Aggregate Comfort Noise Generation for Audio Conferencing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio devices generate comfort noise with inconsistent spectral shapes, leading to fluctuations in background noise when mixed signals are transmitted from a local environment to a remote location during conferencing sessions.
Innovation Solution
A system and method that generate and add aggregate comfort noise to a mixed audio signal, based on noise spectral estimates from each audio device in the environment, and share metrics related to speech and noise levels to modify noise reduction processing on each device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If each audio device generates comfort noise independently based on local noise sources, then comfort noise provides audible confirmation of connection, but the spectral shapes become inconsistent and cause background noise fluctuations in the mixed output signal
Solution Approach 1:
The patent merges the comfort noise generation function from individual audio devices to a central audio device. Instead of each device generating its own comfort noise independently, the system collects noise spectral estimates from all devices and uses a single centralized comfort noise generator to create one unified comfort noise signal. This eliminates the spectral inconsistencies that arise from multiple independent generators while maintaining the connection confirmation function.
Solution Approach 2:
The patent introduces a noise spectral estimate collector as an intermediary component between the audio devices and the comfort noise generator. This collector gathers noise spectral estimates from multiple audio devices and transmits them to the centralized generator, enabling the system to account for noise characteristics from all devices when generating the unified comfort noise, thus resolving the inconsistency problem.
2Adaptability or versatility
If multiple audio devices of different types and pickup patterns are used to capture sound, then comprehensive audio coverage is achieved, but the comfort noise spectral shapes differ and cause fluctuations in the mixed output signal
Solution Approach 1:
The patent combines the comfort noise generation functions from multiple audio devices into a single centralized generator. The system collects noise spectral estimates from all audio devices regardless of their types or pickup patterns and uses these estimates to generate one unified comfort noise signal, eliminating the spectral variations that would otherwise cause background noise fluctuations.
Solution Approach 2:
The centralized comfort noise generator serves all audio devices universally, regardless of their specific types, pickup patterns, or locations. The generator uses noise spectral estimates from the entire audio system to create a unified comfort noise that is consistent across all devices, making the system adaptable to various audio capture configurations while maintaining noise consistency.
3Reliability
If comfort noise is generated and added to audio signals before transmission to automixer, then connection confirmation is provided, but inconsistent spectral shapes from different devices cause background noise fluctuations in the mixed output
Solution Approach 1:
The patent merges the comfort noise generation process into a single centralized function that operates after the automixer. Instead of adding comfort noise to individual device signals before mixing, the system generates one unified comfort noise signal using aggregated noise spectral estimates and adds it to the final mixed output. This eliminates the harmful fluctuations caused by multiple independent comfort noise generators with different spectral shapes.
Data Source
AI summary
Systems and methods are disclosed for providing a more consistent background noise for mixed audio transmitted from a local environment to a remote location. The systems and methods may generate and add aggregate comfort noise to a mixed audio signal, and the aggregate comfort noise may be based on noise spectral estimates from each audio device in an environment. Metrics related to speech and noise levels may be shared between the audio devices in an environment to modify the noise reduction processing of audio signals on each audio device.