Pink Noise Generation

[DSPstudent123]

Hello everybody,

I'm trying to figure out how to generate pink noise and I was wondering if somebody could help me out a bit. I'm currently having problems with the Voss-McCartney algorithm for generation of pink noise. (If it matters I'm using this web site as a reference: The Spectrum Produced by the Voss-McCartney Pink Noise Generator ).

As I understand, the idea is to filter white noise and gain pink noise, but that is puzzling me.

In the above paper, an algorithm for computing the PSD (Power spectral density) of pink noise. However, I am uncertain on how to use this information to construct the actual pink noise. Can you please help me with this?

Your help will be greatly appreciated ,
E

 

[permute]

The noise is updated every sample.
+ The noise is updated every second sample
+ The noise is updated every fourth sample.
+ The noise is updated every fourth sample.
+ The noise is updated every sixteenth sample.
= The output of the pink noise generator is the sum of N rows, numbering 0 to N-1.

so that's N different noise source running at N diadically spaced rates, fs, fs/2, fs/4, fs/8, etc...

also, "Note that there is a small amount of ripple (about 1dB) which does not go away regardless of the number of rows added. ", which itself could probably be improved if the something other than the zero-order-hold kernel was used.

 

[zorro]

Hi E,

Please look at this thread on the same subject:
https://www.edaboard.com/threads/170541/
Regards

Z