Spatial Audio Encoding with Energy-Based Direction Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing spatial audio encoding methods face challenges in optimizing compression of spatial metadata at varying bit rates, particularly at lower bitrates, leading to suboptimal accuracy in quantizing multiple spatial directions when multiple concurrent directions are present.
Innovation Solution
The proposed method modifies energy ratios by determining a quantization spatial resolution based on modified energy ratios, using codebooks with varying resolutions for different energy ratios, and adjusting quantization based on diffuse-to-total and direct-to-total energy ratios to optimize accuracy and bit usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If uniform quantization resolution is used for all direction parameters, then encoding complexity is reduced, but quantization accuracy deteriorates for directions with different energy ratios
Solution Approach 1:
The patent applies local quality by using different quantization resolutions for different direction parameters based on their energy ratios. Specifically, direction parameters with higher energy ratios (more important directions) are quantized with higher resolution, while directions with lower energy ratios use lower resolution. This is achieved by determining quantization resolutions adaptively for each direction parameter based on its associated energy ratio, thereby optimizing overall perception quality without uniformly increasing complexity.
2Measurement precision
If higher quantization resolution is used for all direction parameters, then spatial accuracy is improved, but bitrate increases
Solution Approach 1:
The patent changes the parameter of quantization resolution dynamically based on energy ratio values. Instead of using a fixed high resolution for all directions, the system adjusts the quantization resolution parameter for each direction according to its energy importance. This allows the encoder to maintain high spatial accuracy for important directions while using coarser quantization for less important directions, thereby controlling overall bitrate while preserving perceptual quality.
3Quantity of substance
If lower quantization resolution is used for direction parameters, then bitrate is reduced, but spatial accuracy deteriorates
Solution Approach 1:
The patent applies local quality by using different quantization resolutions for different direction parameters based on their energy ratios. Specifically, direction parameters with higher energy ratios (more important directions) are quantized with higher resolution, while directions with lower energy ratios use lower resolution. This is achieved by determining quantization resolutions adaptively for each direction parameter based on its associated energy ratio, thereby optimizing overall perception quality without uniformly increasing complexity.
4Ease of operation
If multiple direction parameters are encoded with equal precision, then encoding simplicity is maintained, but perceptual quality deteriorates when directions have different energy importance
Solution Approach 1:
The patent changes the parameter of quantization resolution dynamically based on energy ratio values. Instead of using a fixed high resolution for all directions, the system adjusts the quantization resolution parameter for each direction according to its energy importance. This allows the encoder to maintain high spatial accuracy for important directions while using coarser quantization for less important directions, thereby controlling overall bitrate while preserving perceptual quality.
Data Source
AI summary
An apparatus comprising means configured to: obtain at least one direction parameter value for a time-frequency part of at least one audio signal (301); obtain at least one energy ratio for the time-frequency part (301), wherein each energy ratio is associated with a respective direction parameter value; generate respective at least one modified energy ratio from the at least one energy ratio for the time-frequency part (304); determine a quantization spatial resolution for encoding the at least one obtained direction parameter value based on the at least one modified energy ratio (305); and encode the obtained direction parameter values based on the quantization spatial resolution (306).