Audio Reproducing Device Bandwidth Extension Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional audio decoding techniques face challenges in complying with HE-AACv2 profile and higher levels, leading to increased computation and generation of abnormal sounds, especially when decoding multi-channel signals, due to the need for HQ-SBR which performs complex arithmetic and has different delay information formats compared to LP-SBR.
Innovation Solution
An audio reproducing device that switches between two bandwidth extension processing methods based on analysis information of the basic codec, selecting less computationally intensive processing for multi-channel signals and using buffer-stored stereo extension information to maintain continuous delay information and prevent abnormal sounds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If HQ-SBR is used to comply with HE-AACv2 profile and decode multi-channel signals, then sound quality is improved, but computation amount significantly increases
Solution Approach 1:
The patent dynamically switches between HQ-SBR and LP-SBR processing modes based on the presence of PS data and channel configuration. The system adapts its processing complexity in real-time, using HQ-SBR only when necessary for HE-AACv2 compliance with PS data, and LP-SBR for other cases to reduce computation while maintaining acceptable sound quality.
2Productivity
If SBR switching is performed based on PS data presence, then computation amount is reduced, but abnormal sounds are generated due to delay information discontinuity
Solution Approach 1:
The patent performs preliminary analysis of the basic codec type and PS data presence before selecting the SBR processing mode. By determining the appropriate processing path in advance based on stream characteristics, the system avoids mid-processing switches that would cause delay information discontinuity and abnormal sounds, while still achieving computation reduction where possible.
3Productivity
If LP-SBR is used instead of HQ-SBR, then computation amount is reduced, but compliance with HE-AACv2 profile and higher levels cannot be achieved
Solution Approach 1:
The patent changes the processing parameters (switching between LP-SBR and HQ-SBR) based on the specific requirements indicated by PS data presence and channel configuration. This allows the system to achieve HE-AACv2 profile compliance only when necessary (with PS data and mono/basic codec), while using the more efficient LP-SBR for other scenarios, thus balancing profile compliance with computation reduction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
This invention prevents occurrence of abnormal sounds without significantly increasing calculation amount even for multi-channel signals. An audio reproducing device (100) decodes and reproduces a stream including basic codec. The audio reproducing device (100) includes: a stream separating unit (101) which separates the stream into basic codec and bandwidth extension data; a basic codec analyzing unit (102) which analyzes the basic codec; a basic codec decoding unit (103) which decodes the basic codec in accordance with basic codec analysis information; a bandwidth extension data analyzing unit (104) which analyzes bandwidth extension data; a first bandwidth extension processing unit (106) which extends a decoded basic codec signal by using the bandwidth extension information; a second bandwidth extension processing unit (107) which extends a decoded basic codec signal by using the bandwidth extension information with an accuracy higher than an accuracy of the first bandwidth extension processing unit (106); and a switching unit (109) which switches between the first bandwidth extension processing unit (106) and the second bandwidth extension processing unit (107) based on the basic codec analysis information.