High Pass Filter for Spatial Audio Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current teleconference systems using Algebraic Code Excited Linear Prediction (ACELP) based Adaptive Multi-Rate Wide Band (AMR-WB) codecs introduce disturbing spatial noise artifacts, particularly at low bit rates, which affect the localization of sound sources in spatial audio, and existing enhancement methods like Voiceage's frequency-selective pitch enhancement are insufficient in addressing these issues.
Innovation Solution
A method involving high pass filtering of ACELP coded audio signals using time-varying filters, where cutoff frequencies are adjusted based on estimated signal type and pitch frequencies to reduce spatial noise, employing Finite Impulse Response (FIR) or Infinite Impulse Response (IIR) filters, preferably elliptical IIR filters, to suppress noise without distorting the pitch component.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If ACELP coded audio signals are transmitted at low bit rates, then bandwidth efficiency is improved, but spatial noise artifacts increase and degrade audio quality
Solution Approach 1:
The patent applies dynamic filtering by adjusting the cutoff frequency of the high-pass filter based on the estimated pitch frequency of the audio signal. This dynamic adaptation allows the filter to maintain optimal performance across varying speech characteristics while effectively suppressing spatial noise artifacts introduced by low-bit-rate ACELP coding
Solution Approach 2:
The patent changes the filtering parameters (cutoff frequency) based on the signal characteristics, specifically adapting the filter cutoff frequency to the estimated pitch frequency. This parameter adjustment optimizes the balance between noise suppression and preservation of speech quality at low bit rates
2Object-affected harmful factors
If high pass filtering is applied to reduce spatial noise, then audio quality is improved, but pitch components may be distorted
Solution Approach 1:
The filter cutoff frequency is dynamically adjusted to track the pitch frequency of the speech signal. By making the filter adaptive rather than static, the system maintains the pitch components while suppressing spatial noise, preventing distortion of speech fundamentals
Solution Approach 2:
The system uses pitch estimation feedback to continuously adjust the filter parameters. The estimated pitch frequency feeds back into the filter design, creating a closed-loop system that adapts to the speech signal characteristics and preserves pitch accuracy while reducing noise
3Object-affected harmful factors
If existing enhancement methods are used, then some noise reduction is achieved, but they are insufficient for low bit rate ACELP coded signals
Solution Approach 1:
The patent specifically tailors the filtering parameters to the characteristics of ACELP coded signals, adjusting the cutoff frequency based on the pitch estimation that accounts for the coding artifacts introduced by ACELP. This specialized parameter adaptation makes the method effective for low bit rate ACELP signals where generic enhancement methods fail
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly enhances spatial audio signals by reducing spatial noise, improving teleconference quality and enabling effective use of low bit rates for coding stereo and binaural signals, as demonstrated by MUSHRA test results showing a 15-point quality improvement and maintaining audio fidelity without introducing additional distortions.
Implementation Method 1
filtering parameters are determined based on at least one of the estimated signal type and the estimated pitch frequency. The received signal is then high pass filtered based on the determined filter parameters
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
In a method of enhancing spatial audio signals, receiving (SlO) an ACELP coded signal comprising a plurality of blocks. For each received block estimating (S20) a signal type based on at least one of the received signal and a set of decoder parameters, estimating (S30) a pitch frequency based on at least one of the received signal and the set of decoder parameters, and determining (S40) filtering parameters based on at least one of the estimated signal type and the estimated pitch frequency. Finally, high pass filtering (S50) the received signal based on the determined filter parameters to provide a high pass filtered output signal.