Scalable Audio Encoding with Encoder-Side Quantization Noise Shaping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ADPCM coding methods, such as those in the G.722 and G.727 standards, face challenges in maintaining audio signal quality due to flat quantization noise spectra, which can become audible in low-energy frequency regions, particularly around 2000-2500 Hz, and require additional noise shaping techniques that increase decoder complexity.
Innovation Solution
A hierarchical coding method is introduced, which includes core coding and enhancement coding stages. The enhancement coding stage uses a noise shaping filter to determine a target signal by minimizing the error between possible scalar quantization values and the shaped noise, allowing for improved noise shaping without additional processing at the decoder, thus enhancing signal quality across various bit rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If ADPCM coding with nested codes is used according to G.722 or G.727 standards, then the encoding efficiency and scalability are improved, but the quantization noise spectrum remains flat and becomes audible in low-energy frequency regions
Solution Approach 1:
The patent applies noise shaping by modifying the quantization process parameters to redistribute quantization noise energy across the frequency spectrum. A noise shaping filter is introduced that processes the prediction error signal before quantization, shaping the noise spectrum to push quantization noise away from low-energy frequency regions (particularly 2000-2500 Hz) where it would be most audible, while maintaining the scalable nested code structure of G.722/G.727 standards.
2Object-affected harmful factors
If noise shaping techniques are added to improve audio signal quality, then the perceptual quality is enhanced, but the decoder complexity increases
Solution Approach 1:
The noise shaping filter is applied at the encoder side before quantization, performing the noise shaping action in advance. This preliminary action allows the decoder to simply reconstruct the signal using the transmitted quantized indices without requiring complex noise shaping processing, thereby improving perceptual quality while avoiding increased decoder complexity.
3Object-affected harmful factors
If the quantization noise is reduced in low-energy regions, then the audibility is reduced, but the overall signal fidelity may be compromised
Solution Approach 1:
The noise shaping filter applies frequency-dependent processing to the quantization noise, targeting specific frequency regions (particularly low-energy regions around 2000-2500 Hz) where quantization noise is most audible. The filter shape is adapted based on the signal's spectral characteristics, applying stronger noise shaping in regions where the signal energy is low and less shaping where signal energy is high, thereby reducing audibility while preserving overall signal fidelity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a method for scalable encoding of an audio-digital signal comprising for a common section of the input signal: - a core encoding step, producing an index (I B (n)) of scalar quantification for each sample n of the common section and - at least one enhancement encoding step, producing indices (Jk {n)) of scalar quantification (Qk enh) for each encoded sample of an enhancement signal. The method is such that the enhancement encoding step comprises a step of obtaining a filter (W(z)) for transforming the encoding noise used to determine a target signal, and that the indices (Jk (n)) for scalar quantification (Qk enh (n)) of said enhancement signal are determined by minimising the error between a set of possible scalar quantification values and said target signal. The encoding method according to the invention can also comprise a transformation of the encoding noise for encoding the core flow. The invention also relates to an encoder implementing the encoding method such as described.