An experimental investigation of criteria for continuous variable entanglement

The criterion for Gaussian continuous variable entanglement is experimentally compare in the presence of loss, to a criterion for observation of the EPR-paradox and demonstrated qualitative and quantitative differences. The differences arise from the sensitivity of the EPR-paradox to state mixedness, a property that the inseparability criterion is entirely insensitive to. Entanglement on a diagram of the sideband photon number required to generate the entanglement - a property synonymous to inseparability - versus the excess sideband photons is presented. Efficacy contours of some common quantum information protocols on the diagram is displayed.