Audio Buffer Synchronization for Echo Reduction in VoIP

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Audio conferencing systems using VoIP often suffer from echo and unwanted reverberation due to simultaneous microphone inputs from participants in close proximity, leading to poor sound quality.

Innovation Solution

The system synchronizes and corrects audio buffers from multiple microphone inputs to reduce echo by emphasizing the active speaker's audio and reducing the input from inactive microphones, either locally or through a server, ensuring synchronous playback across remote peer computer systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple microphones are used to capture audio from participants in close proximity, then audio coverage is improved, but echo and reverberation increase

Engineering Contradiction:

Improveaudio coverageVSAvoidecho and reverberation

Core Design Contradiction:

Measurement precisionVSObject-generated harmful factors

Solution Approach 1:

The patent extracts and removes the harmful echo and reverberation components from the audio signal by analyzing the audio buffers from multiple microphones, identifying portions that correspond to echo, and selectively attenuating or removing these portions while preserving the direct speech signals

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the gain parameter of audio buffers from different microphones based on their spatial relationship and contribution to echo, dynamically adjusting the mixing coefficients to emphasize direct speech and suppress reverberating sounds

Inventive Principle:

Principle #35Parameter changes

2Quantity of substance

If audio buffers from multiple microphones are mixed simultaneously, then audio input coverage is improved, but sound quality deteriorates due to echo

Engineering Contradiction:

Improveaudio input coverageVSAvoidsound quality

Core Design Contradiction:

Quantity of substanceVSManufacturing precision

Solution Approach 1:

The patent introduces an intermediary processing stage between microphone input and output where audio buffers are analyzed, synchronized, and selectively combined, using algorithms to determine the optimal mixing strategy that maximizes speech clarity while minimizing echo

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent dynamically adjusts the mixing of audio buffers from multiple microphones based on real-time analysis of speech activity, speaker identification, and echo detection, rather than using a static mixing approach, allowing the system to adapt to changing acoustic conditions

Inventive Principle:

Principle #15Dynamics

3Productivity

If audio buffers are sent to multiple computer systems, then audio distribution is improved, but synchronization issues cause echo at remote peers

Engineering Contradiction:

Improveaudio distributionVSAvoidsynchronization delay

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The patent performs preliminary synchronization of audio buffers before distribution to multiple computer systems, calculating and applying appropriate time offsets to ensure that audio from different microphones arrives at remote peers at the correct time, preventing echo caused by delayed playback

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11425258B2Audio conferencing in a room

Publication Date: 2022.08.23 WAVES AUDIO

AI summary

First and second computer systems and respective first and second microphones thereof receive respective portions of a same audio input signal. Audio buffers received respectively from the first and second computer system include data encoded from respective microphone inputs of the first and the second computer systems. The received audio buffers are synchronized and corrected for gain differences between the received audio buffers to produce corrected audio buffers. The corrected audio buffers are mixed into an output buffer. The synchronization reduces echo when the output buffer is played at a remote peer computer system.