Incrementally baked global illumination

Christan Luksch*, Michael Wimmer, Michael Schwarzler

*Corresponding author for this work

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

9 Citations (Scopus)
47 Downloads (Pure)

Abstract

Global Illumination is affected by the slightest change in a 3D scene, requiring a complete reevaluation of the distributed light. In cases where real-time algorithms are not applicable due to high demands on the achievable accuracy, this recomputation from scratch results in artifacts like flickering or noise, disturbing the visual appearance and negatively affecting interactive lighting design workflows. We propose a novel system tackling this problem by providing incremental updates of a baked global illumination solution after scene modifications, and a re-convergence after a few seconds. Using specifically targeted incremental data structures and prioritization strategies in a many-light global illumination algorithm, we compute a differential update from one illumination state to another. We further demonstrate the use of a novel error balancing strategy making it possible to prioritize the illumination updates.

Original languageEnglish
Title of host publicationProceedings - I3D 2019
Subtitle of host publicationACM SIGGRAPH Symposium on Interactive 3D Graphics and Games
EditorsStephen N. Spencer
PublisherAssociation for Computing Machinery (ACM)
Pages1-10
Number of pages10
ISBN (Electronic)9781450363105
DOIs
Publication statusPublished - 21 May 2019
Event2019 ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games, I3D 2019 - Montreal, Canada
Duration: 21 May 201923 May 2019

Conference

Conference2019 ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games, I3D 2019
Country/TerritoryCanada
CityMontreal
Period21/05/1923/05/19

Keywords

  • Global Illumination
  • Instant Radiosity
  • Lightmaps

Fingerprint

Dive into the research topics of 'Incrementally baked global illumination'. Together they form a unique fingerprint.

Cite this