Pan

2002 - http://www.users.bigpond.com/tstex/pan.htm - Pan is a system which renders 3D sound at the push of a button using information from a 3D animation. The system spatialises sound sources relative to the viewer as part of a dynamic acoustic simulation of the room in which the animation is taking place. The sound source may be moving, the camera may be moving, the room may be changing size, the materials the walls are made out of may even be changing, and the system will render the 3D audio specific to any given animation in Dolby Digital (5.1 channel) specification audio.

For those with some knowledge of 3D sound, the system's output has the following characteristics: * Doppler shifted and spatialised unreflected sound sources. * Doppler shifted and spatialised primary phantom sound sources. (Six early reflections.) * Dynamic reverberation for varying room shapes and surface materials. * Wall proximity determination. * Enveloped source proximity limitation to avoid clipping. * Attenuation scaling to simulate differing distance related decays for varying relative sound source amplitudes. * HRTF generated elevation cues. (pending)

For those with some knowledge of 3D animation, the system runs within the Houdini 3D (tm) software using CHOPS as an interface with the main 3D sound engine. A special CHOPnet is used to acquire and automatically prepare relevant information about the animation scene before all relevant scene data is exported to the Csound software for 3D sound rendering by a hscript.

This 8meg mpeg animation was prepared especially to demontrate the special features of the Time Space Texture 3D sound system. The audio is a stereo mixdown of what was originally a 5.1 channel recording.

Category:

Film & Animation

Tags:

• Houdini
• 3d sound
• 3d graphics
• binaural
• CHOPs
• Csound
• John Chowning
• Andrew Lyons