM
yes, what exaclty could you do with a good mic with a known response curve and the built in pink noise genrator to tune the room, speaker placement and mix eq base settings? oh yes and some sort of anylyzer.
FUMoney
08-26-2001, 11:25 AM
Go to www.siasoft.com. (http://www.siasoft.com.) They have a computer based RTA that works pretty well. I used it on my system,(I have Event 20/20bas's) and it worked great. I used an AKG414 mic. The company is owned by EAW so I guess they're pretty serious audio people, oh and did I mention, you can download a 30 day trial version! Good luck!
marcusb
08-29-2001, 10:42 PM
Personally I wouldn't recommend tuning your room using a 1/3octave graphic, or any sort of graphic for that matter.
However it does pay off to do a pink noise test and analyse your room's response from your mix position then try to optimise by moving your monitors adding furniture and or basic acoustic treatment (diffusors, absorbers). When you've done your best an RTA can assist you in identifying if the room is exaggerating any particular frequencies.
Yet another tool to further train your mixing skills images/icons/smile.gif
Cheers,
Marcus
vBulletin® v3.8.5, Copyright ©2000-2010, Jelsoft Enterprises Ltd.