Aspects of spatial audio rendering from spherical microphone arrays

Higher order ambisonics (HOA) has become a popular framework for scene-based spatial rendering of acoustic signals. Common use cases are spherical harmonic modal beamforming, loudspeaker array playback, and binaural playback for VR applications. In most cases, the acoustic sources are mono signals that are placed in space and then transformed into the HOA domain. However, some applications, e.g., re-creating complex acoustic scenes or archiving sound fields, do not lend themselves to acoustic simulations but require the recording with higher-order spherical arrays. Although conceptually identical, there are some important differences between the HOA rendering of a simulated sound field versus a recorded real sound field. The differences depend greatly on the specific spatial audio rending method. This contribution reports on an analysis of the differences between simulated and recorded sound fields based on the rendering method.