Audio Signal Processing Apparatus for Higher Order Ambisonics Rendering

The patent segments the HOA signal into two distinct components: object signals (individual sound sources) and residual signals (ambient audio scene). This segmentation is achieved through matrix decomposition of the HOA coefficients, allowing separate processing of object and ambient components. The object-based rendering preserves position accuracy while channel-based rendering maintains immersive quality.

Different rendering techniques are applied to different signal components: object-based rendering is used for object signals to preserve position accuracy, while channel-based rendering is used for residual signals to maintain immersive sound quality. This local differentiation of processing methods resolves the contradiction between the two performance requirements.

The patent segments the HOA signal to identify only the object signal components that require position-accurate rendering. By separating object signals from residual signals through matrix decomposition, computational resources are focused only on the necessary components, reducing overall complexity while maintaining position accuracy where needed.

Instead of applying computationally intensive object-based rendering to all HOA components, the patent applies it only partially to the object signal components identified through decomposition. The residual components use simpler channel-based rendering, achieving adequate performance with reduced computational burden.