How To Generate Pink Noise?

oshifis · 7th January 2012, 07:06 PM

I would like to computer generate pink noise, and burn it on CD. Is there an easy and exact way to do it? I have seen several software where part of the functions is pink noise, but I am not sure how accurate they are. Especially the crest factor (peak to average ratio, I need 12 dB), frequency range (should be 5...10 Hz to 22 kHz within fractions of a dB), and peak level (should be near 0 dB). It can be PRBS-based with several seconds repetion period, if that is simpler. Any suggestions?

Ron E · 7th January 2012, 07:11 PM

Quote:

Originally Posted by oshifis

I would like to computer generate pink noise, and burn it on CD. Is there an easy and exact way to do it? I have seen several software where part of the functions is pink noise, but I am not sure how accurate they are. Especially the crest factor (peak to average ratio, I need 12 dB), frequency range (should be 5...10 Hz to 22 kHz within fractions of a dB), and peak level (should be near 0 dB). It can be PRBS-based with several seconds repetion period, if that is simpler. Any suggestions?

I am not sure how to specify or measure crest factor in pink noise, but audacity comes with a pink noise generator. Generate the pink noise, save as wav and burn to CD. Noise isn't really "flat" except over long averaging times, especially at low frequencies.

Ron E · 9th January 2012, 04:45 PM

The foobar2000 media player has a dynamic range add in component that seems to measure peak and RMS dB in sound files. I don't think it exactly correlates to crest factor, though.

A sine wave should have a 3dB crest factor and it reports 0dB, and a square wave should have a 0dB crest factor and it reports -3dB. So when it measures 10dB on Audacity pink noise, perhaps it is really 13dB...

Pano · 9th January 2012, 05:30 PM

I generated Pink Noise in Audacity V1.3 and it seems to be an RMS average of -12.22dB with the peaks at -0.94dB. Very close to what you are looking for.

oshifis · 10th January 2012, 02:40 PM

Thanks for all. If I don't want (trust) these commercial software, is there any algorithm for calculating the individual samples? Also I want a brickwall filter below 20 Hz and above 20 kHz, how do I realize it? My goal is to create a test CD, and get a ruler-flat FFT plot with 1/12th or 1/24th octave bandwidth setting within this range. Then compare the FFT plot of the WAV file against the FFT plot of the analog audio output when playing back the CD on a standalone CD player. Just for playing around...

Pano · 10th January 2012, 02:56 PM

I don't think you'll get a ruler flat response this way, unless you use very long averaging. Maybe someone else knows better.

If you want a noise signal and a good FR graph of it, I would suggest HOLMIpulse and ARTA. Both can use a noise signal for FR. ARTA uses either Pink or White, HOLM uses a modified MLS. Although normal use is thru the ins and outs of a soundcard, you can save the signal, burn it to CD, record it and measure. A bit of work, but with a CD player I would not know how else to do it.

The nice thing about these softwares is that they know what signal to expect (their own), therefore you don't need a "perfect" pink noise.

phofman · 10th January 2012, 03:10 PM

Quote:

Originally Posted by oshifis

Thanks for all. If I don't want (trust) these commercial software, is there any algorithm for calculating the individual samples? Also I want a brickwall filter below 20 Hz and above 20 kHz, how do I realize it? My goal is to create a test CD, and get a ruler-flat FFT plot with 1/12th or 1/24th octave bandwidth setting within this range. Then compare the FFT plot of the WAV file against the FFT plot of the analog audio output when playing back the CD on a standalone CD player. Just for playing around...

Did you look at sox SoX - Sound eXchange | HomePage ? It is the most versatile tool for playing with audio I know of. Its pinknoise generator does not have an option for crest factor, but the source code is freely available and rather simple to read (plain C).

The brickwall filter is simple to do with sox, see SoX

7th January 2012, 07:06 PM	#1
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	How To Generate Pink Noise? I would like to computer generate pink noise, and burn it on CD. Is there an easy and exact way to do it? I have seen several software where part of the functions is pink noise, but I am not sure how accurate they are. Especially the crest factor (peak to average ratio, I need 12 dB), frequency range (should be 5...10 Hz to 22 kHz within fractions of a dB), and peak level (should be near 0 dB). It can be PRBS-based with several seconds repetion period, if that is simpler. Any suggestions?

9th January 2012, 04:45 PM	#3
Ron E diyAudio Member Join Date: Jun 2002 Location: USA, MN	The foobar2000 media player has a dynamic range add in component that seems to measure peak and RMS dB in sound files. I don't think it exactly correlates to crest factor, though. A sine wave should have a 3dB crest factor and it reports 0dB, and a square wave should have a 0dB crest factor and it reports -3dB. So when it measures 10dB on Audacity pink noise, perhaps it is really 13dB... __________________ Our species needs, and deserves, a citizenry with minds wide awake and a basic understanding of how the world works. --Carl Sagan Armaments, universal debt, and planned obsolescence--those are the three pillars of Western prosperity. —Aldous Huxley

10th January 2012, 02:56 PM	#6
Pano diyAudio Moderator Join Date: Oct 2004 Location: North Carolina Blog Entries: 4	I don't think you'll get a ruler flat response this way, unless you use very long averaging. Maybe someone else knows better. If you want a noise signal and a good FR graph of it, I would suggest HOLMIpulse and ARTA. Both can use a noise signal for FR. ARTA uses either Pink or White, HOLM uses a modified MLS. Although normal use is thru the ins and outs of a soundcard, you can save the signal, burn it to CD, record it and measure. A bit of work, but with a CD player I would not know how else to do it. The nice thing about these softwares is that they know what signal to expect (their own), therefore you don't need a "perfect" pink noise. __________________ Take the Speaker Voltage Test!

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Does Pink noise follow the V x V/R method?	OMNIFEX	Solid State	18	5th January 2009 03:46 PM
What does pink noise sound like?	jamikl	Multi-Way	23	31st July 2008 02:23 AM
Pink Noise generator	Lyra	Analog Line Level	12	22nd March 2006 09:30 PM
Digital Pink Noise generator...?	Lyra	Digital Source	12	22nd March 2006 02:06 AM
level of recorded pink noise	MCPete	Everything Else	6	17th January 2006 10:16 AM


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

9th January 2012, 05:30 PM	#4
Pano diyAudio Moderator Join Date: Oct 2004 Location: North Carolina Blog Entries: 4	I generated Pink Noise in Audacity V1.3 and it seems to be an RMS average of -12.22dB with the peaks at -0.94dB. Very close to what you are looking for.

10th January 2012, 02:40 PM	#5
oshifis diyAudio Member Join Date: Mar 2004 Location: Budapest, Hungary	Thanks for all. If I don't want (trust) these commercial software, is there any algorithm for calculating the individual samples? Also I want a brickwall filter below 20 Hz and above 20 kHz, how do I realize it? My goal is to create a test CD, and get a ruler-flat FFT plot with 1/12th or 1/24th octave bandwidth setting within this range. Then compare the FFT plot of the WAV file against the FFT plot of the analog audio output when playing back the CD on a standalone CD player. Just for playing around...

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read