CELP–MDCT Audio Decoder Transition Processing with Zero-Input Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing switched audio codecs face issues with discontinuities and aliasing during transitions between CELP and MDCT coding schemes, leading to artifacts and inefficiencies in bitrate, delay, and computational complexity.
Innovation Solution
An audio decoder that utilizes a transition processor to modify decoded audio information using a zero-input response of a linear predictive filter, considering both initial states of frames encoded in linear and frequency domains, to achieve a smooth transition without altering the first decoded audio information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If switched audio codecs switch between CELP and MDCT coding schemes, then coding quality and bitrate efficiency are improved, but discontinuities and aliasing artifacts are introduced during transitions
Solution Approach 1:
The patent applies preliminary action by computing the zero-input response of the linear predictive filter in advance during the CELP frame processing, before the MDCT frame is fully decoded. This pre-computed response is then used to modify the MDCT frame during transition, eliminating discontinuities and aliasing artifacts without requiring additional complex processing at the transition point.
Solution Approach 2:
The zero-input response acts as an intermediary element between the CELP and MDCT coding schemes. It serves as a bridge that connects the linear prediction domain and frequency domain representations, allowing smooth transition between the two coding schemes while maintaining audio quality and eliminating artifacts.
2Reliability
If conventional methods modify the first decoded audio information to achieve smooth transition, then audio quality is improved, but processing delay is introduced
Solution Approach 1:
Instead of modifying the first decoded audio information (CELP frame) to achieve smooth transition, the patent inverts the approach by modifying the second decoded audio information (MDCT frame) using the pre-computed zero-input response. This reversal eliminates the need to reprocess the first frame, thereby avoiding additional processing delay while maintaining audio quality.
Solution Approach 2:
By pre-computing the zero-input response during CELP frame processing, the patent eliminates the need for additional processing of the first frame during transition. The pre-computed response is directly applied to the MDCT frame, avoiding the delay that would result from modifying the first frame after it has been decoded and potentially reprocessed.
3Reliability
If overlap-and-add operation is used to solve discontinuity, then audio quality is improved, but computational complexity and delay are increased
Solution Approach 1:
The patent extracts the essential function of the zero-input response from the complex overlap-and-add operation. Instead of performing the full overlap-and-add process which requires modifying both frames and combining them, the patent extracts only the necessary zero-input response component and applies it directly to the MDCT frame, significantly reducing computational complexity while maintaining audio quality.
Solution Approach 2:
The zero-input response serves as a simplified intermediary that captures the essential transition information without requiring the complex overlap-and-add operation. It acts as a mediator that directly connects the CELP and MDCT frames, eliminating the need for computationally intensive processing while maintaining smooth transitions and audio quality.
Data Source
AI summary
An audio decoder is disclosed. In one example, the audio decoder is for providing a decoded audio information on the basis of an encoded audio information includes a linear-prediction-domain decoder configured to provide a first decoded audio information on the basis of an audio frame encoded in a linear prediction domain, a frequency domain decoder configured to provide a second decoded audio information on the basis of an audio frame encoded in a frequency domain, and a transition processor. The transition processor is configured to obtain a zero-input-response of a linear predictive filtering, wherein an initial state of the linear predictive filtering is defined depending on the first decoded audio information and the second decoded audio information, and modify the second decoded audio information depending on the zero-input-response, to obtain a smooth transition between the first and the modified second decoded audio information.