Fast Noise Source

Similar to the Noise Source block, except it uses less CPU by pre-generating a pool of random variates taken from the specified distribution. At runtime, samples are then uniform randomly chosen from this pool which is a very fast operation.

Supports output of type complex, float, int, short

Args:
 * Noise Type - the random distribution to use, only Gaussian and Uniform are supported
 * Amplitude - the standard deviation of a 1-d noise process. If this is the complex source, this parameter is split among the real and imaginary parts
 * Seed - seed for random generators. Note that for uniform and Gaussian distributions, this should be a negative number.
 * Variate Pool Size - Number of samples to pre-generate