Scalable Audio Encoding Using Spectral Outline Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional scalable encoding techniques face decreased encoding efficiency in the second layer due to the characteristics of residual signals, which are noise-like, making it difficult to enhance the quality of the original signal when decoding.
Innovation Solution
A scalable coding apparatus that includes a predictive information coding section to predict and encode the spectral outline of the input signal from the spectral outline of the lower layer decoded signal, improving the encoding efficiency of the upper layer and enhancing the quality of the decoded signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If encoding is performed on the residual signal obtained by subtracting the first layer decoded signal from the original signal, then the scalable encoding structure is established, but the encoding efficiency in the second layer decreases due to noise-like characteristics of the residual signal
Solution Approach 1:
The patent introduces a spectral outline signal as an intermediary between the first layer decoded signal and the second layer encoding. Instead of directly encoding the noise-like residual signal, the system extracts spectral outline information from the original signal, predicts it from the first layer decoded signal, and encodes only the prediction error. This intermediary approach preserves the scalable structure while improving second layer encoding efficiency by working with a signal that has stronger correlation properties.
Solution Approach 2:
The patent creates a simplified copy of the spectral characteristics by extracting spectral outline information from the original signal. This copy represents the essential spectral shape without the detailed noise components, allowing for more efficient prediction and encoding in the second layer while maintaining the ability to reconstruct the full signal when combined with the first layer.
2Adaptability or versatility
If the residual signal characteristics are used for second layer encoding, then the scalable decoding is achieved, but the quality of the original signal is difficult to enhance
Solution Approach 1:
The spectral outline signal serves as a mediator that carries essential spectral information from the original signal to the second layer decoder. By predicting this intermediary signal from the first layer decoded signal and encoding only the prediction error, the system enables quality enhancement while maintaining scalable decoding capability. The spectral outline acts as a bridge that preserves the relationship between layers while improving reconstruction quality.
3Productivity
If spectral outline prediction is performed from the lower layer decoded signal, then the encoding efficiency of the upper layer is improved, but the device complexity increases due to additional calculation sections
Solution Approach 1:
The patent extracts only the essential spectral outline information from the full signal processing chain, separating it from the detailed spectral components. By focusing calculation efforts on extracting and predicting only the spectral outline rather than processing the entire signal spectrum, the system improves upper layer encoding efficiency while limiting the increase in device complexity to only the necessary spectral outline calculation sections.
Data Source
AI summary
A scalable encoder enabling improvement of the encoding efficiency in the second layer and improvement of the quality of the original signal decoded using the encoding signal in the second layer. A predictive coefficient encoder of the scalable encoder has a predictive coefficient codebook where candidates of the predictive coefficient are recorded. After searching the predictive coefficient codebook, the scale factor of the first layer decoded signal inputted from a scale factor calculator is multiplied, and a predictive coefficient which most approximates the multiplication result to the scale factor of the original signal inputted from the scale factor calculator is determined and encoded, and the coded code is inputted to a multiplexer.