generates a vector of samples for t seconds at frequency freq Hz.
samples=soundsec(t [,freq])
a 1-by-1 matrix of positive real numbers, the time length to sample.
a 1-by-1 matrix of positive real number, the frequency of samples per second.
default value is 22050 Hz
Generates a vector coding time from 0 to t seconds at sampled rate freq.
Samples are always generated to fit in [0,t[ (excluding upper bound).
Value 0 for either t or freq return an empty matrix
