Audio Encoder Noise Shaping Filter for PCM Quantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PCM coders, such as G.711, face challenges in reducing quantizing noise, especially in areas of low signal energy, as they lack noise shaping capabilities due to the absence of synthesis filters, leading to audible noise in wide dynamic range signals.
Innovation Solution
A method that combines the input audio signal with an intermediate feedback signal, applies bit rate scalable quantization, and determines a quantizing noise filter function to shape the noise, allowing for improved noise reduction without requiring transmission of linear prediction coefficients, suitable for PCM coders without synthesis filters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If PCM coding with 256 quantizing levels is used, then the signal can be transmitted at 64 kbit/s in PSTN, but quantizing noise becomes audible in low energy areas
Solution Approach 1:
The patent applies noise shaping to convert the harmful quantizing noise into a beneficial form by redistributing it spectrally. The quantizing noise is filtered to concentrate energy in high-frequency regions where it is masked by the signal, transforming an audible defect into an imperceptible characteristic. This allows maintaining 64 kbit/s transmission while eliminating audible noise in low energy areas.
2Object-affected harmful factors
If noise shaping with linear prediction coefficients is applied, then quantizing noise is reduced in low energy areas, but the complexity increases due to transmission of prediction coefficients
Solution Approach 1:
The patent extracts only the essential noise shaping functionality from complex linear prediction systems. Instead of transmitting full linear prediction coefficients and synthesis filters, it uses a simplified filter derived from the reconstructed signal that achieves noise shaping without requiring additional coefficient transmission. This extracts the beneficial noise reduction while eliminating the complexity overhead.
Solution Approach 2:
The patent introduces an intermediary approach by using the reconstructed signal itself as the basis for deriving the noise shaping filter, rather than relying on transmitted linear prediction coefficients. This intermediary method allows the decoder to generate the necessary filter locally from available data, avoiding the need to transmit additional coefficient information.
3Object-affected harmful factors
If quantizing noise shaping is implemented in adaptive predictive coders, then noise is masked in high energy areas, but the technique is not suitable for non-predictive coders like PCM
Solution Approach 1:
The patent creates a universal noise shaping solution that works for both adaptive predictive coders and non-predictive PCM coders. By deriving the noise shaping filter from the reconstructed signal rather than from linear prediction coefficients, the method becomes adaptable to different coder types. This universal approach allows PCM systems to implement noise shaping without requiring synthesis filters or coefficient transmission.
Data Source
AI summary
The invention concerns an encoder for an input audio signal (S(z)) comprising a combination module combining the input audio signal with an intermediate counter-reaction signal forming a modified input signal and a quantification module scalable for the rate (91) of said modified input signal, delivering a binary raster of quantification indexes of a predetermined rate. This encoder is such that it comprises: a reverse quantification module (93) of a part of the quantification indexes of the binary raster corresponding to the indexes of a rate lower than the predetermined rate, for predetermining a reconstructed signal (SMic(z)), a module for determining (94) a quantification noise (QMic(z)) derived from at least quantification modules and reverse quantification modules in series, a determination module (111) for a filtering function of the quantification noise from said reconstructed signal and a filtering module fit for applying (33) said filtering function to said quantification noise in order to obtain said intermediary counter-reaction signal. The invention also relates to an encoding method applied by the described encoder and a corresponding decoding method applied by a decoder.