Decorrelation Filters for Simulated Audio Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for simulating early reflections in virtual environments often result in undesired phase artifacts due to constructive and destructive interference from multiple reflections of the same audio signal, detracting from the immersive experience.
Innovation Solution
A computer-implemented method for decorrelating a set of simulated audio signals using a subset of decorrelation filters selected based on a computed combination metric, which measures decorrelation and cross-correlation to ensure the filters are maximally different and effective in reducing phase artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple reflections of the same audio signal are used to simulate early reflections, then the spatial information about the environment is improved, but phase artifacts occur due to constructive and destructive interference
Solution Approach 1:
The audio signal is divided into multiple reflections, each processed by a different decorrelation filter. This segmentation allows each reflection to be independently decorrelated, preventing the phase artifacts that would occur if all reflections used the same filter while preserving the spatial information provided by multiple reflections.
Solution Approach 2:
Decorrelation filters are introduced as intermediary elements between the audio source and the output. These filters act as mediators that modify the phase relationships of reflected signals, eliminating the harmful constructive and destructive interference patterns while maintaining the spatial characteristics of the environment.
2Reliability
If decorrelation filters are applied to reduce phase artifacts, then the immersive experience is improved, but computational complexity increases due to filter selection and application
Solution Approach 1:
A set of candidate decorrelation filters is pre-generated and stored before runtime. This preliminary action allows the system to have multiple filter options available without performing complex filter generation during real-time audio processing, reducing computational complexity while maintaining the ability to select appropriate filters for different acoustic scenarios.
Solution Approach 2:
The system changes parameters of existing filters (such as delay values and frequency responses) to generate a diverse set of candidate filters. This approach is computationally more efficient than generating entirely new filters, as it involves modifying existing filter parameters rather than performing full filter synthesis during runtime.
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
The method effectively reduces phase artifacts and improves the immersive experience by selecting the most suitable subset of decorrelation filters for the specific set of simulated audio signals, ensuring effective decorrelation and minimizing interference.
Implementation Method 1
the playback of multiple copies of the same signal with slightly differing delays causes constructive and destructive interference, resulting in a 'whooshing' sound
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Described herein is a computer-implemented method for decorrelating a set of simulated audio signals in a virtual environment, the method comprising: obtaining (102) a set of decorrelation filters; determining (104) a set of simulated audio signals to be decorrelated; selecting (106) a subset of decorrelation filters from the set of decorrelation filters which are suitable for decorrelating the set of simulated audio signals, wherein the selection is based on a computed combination metric, the combination metric comprising a measure of decorrelation provided by a combination of filters within the set of decorrelation filters; and applying (108) the selected subset to the simulated audio signals.