Toward efficient and accurate order-independent transparency
Ethan Kerzner, Chris Wyman, Lee A. Butler, Christiaan Gribble · 2013
Correctly rendering multi-layered transparent geometry requires accumulating contributions from multiple fragments per pixel. Dynamic A-buffers (e.g., Yang et al's [2010] per-pixel linked lists) achieve this by storing and sorting fragments on-the-fly. We introduce two improvements to recent GPU-based interactive A-buffer techniques. First, our redesigned algorithm uses fewer costly global atomic operations to construct linked lists. Second, we decouple visibility and shading to reduce memory demands of multi-fragment rendering.