KOMPASSI-Renderer

KOMPASSI-Renderer is a rendering plugin (VST/LV2/VST3) that serves as a demo/wrapper for the kompassi-encoder and kompassi-decoder.

This is a demo implementation of our new parametric spatial audio renderer for headphone or loudspeaker rendering of multichannel recordings. It retains the flexibility of our earlier method called COMPASS while improving significantly in terms of robustness and audio quality. It uses a new and more advanced parametric model of the sound scene, and an improved rendering solution. The demo plugin is limited at the moment to Ambisonic input and head-tracked binaural delivery, but we intend to gradually update the plugin with additional functionality, such as loudspeaker rendering for flexible setups, and support for custom microphone arrays.

An Advanced Spatial Audio Codec

Our most advanced parametric rendering algorithms all re-written and integrated into a highly configurable encoder and decoder. The renderer plugin, which serves as a demo of the codec, currently supports:

• Ambisonics or tetrahedral/A-format microphone array recordings as input.
• Binaural playback.
• A SOFA loader, allowing for personalised binaural listening.

Sound Scene Modifications

Optionally, the spatial parameter stream produced by the encoder, may be intercepted and augmented before being passed onto the decoder. Currently, this is conducted by the renderer plugin to enable:

• Six-degrees of Freedom (6DoF) support (i.e., accounting for both listener head-rotations and translations away from the recording point).
• Biasing the reproduction to give more prominence to either the ambience in the scene or directional sounds.