Comments on “Primary-Ambient Extraction Using Ambient Spectrum Estimation for Immersive Spatial Audio Reproduction”

Haisheng Lu,Jiangnan Liang,Chuang Shi
DOI: https://doi.org/10.1109/taslp.2023.3328286
2024-01-01
Abstract:In the above paper, He et al. propose a primary-ambient extraction method using ambient phase estimation with a sparsity constraint (APES). The primary-ambient extraction problem is formulated as a $L1$-norm minimization of the primary component with respect to the phase of the ambient component, based on a stereo signal model they have assumed. Discrete searching (DS) is proposed as the approach for solving the APES in each time-frequency bin, which is computationally complicated and potentially imprecise. This correspondence provides an analytical solution to the APES by exploring the geometric relation between the primary and ambient components in a rotated coordinate.
engineering, electrical & electronic,acoustics
What problem does this paper attempt to address?