The sonic detection and ranging (SODAR) technology developed previously (Wu and Zhu, JASA, 2013) is automated for performing blind sources localization and separation. In particular, source localization results are displayed in terms of space-frequency or space-time correlations. In other words, one can either view distributions of sound sources in three-dimensional (3-D) space versus any user-defined frequency bands, or distributions of sound sources in 3-D space versus time history. The sound pressure level (SPL) values associated with the identified sources are also calculated and displayed in these space-frequency and space-time correlation graphs. To acquire a better understanding of the distribution of sound sources together with their SPL values in 3-D space, a 3-D viewer using Google SketchUp software is employed to view these results in both space-frequency and space-time correlations. With this 3-D viewer, one is able to rotate and look at any source from any perspective, and zoom in and out to examine the details of relative positions of individual sources. The information on source locations together with windowing and filtering technologies enable us to separate the individual source signals. Examples of using this blind sources localization and separations in a non-ideal environment that involves random background noise and unspecified interfering signals are demonstrated.