Stereo Audio Decoding Using Adaptive Prediction Direction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-channel audio coding techniques face challenges in achieving high coding gain while maintaining audio quality and reducing computational complexity, particularly in situations where mid/side coding does not yield significant coding gains due to phase shifts or equal waveforms, and parametric coding introduces lossy information and artifacts.
Innovation Solution
A method involving frequency-domain prediction of a second combination signal from a first combination signal, using a modified discrete cosine transform (MDCT) for efficient stereo coding, where the prediction direction is adaptively chosen based on energy distribution, and phase differences are handled using real-to-imaginary or imaginary-to-real converters to minimize computational complexity and enhance coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If mid/side coding is applied to achieve coding gain, then coding efficiency is improved, but audio quality deteriorates when channels have phase shifts or equal waveforms
Solution Approach 1:
The patent applies frequency-selective mid/side coding by changing the coding parameter (coding mode) based on the spectral characteristics of the audio signal. The encoder analyzes each frequency band and applies mid/side coding only in bands where the side signal energy is sufficiently smaller than the mid signal energy, thereby maintaining coding efficiency while preserving audio quality in bands where phase shifts or equal waveforms would otherwise degrade performance.
Solution Approach 2:
The coding mode is dynamically adjusted on a per-band basis rather than applying a fixed coding mode to the entire spectrum. The encoder determines for each frequency band whether to apply mid/side coding or conventional stereo coding based on the energy relationship between mid and side signals in that band, allowing the system to adapt to local spectral characteristics and optimize both coding efficiency and audio quality.
2Quantity of substance
If parametric coding is applied to reduce bit rate, then bit rate is reduced, but information loss and artifacts increase
Solution Approach 1:
The patent applies different coding strategies to different frequency bands based on local signal characteristics. In bands where mid/side coding is beneficial (side signal energy significantly smaller than mid signal energy), the patent applies this more efficient coding method. In other bands, conventional stereo coding is used to preserve signal fidelity. This local adaptation ensures that information loss is minimized while still achieving bit rate reduction where appropriate.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
An audio or video encoder and an audio or video decoder are based on a combination of two audio or video channels (201, 202) to obtain a first combination signal (204) as a mid signal and a residual signal (205) which can be derived using a predicted side signal derived from the mid signal. The first combination signal and the prediction residual signal are encoded (209) and written (212) into a data stream (213) together with the prediction information (206) derived by an optimizer (207) based on an optimization target (208) and a prediction direction indicator indicating a prediction direction associated with the residual signal. A decoder uses the prediction residual signal, the first combination signal, the prediction direction indicator and the prediction information to derive a decoded first channel signal and a decoded second channel signal. In an encoder example or in a decoder example, a real-to-imaginary transform can be applied for estimating the imaginary part of the spectrum of the first combination signal. For calculating the prediction signal used in the derivation of the prediction residual signal, the real-valued first combination signal is multiplied by a real portion of the complex prediction information and the estimated imaginary part of the first combination signal is multiplied by an imaginary portion of the complex prediction information.