the x,p quadratures of the target wires follow the following multinormal distribution
the covariance should be $\sigma_b+\sigma_m$ when performing random sampling.
impossible for obtaining the quadrature x and p samples via one single homodyne measurement angle, therefore the output results should not contain x and p at the same time.
Inconsistence covariance results for two ways of homodyne measurements