TDAC Noise Shaping Interpolation for Audio Transition Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current multi-mode audio codecs face challenges in simultaneously shaping quantization noise in both time-domain and frequency-domain for Time-Domain Aliasing Cancellation (TDAC) transforms, leading to discontinuities and artifacts during mode transitions.
Innovation Solution
A frequency-domain noise shaping method and device that interpolates the spectral shape and time-domain envelope of quantization noise, allowing for simultaneous noise shaping across transform-coded blocks, using band-specific filters and noise gains to ensure smooth transitions and efficient energy compaction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If separate noise shaping procedures are used for transform coding and predictive coding modes, then each mode can be optimized independently, but discontinuities and artifacts occur during mode transitions
Solution Approach 1:
The patent combines time-domain noise shaping filtering with frequency-domain noise shaping by applying the filtering operation to the transform coefficients after inverse transform. This merging allows the same filtering operation to serve both time-domain and frequency-domain noise shaping requirements, ensuring consistency across coding mode transitions while maintaining the benefits of separate optimization procedures for each mode
Solution Approach 2:
The patent introduces an intermediate filtering operation in the time-domain that acts as a bridge between transform coding and predictive coding noise shaping approaches. By applying the noise shaping filter to the reconstructed signal and using the filtered signal as input to the transform, the system creates a smooth transition mechanism that maintains reliability while preserving mode-specific optimization
2Productivity
If quantization noise is shaped in the transform domain using scale factors, then bit rate efficiency is improved, but time-domain noise characteristics are not controlled
Solution Approach 1:
The patent segments the noise shaping process into two distinct stages: frequency-domain noise shaping through scale factor control of transform coefficients, and time-domain noise shaping through filtering of the reconstructed signal. This segmentation allows each stage to independently optimize its respective domain while the combination of both stages achieves comprehensive noise control across frequency and time domains
Solution Approach 2:
The patent extends noise shaping from the frequency-domain only to include the time-domain by applying filtering operations after inverse transform. This dimensional extension allows the system to control noise characteristics in both frequency and time domains simultaneously, maintaining bit rate efficiency while adding time-domain noise control capability
3Ease of operation
If time-domain filtering is used for noise shaping in predictive coding, then time-domain noise characteristics are controlled, but frequency-domain noise shaping is less effective
Solution Approach 1:
The patent merges time-domain filtering operations with frequency-domain transform coding by applying the noise shaping filter to the transform coefficients in the frequency-domain. This merging allows the filtering operation to simultaneously affect both time-domain and frequency-domain noise characteristics, maintaining the effectiveness of frequency-domain noise shaping while adding time-domain control capability
Solution Approach 2:
The patent substitutes the traditional time-domain filtering approach with a frequency-domain filtering operation applied to transform coefficients. By performing the filtering operation in the frequency-domain on the MDCT coefficients, the system maintains the efficiency benefits of frequency-domain processing while achieving the time-domain noise control objectives traditionally requiring time-domain filtering
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A frequency-domain noise shaping method and device interpolates a spectral shape and a time-domain envelope of a quantization noise in a windowed and transform-coded audio signal. In the method and device, transform coefficients of the windowed and transform-coded audio signal are split into a plurality of spectral bands. For each spectral band, a first gain representing a spectral shape of the quantization noise at a first transition between a first time window and a second time window is calculated, a second gain representing a spectral shape of the quantization noise at a second transition between the second time window and a third time window is calculated, and the transform coefficients of the second time window are filtered based on the first and second gains, to interpolate between the first and second transitions the spectral shape and the time-domain envelope of the quantization noise.