Blendercave: A Multimodal Scene Graph Editor For Virtual Reality

Abstract

This paper presents the BlenderCAVE project, which extends the 3D creation content software Blender and its Game Engine (BGE) to Virtual Reality (VR) applications. Based on a multi-screen nonstereoscopic adaptation of the BGE [Gascon et al., 2010], Blender- CAVE now integrates a complete framework dedicated to Virtual Reality (VR), compatible with the three main Operating Systems for any given VR architecture configuration. It has been developed by audio and VR researchers with support from the Blender Community on LIMSI’s state of the art VR platforms. Acting as a Scene Graph, BlenderCAVE handles multi-screen/multi-user tracked stereoscopic rendering through an efficient low-level master/ slave synchronization process while controlling spatial audio rendering (ambisonic, multi-user binaural, WFS, etc.) and haptic events through OSC and VRPN protocols. The scene creation process itself is reduced to simple Blender manipulations including basic python programming easily carried out using standard laptops. OSC client and spatial audio rendering methods have thus far been implemented in the Max/MSP Audio Programming Environment.