Predictive Audio Coding Error Recovery for Stable Sound Location
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Predictive coding schemes in stereo and multichannel audio encoding are sensitive to errors, leading to error propagation and unwanted changes in the perceived location of sound sources, especially in stable audio scenes, and existing remedies consume bandwidth even in error-free channels.
Innovation Solution
A method and decoder that use parameter memory to replace decoded parameters when errors are detected, determining the stability and concentration of sound sources in a subset of channels, and activating parameter recovery to reduce error propagation without redundant transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If predictive coding schemes are used to reduce bandwidth consumption, then bandwidth efficiency is improved, but error propagation and unwanted changes in perceived sound source location occur when errors are detected
Solution Approach 1:
The system performs preliminary analysis of sound source stability and concentration before error recovery activation. By pre-evaluating whether the sound source is stable and concentrated in a subset of channels, the system prepares to switch to parameter memory replacement only when appropriate, avoiding unnecessary bandwidth consumption while protecting against error propagation in vulnerable scenarios
Solution Approach 2:
The system dynamically changes the coding parameter selection based on detected error conditions and sound source characteristics. When errors are detected and the sound source is determined to be stable and concentrated, the system switches from using transmitted predictive parameters to using parameters from memory, thereby changing the operational state to eliminate error propagation while maintaining bandwidth efficiency
2Reliability
If parameter recovery is activated to replace decoded parameters with stored parameters, then error propagation is reduced, but bandwidth is wasted in error-free channels
Solution Approach 1:
The system dynamically adjusts the error recovery mechanism based on real-time detection of error conditions and sound source characteristics. The parameter recovery is not continuously activated but rather switched on/off based on whether errors are present and whether the sound source is stable and concentrated, making the system adaptive rather than static
Solution Approach 2:
The system uses parameter copies from memory to replace decoded parameters only when necessary. By maintaining parameter memory as a backup copy and selectively using it during error conditions with stable, concentrated sound sources, the system avoids the overhead of continuous parameter transmission while ensuring reliability when needed
3Loss of energy
If transmitted parameters are used in predictive coding mode, then bandwidth efficiency is improved, but unnatural sound shifts occur during stable audio scenes when errors are detected
Solution Approach 1:
The system changes the operational parameter source from transmitted predictive parameters to memory-based parameters when errors are detected and the sound source is determined to be stable and concentrated. This parameter switching eliminates the unnatural sound shifts caused by erroneous predictive parameters while maintaining bandwidth efficiency by not continuously transmitting redundant parameters
Solution Approach 2:
The system incorporates feedback mechanisms that monitor error conditions and sound source characteristics (stability and concentration) to determine when parameter recovery should be activated. This feedback loop ensures that parameter replacement occurs only when it will improve sound source location stability, avoiding unnecessary changes that could introduce artifacts
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, apparatuses, decoders, and computer programs for audio coding are provided. Responsive to a bad frame being indicated, it is determined that a parameter memory is corrupted. Responsive to a bad frame not being indicated decoding multichannel parameters. Responsive to the parameter memory being corrupted, it is determined, based on the location measure, whether the reconstructed sound source is stable and predominantly concentrated in a subset of channels of multichannels of the received multichannel signal. Responsive to the reconstructed sound source being concentrated in the subset of channels of the multichannels and being stable, replacing decoded multichannel parameters with stored multichannel parameters.