Premium
VidNet: Distributed Processing Environment for Computer Generated Animation
Author(s) -
Weber D.,
Spezialetti M.,
Barada H.
Publication year - 1996
Publication title -
software: practice and experience
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.437
H-Index - 70
eISSN - 1097-024X
pISSN - 0038-0644
DOI - 10.1002/(sici)1097-024x(199602)26:2<237::aid-spe13>3.0.co;2-g
Subject(s) - computer science , animation , computer graphics (images) , computer animation , human–computer interaction
This paper describes the design, protocol, and implementation of VidNet (Video Network), an environment for distributing the generation of photo‐realistic images across the workstations of a local area network. Photo‐realistic images are produced by tracing the interacting light rays within a three‐dimensional computer model. The number of calculations required to ray‐trace a single image severely limits the number of images that can be generated by a single‐processor system. VidNet uses distributed processing techniques to reduce the time required to produce computer generated animation, which may consist of several thousand ray‐traced images. In designing VidNet, however, consideration was given to a number of requirements which would define its impact on the overall computing environment. First, VidNet was designed to minimize its visibility to the users of the workstations. The two scheduling strategies, the assigned and the automatic strategies, used by VidNet to determine the assignment of frame processing to workstations were directly influenced by this restriction. Secondly, VidNet was implemented using only existing technologies. No additional hardware or software was required. The existing software package for ray‐tracing was used to render individual images. Also, VidNet provides a degree of fault tolerance and error recovery for some basic types of failures.