Multichannel Audio Coding With Jointly Encoded Residual Signals

The patent combines multiple residual signals (first residual signal from left channel, second residual signal from right channel) into a single jointly encoded representation. By merging these signals and exploiting their similarities and dependencies, the system achieves efficient coding with reduced bit rate while maintaining audio quality through residual-signal-assisted multi-channel decoding.

The patent uses the jointly encoded representation of residual signals as a compact copy that can be used to reconstruct multiple audio channels. Instead of transmitting full multi-channel audio data, the system transmits a compressed residual representation that can be copied and applied to reconstruct the original channels with high fidelity.

The patent merges the encoding processes for multiple residual signals into a single joint encoding operation. By combining the first residual signal and second residual signal into one jointly encoded representation, the system improves coding efficiency through exploitation of signal dependencies, reducing computational redundancy while preserving audio quality.

The patent changes the encoding parameters by transitioning from separate encoding of residual signals to joint encoding with shared parameters. This parameter change enables the system to exploit correlations between channels more effectively, improving coding efficiency while maintaining the ability to reconstruct high-quality audio signals.