Real-time interactive applications today create immersive experiences with realistic graphics and physical behavior. Realistic sound can add an extra dimension of realism to these applications but is still an area of open research. Today’s heterogeneous platforms have plenty of compute power available that can be harvested to run advanced sound rendering pipelines in real time. In our work, we show that we can improve interactive applications with novel sound experiences on today’s hardware with OpenCL.
We show that the programming framework provided by the OpenCL API is well suited to implement certain stages of a sound rendering pipeline, and it allows developers to plug these stages into realistic and efficient applications that integrate graphics, physics and sound while leveraging all available compute power.
In this presentation we show a real-time demo of a novel sound rendering pipeline that effectively produces full 3D sound and runs in real time with Intel® SDK for OpenCL* Applications on the CPU.
Next Generation Sound Rendering with OpenCL [PDF] (SIGGRAPH'11 Khronos OpenCL BOF)
Read more about sound rendering research at: http://gamma.cs.unc.edu/research/sound/
OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Khronos.