This paper presents software development and research on the field of digital audio synthesis of spectral data using the Pure Data environment (Miller Puckette et al) and the ATS spectral analysis technique (by Juan Pampin [6]). The ATS technique produces spectral data using a deterministic-plus-stochastic representation. The focus is on the methods by which such data may be real-time read and synthesized using several Pure Data externals developed by the author and others, as well as on the involved audio synthesis strategies. All the software involved in this development is GNU Licensed and runs under Linux.