Dynamic Stochastic Synthesis, proposed by Iannis Xenakis in 1992, consists in the stochastic variation of a waveform on each cycle. Inspired by this method, a Diffusion Dynamic Stochastic Synthesis method is proposed in this article by solving a diffusion equation to simulate the process with particles and mapping their positions to amplitude values of a waveform. An implementation in Max/MSP is also shown and the results are compared with a simplified version of conventional DSS, analyzing the spectrum of the signals generated by both methods where, despite the similarity in the overall form of the frequency analysis, noticeable differences are found. In order to show the potential of the method proposed as a musical application in the form of an instrument. This article concludes summarizing the work, analyzing the main results and finally discussing future work.