Abstract: The methods described herein are configured to accurately propagate sound through a virtual environment. During runtime of an application using the virtual environment, a shortest path is calculated from a sound source in a first zone to a sound destination in a second zone that passes through at least one portal. The direction of the calculated shortest path from the portal to the sound destination is determined. Then, a virtual sound source is generated based on the length and determined direction of the calculated path. Further, obstructions in the virtual environment are processed to determine an attenuation of audio data associated with the sound source. The generated virtual sound source and the attenuated audio data are provided to an audio engine for rendering to a user. Propagating sound using a combination of these methods provides an accurate, realistic sound experience in the virtual environment.