This paper presents a method for generating procedural audio using a form of granular synthesis where several aspects of the sound can be controlled dynamically in real-time. Original recordings are partitioned into small parts, a.k.a. grains, that are stored in a multi-dimensional database. Using a weighted random selection of grains, a natural output signal can be generated without repeating grains. The granular databases can be stored in a newly defined file format based on the open storage format HDF5, which means that granular databases can be created and shared for re-use in multiple productions. The method described in this paper has been adopted into the new MPEG-I Immersive audio standard for 6DoF VR and AR audio.
(See document for exact affiliation information.)
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member Join the AES. If you need to check your member status, login to the Member Portal.
