Audio Conference Device Encoder Switchover Synchronization

The patent segments the N participant signals into multiple groups, where only a subset of signals (M < N-1) are superimposed to form the mixed signal for each participant. This segmentation reduces the computational complexity from O(N²) to O(M×N) while maintaining acceptable audio coverage by selectively including only the most relevant participant signals in each mixed signal.

The patent applies local quality by differentiating between useful speech signals and background noise sources. Instead of uniformly superimposing all participant signals, the system selectively includes only M participants whose signals contain useful speech content, excluding those contributing primarily to background noise. This selective approach maintains audio coverage completeness while improving speech intelligibility by reducing noise interference.

The patent applies preliminary action by pre-defining a fixed set of M participants whose signals will be superimposed to form the mixed signal, rather than dynamically adapting the participant set during the conference. This preliminary selection is based on initial audio signal analysis, and the same M participants are consistently included throughout the conference duration. This approach prevents abrupt quality changes and speech clipping that would occur with dynamic adaptation, while still reducing computing outlay compared to including all N-1 participants.

The patent applies universality by using a fixed set of M participants for all mixed signal generation throughout the conference, rather than dynamically reconfiguring which participants are included. This universal approach ensures that the same participant signals are consistently processed by the same encoders, eliminating the encoding and decoding errors that would occur during dynamic switchover between different encoder configurations. The system maintains adaptability through volume control adjustments while preserving encoding stability.