Server-Based Sound Mixing for Multiuser Voice Chat
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multiuser voice chat systems face inefficiencies in audio stream management, as existing methods often result in a cascade of audio streams being transmitted between participants, leading to increased bandwidth usage and complexity.
Innovation Solution
A server-based sound mixing system where each participant's audio is transmitted to a central server, which mixes and modifies the audio streams to exclude the participant's own voice, creating unique streams for each user based on their pose information, such as head orientation and location, and sends these modified streams back to the participants, optimizing audio delivery and reducing unnecessary data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If every participant broadcasts audio to every other participant in a multiuser voice chat system, then complete audio coverage is achieved, but bandwidth usage and system complexity increase significantly
Solution Approach 1:
The patent introduces a server as an intermediary that receives audio from all participants, performs mixing operations, and distributes the mixed audio back to participants. This centralised approach replaces the complex peer-to-peer broadcast model, reducing the number of audio streams from O(n²) to O(n) while maintaining complete audio coverage.
Solution Approach 2:
The patent merges all individual audio streams into a single mixed audio stream on the server. Instead of transmitting separate audio streams between each pair of participants, the server combines all audio sources into unified streams that are then distributed to participants, significantly reducing bandwidth usage and system complexity.
2Reliability
If audio streams are transmitted between all participants in a multiuser voice chat, then direct communication is maintained, but bandwidth consumption increases
Solution Approach 1:
The patent merges all individual audio streams into consolidated mixed streams on the server before distribution. This approach maintains communication quality by preserving all audio sources while reducing total bandwidth consumption by eliminating redundant transmissions of the same audio data to multiple participants.
Solution Approach 2:
The server performs multiple functions: receiving audio from all participants, mixing the audio streams, excluding each participant's own voice, and distributing the customised mixed streams. This multi-functional approach replaces multiple point-to-point transmission channels with a single universal audio distribution system.
3Loss of information
If each participant receives audio from all other participants, then complete audio information is provided, but unnecessary audio data is transmitted
Solution Approach 1:
The patent extracts and excludes each participant's own audio from the mixed stream before transmitting it back to that participant. This extraction process ensures that participants receive complete audio information from all other participants while eliminating the redundancy of hearing their own voice, thereby improving data transmission efficiency.
4Ease of manufacture
If standard audio mixing is used without pose information, then processing simplicity is maintained, but audio spatial accuracy is reduced
Solution Approach 1:
The patent applies local quality by using each participant's pose information (head orientation, position) to customise the audio mixing specifically for that participant. Instead of a universal mixing approach, the server creates participant-specific mixed streams where audio from other participants is spatially positioned according to the receiving participant's orientation, enhancing audio spatial accuracy while maintaining processing feasibility.
Data Source
AI summary
One or more chat servers receives voice signals and pose (location and orientation) signals from devices such as VR headsets associated with respective chat participants. For each participant, the server renders a single stream representing the voices of the other participants, with the voice data in each stream being modified to account for the orientation of the head of the receiving participant. The server sends the streams to the participants for whom the streams are tailored. The voice information representing the chat of the other participants in a stream intended for a particular participant can also be modified to account for the distances between participants and orientations of speakers' heads relative to the particular participant for whom the stream is tailored.