MDCT Packet Loss Concealment Using Prototype Buffer Overlap-Add
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio codecs face challenges in maintaining high spectral resolution and reducing complexity when dealing with packet loss concealment, particularly due to the unreliability of IFFT endpoints and the use of asymmetric MDCT windows, which can lead to lower quality reconstructed signals.
Innovation Solution
The solution involves using a decoder that copies and overlap adds segments from a PLC prototype buffer to an MDCT processing buffer, employing asymmetric MDCT windows and overlap-add (OLA) operations to fill the analysis window length, thereby improving the quality of the reconstructed signal.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If asymmetric MDCT windows with fewer trailing zeros are used to reduce algorithmic delay, then processing delay is reduced, but the reliability of IFFT endpoints deteriorates leading to lower quality reconstructed signals
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing prototype values for the unreliable IFFT endpoint regions before actual processing. When packet loss occurs, these pre-computed prototypes are used to fill in the unreliable endpoint areas, preventing the reliability issue from affecting the final reconstruction quality while maintaining the low-delay asymmetric window benefits
Solution Approach 2:
The patent introduces an intermediary mechanism (prototype buffer and selective copying logic) that mediates between the low-delay asymmetric window processing and the reliability requirements. The intermediary identifies unreliable regions and replaces them with trusted prototype values, thus bridging the gap between speed optimization and quality assurance
2Measurement precision
If longer MDCT windows are used to improve frequency resolution, then spectral resolution is improved, but processing complexity increases
Solution Approach 1:
The patent applies local quality by using different window lengths in different time regions. Instead of uniformly using long windows throughout, it employs shorter windows in regions where high frequency resolution is less critical and longer windows only where needed, thus achieving good frequency resolution where necessary while reducing overall processing complexity
Solution Approach 2:
The patent segments the processing into different regions: reliable regions processed with standard short-window MDCT and unreliable regions handled with prototype copying. This segmentation allows the system to achieve high frequency resolution in critical areas without applying the computational overhead of long windows globally
Data Source
AI summary
A method, decoder, and program code for filling an analysis window length with a time domain signal for concealing a lost audio frame associated with a received audio signal. A first segment of a previously received part of a received audio signal is copied from a prototype buffer. A second segment of the previously received part of the received audio signal is overlap added from the prototype buffer to an initial portion of a reconstructed part of the received audio signal followed by a remaining portion of the reconstructed part of the received audio signal.