On Floating-Point Normal Vectors
Loading...
Date
2010
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association and Blackwell Publishing Ltd
Abstract
In this paper we analyze normal vector representations. We derive the error of the most widely used representation, namely 3D floating-point normal vectors. Based on this analysis, we show that, in theory, the discretization error inherent to single precision floating-point normals can be achieved by 250.2 uniformly distributed normals, addressable by 51 bits. We review common sphere parameterizations and show that octahedron normal vectors perform best: they are fast and stable to compute, have a controllable error, and require only 1 bit more than the theoretical optimal discretization with the same error.
Description
@article{10.1111:j.1467-8659.2010.01737.x,
journal = {Computer Graphics Forum},
title = {{On Floating-Point Normal Vectors}},
author = {Meyer, Quirin and Suessmuth, Jochen and Sussner, Gerd and Stamminger, Marc and Greiner, Guenther},
year = {2010},
publisher = {The Eurographics Association and Blackwell Publishing Ltd},
ISSN = {1467-8659},
DOI = {10.1111/j.1467-8659.2010.01737.x}
}