Audio Frame Error Concealment Using Saved Decoder Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio codecs, such as G.718 and G.729.1, face challenges in maintaining signal quality and perceptual characteristics during frame erasures due to channel errors, which can lead to audible distortions and disruptions in the reconstructed output signal.
Innovation Solution
The method involves extension coding, where lower frequency content is mapped to higher frequencies using modified discrete cosine transform spectrum values, with scaling factors applied to maintain signal quality and minimize artifacts, employing both generic and tonal coding modes to handle frame errors effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If channel errors occur in wireless communications networks, then data segments may be corrupted or lost, but this leads to frame erasures and audible distortions in the reconstructed output signal
Solution Approach 1:
The patent applies preliminary action by saving parameter values from previously decoded frames before errors occur. When a frame error is detected, these pre-saved parameters are immediately used to reconstruct the corrupted frame, avoiding audible distortions without requiring retransmission or complex real-time analysis.
Solution Approach 2:
The patent uses copying by creating replicas of parameter values from valid frames and applying them to replace corrupted parameters in erroneous frames. This allows the decoder to reconstruct lost or corrupted audio data by copying relevant spectral and temporal parameters from neighboring valid frames, thereby eliminating audible distortions caused by frame erasures.
2Manufacturing precision
If parameter values from previous frames are used to reconstruct frames with errors, then signal quality is maintained, but computational complexity increases due to multiple scaling factor calculations
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting multiple scaling factors (fac_spect, fac_sin) based on the specific frame error conditions and the type of preceding valid frame (generic or tonal). These parameter modifications enable precise control over the reconstruction process, maintaining signal quality by adapting the blending between original and reconstructed spectral components to match the local signal characteristics.
Solution Approach 2:
The patent implements local quality by applying different scaling factors to different spectral components and frequency bands during frame reconstruction. Specifically, fac_spect controls the overall spectral envelope while fac_sin controls the sinusoidal components, allowing localized adjustment of reconstruction quality in different frequency regions based on the signal's local characteristics, thereby maintaining high signal quality without uniform complexity across all frequencies.
Data Source
Figure 1
Figure 2A~3B
Figure 4
AI summary
A method of frame error concealment in encoded audio data comprises receiving encoded audio data in a plurality of frames; and using saved one or more parameter values from one or more previous frames to reconstruct a frame with frame error. Using the saved one or more parameter values comprises deriving parameter values based at least part on the saved one or more parameter values and applying the derived values to the frame with frame error.