Advanced

Depth buffer compression for stochastic motion blur rasterization

Andersson, Magnus LU ; Hasselgren, Jon LU and Akenine-Möller, Tomas LU (2011) High Performance Graphics, 2011 In [Host publication title missing] p.127-134
Abstract
Previous depth buffer compression schemes are tuned for compressing depths values generated when rasterizing static triangles. They provide generous bandwidth usage savings, and are of great importance to graphics processors. However, stochastic rasterization for motion blur and depth of field is becoming a reality even for real-time graphics, and previous depth buffer compression algorithms fail to compress such buffers due to the irregularity of the positions and depths of the rendered samples. Therefore, we present a new algorithm that targets compression of scenes rendered with stochastic motion blur rasterization. If possible, our algorithm fits a single time-dependent predictor function for all the samples in a tile. However,... (More)
Previous depth buffer compression schemes are tuned for compressing depths values generated when rasterizing static triangles. They provide generous bandwidth usage savings, and are of great importance to graphics processors. However, stochastic rasterization for motion blur and depth of field is becoming a reality even for real-time graphics, and previous depth buffer compression algorithms fail to compress such buffers due to the irregularity of the positions and depths of the rendered samples. Therefore, we present a new algorithm that targets compression of scenes rendered with stochastic motion blur rasterization. If possible, our algorithm fits a single time-dependent predictor function for all the samples in a tile. However, sometimes the depths are localized in more than one layer, and we therefore apply a clustering algorithm to split the tile of samples into two layers. One time-dependent predictor function is then created per layer. The residuals between the predictor and the actual depths are then stored as delta corrections. For scenes with moderate motion, our algorithm can compress down to 65% compared to 75% for the previously best algorithm for stochastic buffers. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
in
[Host publication title missing]
editor
Dachsbacher, Carsten; Mark, William and Pantaleoni, Jacopo
pages
127 - 134
publisher
Eurographics Association
conference name
High Performance Graphics, 2011
external identifiers
  • Scopus:80052445146
language
English
LU publication?
yes
id
7a3a7a6e-a1aa-429d-885e-c58d49fe1df4 (old id 2299934)
alternative location
http://fileadmin.cs.lth.se/graphics/research/papers/2011/depth_compress_mb/depth_compress.pdf
date added to LUP
2012-01-23 09:29:26
date last changed
2016-10-13 04:50:19
@misc{7a3a7a6e-a1aa-429d-885e-c58d49fe1df4,
  abstract     = {Previous depth buffer compression schemes are tuned for compressing depths values generated when rasterizing static triangles. They provide generous bandwidth usage savings, and are of great importance to graphics processors. However, stochastic rasterization for motion blur and depth of field is becoming a reality even for real-time graphics, and previous depth buffer compression algorithms fail to compress such buffers due to the irregularity of the positions and depths of the rendered samples. Therefore, we present a new algorithm that targets compression of scenes rendered with stochastic motion blur rasterization. If possible, our algorithm fits a single time-dependent predictor function for all the samples in a tile. However, sometimes the depths are localized in more than one layer, and we therefore apply a clustering algorithm to split the tile of samples into two layers. One time-dependent predictor function is then created per layer. The residuals between the predictor and the actual depths are then stored as delta corrections. For scenes with moderate motion, our algorithm can compress down to 65% compared to 75% for the previously best algorithm for stochastic buffers.},
  author       = {Andersson, Magnus and Hasselgren, Jon and Akenine-Möller, Tomas},
  editor       = {Dachsbacher, Carsten and Mark, William and Pantaleoni, Jacopo},
  language     = {eng},
  pages        = {127--134},
  publisher    = {ARRAY(0xba68cc8)},
  series       = {[Host publication title missing]},
  title        = {Depth buffer compression for stochastic motion blur rasterization},
  year         = {2011},
}