Parametric coding of sound fields into directional and diffuse components is a common approach to spatial up-mixing of low-order Ambisonic signals and impulse responses. In this contribution we investigate parametric representations of sound fields considering one or two directional components per time-frequency bin. Additionally, a purely isotropic diffuse stream or a first-order directional ambient stream is considered to account for residual spatial components. A binaural listening experiment was conducted, comparing third-order parametrically synthe-sized scenes to a third-order Ambisonic reference. The improvement gained by a second directional component per time-frequency bin is mainly observable under anechoic conditions. The variants with a first-order directional ambient stream better account for residual spatial components in cases where not all direct sound sources can be represented by directional components and under reverberant conditions. This makes them well-suited as baselines in further studies on parametric coding and upmixing of sound fields and potentially useful as training targets for deep neural network approaches in parametric coding and upmixing.
(See document for exact affiliation information.)
https://aes2.org/publications/elibrary-page/?id=22205
(362KB)
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.
