Adaptive Spherical Transform for HOA Audio Decorrelation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for encoding and decoding Higher Order Ambisonics (HOA) audio signals result in increased coding noise due to high cross-correlation between channels, particularly after matrixing operations, which is exacerbated by transformations to the spatial domain prior to compression, leading to noise unmasking effects.

Innovation Solution

Implement an adaptive Discrete Spherical Harmonics Transform (aDSHT) that decorrelates channels using an inverse adaptive DSHT, followed by perceptual encoding and transmitting rotation information, minimizing noise unmasking by adapting the spatial sampling grid to the signal properties.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If independent perceptual coders are applied to individual Ambisonics coefficient channels, then compression is achieved, but coding noise increases after matrixing operation due to high cross-correlation between channels

Engineering Contradiction:

Improvecompression efficiencyVSAvoidcoding noise

Core Design Contradiction:

Loss of energyVSObject-affected harmful factors

Solution Approach 1:

The patent applies a preliminary spatial transform (DSHT) and decorrelation processing to the HOA signals before compression. This preliminary action reorganizes the channel structure and reduces cross-correlation between channels, so that when independent perceptual coders are subsequently applied, the coding noise does not get amplified during the matrixing operation. The decorrelation step is performed in advance to prevent the noise unmasking problem.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent changes the parameter representation by transforming from the conventional Ambisonics coefficient domain to a spatial domain using Discrete Spherical Harmonics Transform. This parameter transformation allows the signal to be represented in a different basis where cross-correlation between channels is reduced, thereby improving compression performance and reducing coding noise artifacts.

Inventive Principle:

Principle #35Parameter changes

2Adaptability or versatility

If HOA signals are transformed to spatial domain by Discrete Spherical Harmonics Transform prior to compression, then spatial processing is enabled, but noise unmasking effects are exacerbated

Engineering Contradiction:

Improvespatial processing capabilityVSAvoidnoise unmasking

Core Design Contradiction:

Adaptability or versatilityVSObject-affected harmful factors

Solution Approach 1:

The patent applies a preliminary decorrelation processing step after the DSHT transformation. This decorrelation operation reorganizes the spatially transformed signals to reduce inter-channel correlations before compression. By performing this decorrelation action in advance, the patent prevents the exacerbation of noise unmasking effects that would otherwise occur during subsequent matrixing operations.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary processing stage between the DSHT transformation and the compression/matrixing operations. This intermediary decorrelation step acts as a mediator that processes the spatially transformed signals to reduce cross-correlation, thereby preventing the noise unmasking problem from occurring during the final matrixing operation.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Adaptability or versatility

If conventional DSHT is used for spatial encoding, then spatial transformation is achieved, but large amount of side information must be transmitted

Engineering Contradiction:

Improvespatial encodingVSAvoidside information overhead

Core Design Contradiction:

Adaptability or versatilityVSLoss of information

Solution Approach 1:

The patent extracts only the essential parameters needed to define the spatial sampling grid configuration. Instead of transmitting complete grid position information for all channels, the method identifies and transmits only the key parameters (such as rotation angles and grid type identifiers) that allow the receiver to reconstruct the same spatial sampling grid. This extraction of essential parameters significantly reduces side information overhead.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent discards redundant spatial configuration information that can be inferred or reconstructed at the receiver side. By identifying which parameters are necessary to define the spatial sampling grid and which can be derived or assumed, the method transmits only the minimal necessary information, achieving efficient compression of side information while maintaining full spatial encoding capability.

Inventive Principle:

Principle #34Discarding and recovering

Data Source

PatentEP4660999A1Data rate compression of higher order ambisonics audio based on decorrelation by adaptive discrete spherical transform

Publication Date: 2025.12.10 DOLBY INTERNATIONAL AB

Figure 1~3
Figure 4a~4b
Figure 5a~6

AI summary

A method for encoding multi-channel HOA audio signals for noise reduction comprises steps of decorrelating (81) the channels using an inverse adaptive DSHT, the inverse adaptive DSHT comprising a rotation operation (330) and an inverse DSHT (810), with the rotation operation rotating the spatial sampling grid of the iDSHT, perceptually encoding (82) each of the decorrelated channels, encoding rotation information (SI), the rotation information comprising parameters defining said rotation operation, and transmitting or storing the perceptually encoded audio channels and the encoded rotation information.