Multi-mode Audio Decoder Spectral Shaping for Domain Transitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio coders face challenges in transitioning between frames encoded in different domains without significant bit rate loss, particularly when handling both speech and general audio content.
Innovation Solution
A multi-mode audio signal decoder and encoder that performs spectral shaping in the frequency domain for both linear-prediction and frequency-domain modes, allowing for efficient transitions between modes by using spectral shaping of decoded spectral coefficients and applying appropriate parameters for noise shaping, enabling good quality transitions without substantial bit rate overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If transitions between frames encoded in different domains are implemented, then audio quality is improved, but bit rate increases significantly
Solution Approach 1:
The patent applies spectral shaping parameters to transform the spectral characteristics of decoded coefficients, enabling smooth transitions between different coding domains (linear-prediction and frequency-domain modes) without requiring substantial bit rate increases. The spectral shaping modifies the distribution of spectral energy to facilitate domain switching while maintaining audio quality.
2Reliability
If spectral shaping is applied to decoded spectral coefficients, then transitions between encoding modes are improved, but processing complexity increases
Solution Approach 1:
The patent performs spectral shaping on decoded spectral coefficients as a preliminary step before final audio reconstruction. By pre-shaping the spectral characteristics of the decoded coefficients, the system prepares the data for smooth mode transitions, reducing the complexity of subsequent processing steps and improving overall transition reliability.
Data Source
AI summary
A multi-mode audio signal decoder has a spectral value determinator to obtain sets of decoded spectral coefficients for a plurality of portions of an audio content and a spectrum processor configured to apply a spectral shaping to a set of spectral coefficients in dependence on a set of linear-prediction-domain parameters for a portion of the audio content encoded in a linear-prediction mode, and in dependence on a set of scale factor parameters for a portion of the audio content encoded in a frequency-domain mode. The audio signal decoder has a frequency-domain-to-time-domain converter configured to obtain a time-domain audio representation on the basis of a spectrally-shaped set of decoded spectral coefficients for a portion of the audio content encoded in the linear-prediction mode and for a portion of the audio content encoded in the frequency domain mode. An audio signal encoder is also described.