3D Audio Encoding with Adaptive Virtual Speaker Reselection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing three-dimensional audio encoding technologies face challenges in maintaining high-quality sound reconstruction due to fluctuations in virtual speakers used for encoding different frames, leading to poor sound quality and increased data and bandwidth requirements.

Innovation Solution

The method involves determining the coding efficiency of an initial virtual speaker for a current frame and, if it falls below a threshold, selecting an updated virtual speaker from a set of candidates to encode the frame, thereby reducing fluctuations and improving sound quality.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If a virtual speaker is used to compress the three-dimensional audio signal, then data compression is achieved, but the reconstructed audio quality deteriorates due to large fluctuations in virtual speaker selection across frames

Engineering Contradiction:
Improvedata amountVSAvoidreconstructed audio quality
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent applies dynamics by making the virtual speaker selection adaptive and frame-dependent rather than static. The encoder dynamically selects virtual speakers for each frame based on sound field characteristics, and the decoder dynamically adjusts reconstruction parameters accordingly. This dynamic adaptation allows the system to maintain audio quality while achieving data compression, resolving the contradiction between compression and quality.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes parameters by introducing multiple virtual speaker candidates with different characteristics (positions, directions, frequency responses) and selecting among them based on frame-specific sound field properties. The system also changes the parameter of virtual speaker configuration data transmission, sending only necessary selection information rather than complete speaker definitions, achieving both compression and quality maintenance.

Inventive Principle:
Principle #35Parameter changes

2Adaptability or versatility

If the virtual speaker selection fluctuates across different frames, then adaptability to varying sound fields is improved, but sound quality stability deteriorates

Engineering Contradiction:
Improvesound field representation capabilityVSAvoidsound quality consistency
Core Design Contradiction:
Adaptability or versatilityVSStability of the object's composition

Solution Approach 1:

The patent implements feedback by having the encoder transmit virtual speaker selection information and configuration data to the decoder, which then uses this feedback to reconstruct the sound field consistently with the encoder's selection. The decoder adjusts its reconstruction process based on the received speaker selection parameters, ensuring that the adaptability to varying sound fields does not compromise sound quality stability.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent applies universality by designing a virtual speaker system that can serve multiple functions: representing different sound field characteristics across frames, maintaining consistent reconstruction quality, and adapting to varying audio content. The same virtual speaker framework handles both the adaptability requirement and the stability requirement through coordinated encoder-decoder operations.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If more virtual speaker candidates are used to improve sound field reconstruction, then coding efficiency is improved, but device complexity increases

Engineering Contradiction:
Improvesound field reconstruction accuracyVSAvoidencoder and decoder complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies segmentation by dividing the virtual speaker selection process into discrete, manageable components: a finite set of predefined virtual speaker candidates, frame-specific selection based on sound field characteristics, and transmission of selection indices rather than complete speaker definitions. This segmentation maintains high reconstruction accuracy while controlling encoder and decoder complexity through structured, modular processing.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4354431B1Three-dimensional audio signal encoding method and apparatus, encoder, and system
Publication Date: 2025.11.19 HUAWEI TECH CO LTD
  • EP4354431B1 patent drawingFigure 1
  • EP4354431B1 patent drawingFigure 2(a)~2(b)
  • EP4354431B1 patent drawingFigure 3

AI summary

A method and an apparatus for encoding a three-dimensional audio signal, an encoder, a system, and a computer program are provided. The method includes: An encoder obtains a current frame of a three-dimensional audio signal (S510); obtains coding efficiency of an initial virtual speaker for the current frame based on the current frame of the three-dimensional audio signal (S520); and if the coding efficiency of the initial virtual speaker for the current frame meets a preset condition, determines an updated virtual speaker for the current frame from a set of candidate virtual speakers (S540); encodes the current frame based on the updated virtual speaker for the current frame, to obtain a first bitstream (S550); or if the coding efficiency of the initial virtual speaker for the current frame does not meet the preset condition, encodes the current frame based on the initial virtual speaker for the current frame, to obtain a second bitstream (S560). Through reselection of a virtual speaker, the method reduces fluctuation of the virtual speaker used for encoding different frames of the three-dimensional audio signal, and thus improves quality of a reconstructed three-dimensional audio signal at a decoder side, and improves sound quality of a sound played at the decoder side.