HOA Sound Field Compression with Residual Decorrelation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for compressing Higher Order Ambisonics (HOA) representations suffer from high bit rates and perceptual coding noise due to high cross-correlations between HOA coefficient sequences, leading to inefficiencies and perceptible errors, especially when assumptions about sound field composition are violated.
Innovation Solution
The method decomposes the HOA representation into dominant directional signals and a residual component, transforms the residual into a discrete spatial domain, predicts the residual from dominant signals, reduces its order, decorrelates the residual, and encodes both components perceptually to minimize noise and errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If direct encoding of individual HOA coefficient sequences using perceptual coding is employed, then compression is achieved, but perceptual coding noise occurs due to high cross-correlations between HOA coefficient sequences
Solution Approach 1:
The HOA representation is divided into two separate components: dominant directional signals and residual ambient signals. This segmentation allows each component to be processed differently, with dominant signals being decorrelated to reduce cross-correlations before perceptual coding, thereby reducing perceptual coding noise while maintaining compression efficiency.
Solution Approach 2:
The patent applies decorrelation processing to the dominant directional signals, transforming them into a domain where cross-correlations are minimized. This parameter transformation enables more efficient perceptual coding by reducing the harmful interactions between correlated signals, thus lowering perceptual coding noise.
2Measurement precision
If the number of expansion coefficients is increased to improve spatial resolution, then spatial resolution improves, but the bit rate increases quadratically
Solution Approach 1:
By separating the HOA representation into dominant directional components and residual ambient components, the patent enables selective processing. The dominant components, which carry the most spatial information, are decorrelated and efficiently coded, while the residual components can be represented with fewer coefficients, thus maintaining spatial resolution while reducing overall bit rate.
Solution Approach 2:
The patent extracts and separately processes the dominant directional signals from the full HOA representation. By isolating these significant components for special processing (decorrelation and perceptual coding), the system achieves efficient compression of the most important spatial information while reducing the burden on the remaining residual components.
3Adaptability or versatility
If HOA representation is used instead of channel based methods, then independence from specific loudspeaker set-up is achieved, but decoding process complexity increases
Solution Approach 1:
The patent performs decorrelation processing on the dominant directional signals during the encoding phase, preparing the signals in advance for efficient decoding. This preliminary processing reduces the complexity of the decoding operation by pre-organizing the data in a form that requires less computational effort during playback, while maintaining the adaptability benefits of HOA.
Data Source
AI summary
The invention improves HOA sound field representation compression and decompression. A decoder decodes compressed dominant directional signals and compressed residual component signals so as to provide decompressed dominant directional signals and decompressed time domain signals representing a residual HOA component in a spatial domain. A re-correlator re-correlates the decompressed time domain signals to obtain a corresponding reduced-order residual HOA component. A processor determines a decompressed residual HOA component based on the corresponding reduced-order residual HOA component, and determines predicted directional signals based on at least a parameter. The processor is further configured to determine an HOA sound field representation based on the decompressed dominant directional signals, the predicted directional signals, and the decompressed residual HOA component.