Multi-Channel Audio Encoder Smoothing Spatial Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional multi-channel audio coding techniques face challenges in reducing data rate while maintaining subjective quality, particularly due to coarse quantization of spatial parameters leading to artifacts such as 'jumps' or 'modulation' effects in sound source localization, especially for non-tonal signals like slowly moving point sources or rapidly moving sinusoids.
Innovation Solution
An encoder-side adaptive smoothing method is introduced to post-process reconstruction parameters, allowing values between quantization levels and using smoothing control information to mitigate quantization errors, thereby improving audio quality without increasing data rate.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If conventional multi-channel audio coding techniques are used with coarse quantization of spatial parameters, then data rate is reduced, but audible artifacts such as 'jumps' or 'modulation' effects occur in sound source localization
Solution Approach 1:
The encoder performs preliminary smoothing of spatial parameters before quantization. By applying a smoothing filter to the spatial parameters in the encoder, the system pre-compensates for quantization artifacts that would otherwise cause audible jumps and modulation effects in the reconstructed audio signal.
Solution Approach 2:
A smoothing filter acts as an intermediary between the spatial parameter estimation and quantization stages. This intermediary component processes the spatial parameters to remove high-frequency fluctuations before quantization, thereby reducing the harmful artifacts without requiring higher quantization precision.
2Quantity of substance
If reconstruction parameters are quantized to integer multiples of quantizer step size, then data rate is minimized, but non-integer values are lost causing quality degradation
Solution Approach 1:
The system changes the parameter representation by allowing non-integer quantization levels for spatial parameters. Instead of restricting quantization to integer multiples of the quantizer step size, the system enables fractional values, thereby maintaining higher precision while still using efficient quantization. This is achieved by modifying the quantization process to support a broader range of valid parameter values.
3Object-generated harmful factors
If smoothing is applied to reconstruction parameters, then audible artifacts are reduced, but processing complexity increases
Solution Approach 1:
The smoothing filter is applied selectively only to spatial parameters rather than all audio parameters. By limiting the smoothing operation to only those parameters that cause artifacts (the spatial parameters), the system achieves artifact reduction with minimal additional processing complexity, avoiding the need to process entire audio channels through the smoothing filter.
Data Source
AI summary
On an encoder-side, a multi-channel input signal is analyzed for obtaining smoothing control information, which is to be used by a decoder-side multi-channel synthesis for smoothing quantized transmitted parameters or values derived from the quantized transmitted parameters for providing an improved subjective audio quality in particular for slowly moving point sources and rapidly moving point sources having tonal material such as fast moving sinusoids.