3
$\begingroup$

I am trying to generate a band limited noise signal on a Raspberry Pi. I calculate samples at every 1/48000 sec on the Pi. If I generate normally distributed random numbers as data points at every sample would it be gaussian noise signal. Here is what i am having a hard time to understand, how to interpret the frequency of the noise signal. And also how to limit the frequency values to be between a band?

$\endgroup$
2
  • $\begingroup$ I couldn't understand you. Are you trying to generate Gaussian Noise at sampling rate of 48000 Hz? $\endgroup$
    – Royi
    Jun 20, 2014 at 8:40
  • 3
    $\begingroup$ Note that "white" and "gaussian" are two different things (it is not clear from your question if you want to generate white noise or noise with a gaussian distribution, or white gaussian noise). "White" means that each sample is correlated only with itself (autocorrelation function is 1 at 0, 0 everywhere else), and this depends on the quality of your random number source. "Gaussian" refers to the distribution of samples values. $\endgroup$ Jun 20, 2014 at 10:40

1 Answer 1

2
$\begingroup$

If you are generating sample of white noise (Normal distribution value) with frequency of 48 kHz, then you already have this in frequency range of 0 - 24 kHz (at least you should if generator is truly random). If you use same samples with different sampling frequency, then they will also be a white noise.

Now if you want to make it a band-limited white noise, then you must use a digital filter. Choice is up to you, but for sake of simplicity I would recommend FIR filter. You must use it to filter in real time your incoming samples. How to do it? Well, that's totally up to you, you can even use short buffer to perform filtering by linear convolution for each generated sample but it won't be very efficient.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.