Audio Frame Loss Compensation via MDCT and Waveform Adjustment
Solution Overview
Problem
Existing frame loss compensation techniques for audio signals in the transform domain suffer from modest compensation effects, high computational complexity, and significant delays, particularly when dealing with multi-harmonic signals.
Innovation Solution
A method that judges the type of lost frames, using MDCT coefficients from prior frames to calculate initial compensation signals, and performs waveform adjustments through pitch period estimation and overlapped periodic extension to generate time-domain signals without delay and high complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GAPES technique is used to perform frame loss compensation, then the compensation effect is improved, but the computational complexity increases and memory consumption becomes large
Solution Approach 1:
The patent segments the compensation process by frame type, dividing audio frames into multi-harmonic and non-multi-harmonic categories. Different compensation strategies are applied to each segment: simple repetition for non-multi-harmonic frames and MDCT-DSTFT conversion only for multi-harmonic frames. This segmentation resolves the contradiction by applying complex processing only where necessary while using simple processing elsewhere.
Solution Approach 2:
The patent applies local quality by using different compensation methods tailored to specific frame types. Non-multi-harmonic frames receive simple repetition compensation, while multi-harmonic frames receive the more sophisticated MDCT-DSTFT based compensation. This localized approach optimizes the balance between compensation quality and computational complexity for each specific case.
2Ease of manufacture
If simple repetition or mute substitution is used for frame loss compensation, then the implementation is simple and delay is avoided, but the compensation effect is modest
Solution Approach 1:
The patent introduces dynamics by making the compensation method adaptive based on frame type characteristics. The system dynamically selects between simple repetition and MDCT-DSTFT conversion depending on whether the frame is multi-harmonic or not. This dynamic adaptation resolves the contradiction by maintaining simplicity for most frames while enhancing quality for specific multi-harmonic cases.
3Reliability
If noise shaping and inserting technology is used, then the compensation effect on noise-like signals is improved, but the effect on multi-harmonic audio signals becomes very poor
Solution Approach 1:
The patent achieves universality by creating a multi-functional compensation system that handles both noise-like and multi-harmonic signals effectively. The system uses simple repetition for noise-like non-multi-harmonic frames and MDCT-DSTFT conversion for multi-harmonic frames, making the overall system universally applicable to different signal types without compromising performance on either.
4Reliability
If MDCT-DSTFT conversion is performed for all frames, then the compensation effect on multi-harmonic signals is improved, but the computational complexity and processing time increase significantly
Solution Approach 1:
The patent applies partial action by performing the computationally intensive MDCT-DSTFT conversion only partially - specifically for multi-harmonic frames where it is needed, rather than for all frames. This partial application resolves the contradiction by achieving the necessary compensation quality for multi-harmonic signals while avoiding unnecessary processing time for other frame types.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A frame loss compensation method and apparatus for audio signals are disclosed, so as to obtain better compensation effects and at the same time ensure that there is no delay and the complexity is low. The method includes: when a first frame immediately following a correctly received frame is lost, judging a frame type of the first lost frame, and when the first lost frame is a non-multi-harmonic frame, calculating MDCT coefficients of the first lost frame by using MDCT coefficients of one or more frames prior to the first lost frame; obtaining an initially compensated signal of the first lost frame according to the MDCT coefficients of the first lost frame; and performing a first class of waveform adjustment on the initially compensated signal of the first lost frame and taking an adjusted time-domain signal as a time-domain signal of the first lost frame. The apparatus includes a frame type judgment module, an MDCT coefficient acquisition module, an initial compensation signal acquisition module and an adjustment module.