Audio Encoder Adaptive Noise-Aware Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech and audio codecs face challenges in encoding audio signals distorted by background noise, leading to decreased encoding quality, increased listening effort, and reduced intelligibility, as they struggle to differentiate between desired and undesired signal components.
Innovation Solution
An audio encoder that adapts encoding based on noise information, prioritizing higher accuracy for signal components less affected by noise, using a perceptual weighting filter to adjust the encoding process and reduce quantization errors in noise-affected areas, thereby improving signal-to-noise ratio and intelligibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the codec encodes both the desired signal and the undesired distortions together, then the coding problem becomes more complicated, but the encoding quality decreases
Solution Approach 1:
The patent segments the audio signal into different components (desired signal and undesired distortions) and applies different encoding strategies to each component. The desired signal is encoded with higher accuracy while the undesired distortions are encoded with lower accuracy or suppressed, thereby improving overall encoding quality without requiring the codec to treat all components equally complexly.
Solution Approach 2:
The patent applies different quality levels to different parts of the signal based on their importance. Critical signal components are encoded with high precision while less important components (such as background noise or artifacts) are encoded with lower precision or omitted entirely, optimizing the trade-off between encoding quality and computational complexity.
2Loss of information
If the codec encodes the combination of two sources with the same quality as a single clean signal, then the speech part would still be lower quality than the clean signal
Solution Approach 1:
The patent extracts the desired speech signal from the mixed signal containing both desired and undesired components. By separating and independently encoding the speech portion, the codec can maintain high speech quality even when the overall signal contains noise or artifacts, effectively taking out the valuable information from the contaminated signal.
Solution Approach 2:
The patent converts the presence of noise and artifacts in the input signal into beneficial information for encoding decisions. The noise characteristics are analyzed and used to adjust encoding parameters, allowing the codec to allocate bits more effectively and protect the speech components from degradation while tolerating or even utilizing the presence of background noise.
3Measurement precision
If higher encoding accuracy is applied to all parts of the audio signal, then the listening effort increases, but the intelligibility may decrease in noise-affected areas
Solution Approach 1:
The patent applies encoding accuracy selectively rather than uniformly across all signal components. High encoding accuracy is applied only to critical speech portions that require it for intelligibility, while less critical portions receive lower encoding accuracy. This partial application of high precision reduces the overall computational burden and listening effort while maintaining intelligibility where it matters most.
Data Source
AI summary
An audio encoder for providing an encoded representation on the basis of an audio signal, wherein the audio encoder is configured to obtain a noise information describing a noise included in the audio signal, and wherein the audio encoder is configured to adaptively encode the audio signal in dependence on the noise information, such that encoding accuracy is higher for parts of the audio signal that are less affected by the noise included in the audio signal than for parts of the audio signal that are more affected by the noise included in the audio signal.