1 / 126

Scalable many-light methods

Explore the scalable many-light methods for instant radiosity with glossy surfaces, generating and rendering with virtual point lights (VPLs) to approximate indirect illumination. Understand the challenges and scalability issues of this technique.

philipn
Download Presentation

Scalable many-light methods

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Scalable many-light methods Jaroslav Křivánek Charles University in Prague

  2. Instant radiosity • Approximate indirect illumination by Virtual Point Lights (VPLs) • Render with VPLs Generate VPLs

  3. Instant radiosity with glossy surfaces • LargenumberofVPLsrequired • Trueevenfordiffusescenes • Scalabilityissues Ground truth 1,000 VPLs 100,000 VPLs

  4. Scalable many-light methods Generate many, many VPLs Pick only the most relevant VPLs for rendering

  5. Scalable many-light methods • Choosing the right VPLs • Per-pixel basis • Lightcuts [Walter et al 05/06] • Per-image basis • Matrix Row Column Sampling [Hašan et al. 07] • Somewhere in-between • LightSlice[Ou & Pellacini 2011] • Importance caching [Georgiev et al. 2012]

  6. Scalable many-light renderingLightcutsMultidimensional Lightcuts Walter et al., SIGGRAPH 2005/2006 Slides courtesy Bruce Walter: • http://www.graphics.cornell.edu/~bjw/papers.html

  7. Lightcuts • http://www.graphics.cornell.edu/~bjw/papers.html

  8. Complex Lighting • Simulate complex illumination using VPLs • Area lights • HDR environment maps • Sun & sky light • Indirect illumination • Unifies illumination Area lights + Sun/sky + Indirect

  9. Scalable • Scalable solution for many point lights • Thousands to millions • Sub-linear cost Tableau Scene

  10. Lightcuts Problem Visible surface

  11. Lightcuts Problem

  12. Lightcuts Problem Camera

  13. Key Concepts • Light Cluster • Approximate many lights by a single brighter light (the representative light)

  14. Key Concepts • Light Cluster • Light Tree • Binary tree of lights and clusters Clusters Individual Lights

  15. Key Concepts • Light Cluster • Light Tree • A Cut • A set of nodes that partitions the lights into clusters

  16. Simple Example Light Tree Representative 4 #1 #4 #2 #3 Light Clusters 1 4 Individual 1 2 3 4 Lights

  17. Three Example Cuts Three Cuts #1 #1 #1 #4 #4 #2 #3 #4 4 4 4 1 4 1 4 1 4 1 2 3 4 1 2 3 4 1 2 3 4

  18. Three Example Cuts Three Cuts #1 #1 #1 #4 #4 #2 #3 #4 4 4 4 1 4 1 4 1 4 1 2 3 4 1 2 3 4 1 2 3 4 Good Bad Bad

  19. Three Example Cuts Three Cuts #1 #1 #1 #4 #4 #2 #3 #4 4 4 4 1 4 1 4 1 4 1 2 3 4 1 2 3 4 1 2 3 4 Bad Good Bad

  20. Three Example Cuts Three Cuts #1 #1 #1 #4 #4 #2 #3 #4 4 4 4 1 4 1 4 1 4 1 2 3 4 1 2 3 4 1 2 3 4 Good Good Good

  21. Algorithm Overview • Pre-process • Convert illumination to point lights • Build light tree • For each eye ray • Choose a cut to approximate the illumination

  22. Convert Illumination • HDR environment map • Importance sampling • Indirect Illumination • Convert indirect to direct illumination using Instant Radiosity [Keller 97] • Caveats: no caustics, clamping, etc. • More lights = more indirect detail

  23. Build light tree • Cluster spatially close lights with similar orientation Light Tree Representative 4 Light Clusters 1 4 Individual 1 2 3 4 Lights

  24. Choose a cut • Approximate illumination with a bounded error • Different cut for each pixel

  25. Illumination Equation S result=MiGiViIi lights Material term Visibility term Light intensity Geometric term

  26. S lights Cluster Approximation Sum pre-computed during light tree construction result≈ MjGjVjIi j is the representative light Cluster

  27. Cluster Error Bound • Bound each term • Visibility <= 1 (trivial) • Intensity is known • Bound material and geometric terms using cluster bounding volume error<MubGubVubIi S - lights Cluster ub= upper bound

  28. Perceptual Metric • Weber’s Law • Contrast visibility threshold is fixed percentage of signal • Used 2% in our results • Ensure each cluster’s error < visibility threshold • Transitions will not be visible • Used to select cut

  29. Perceptual Metric • Problem: • We don’t know the illumination so we don’t know the threshold either • (because threshold = 2% illumination) • Solution: • As we traverse the tree, gradually improve the illumination estimate. • Stop the traversal if the error bound for all cut nodes is below threshold.

  30. Cut Selection Algorithm • Start with coarse cut (eg, root node) Cut

  31. Cut Selection Algorithm • Select cluster with largest error bound Cut

  32. Cut Selection Algorithm • Refine if error bound > 2% of total Cut

  33. Cut Selection Algorithm Cut

  34. Cut Selection Algorithm Cut

  35. Cut Selection Algorithm Cut

  36. Cut Selection Algorithm • Repeat until cut obeys 2% threshold Cut

  37. Lightcuts (128s) Reference (1096s) Error Error x16 Kitchen, 388K polygons, 4608 lights (72 area sources)

  38. Combined Illumination Lightcuts 128s 4 608 Lights (Area lights only) Avg. 259 shadow rays / pixel Lightcuts 290s 59 672 Lights (Area + Sun/sky + Indirect) Avg. 478 shadow rays / pixel (only 54 to area lights)

  39. Reference Lightcuts Error x 16 Cut size

  40. Scalable • Scalable solution for many point lights • Thousands to millions • Sub-linear cost Tableau Scene Kitchen Scene

  41. Why does it work so well? • Data-driven stratification & importance sampling • Stratification • Clustering of similar lights in the light tree • Importance sampling • Subdividing clusters with high contribution

  42. Main issue Problem: Large cuts in dark areas

  43. Lightcuts Recap • Key ingredients • Upper bound on error • Refinement of the highest-error nodes first

  44. Multidimensional Lightcuts • http://www.graphics.cornell.edu/~bjw/papers.html

  45. Problem • Simulate complex, expensive phenomena • Complex illumination • Anti-aliasing • Motion blur • Participating media • Depth of field

  46. Problem • Simulate complex, expensive phenomena • Complex illumination • Anti-aliasing • Motion blur • Participating media • Depth of field

  47. Problem • Simulate complex, expensive phenomena • Complex illumination • Anti-aliasing • Motion blur • Participating media • Depth of field

  48. Problem • Complex integrals over multiple dimensions • Requires many samples camera

  49. Multidimensional Lightcuts • Solves all integrals simultaneously • Accurate • Scalable

  50. Direct only (relative cost 1x) Direct+Indirect (1.3x) Direct+Indirect+Volume (1.8x) Direct+Indirect+Volume+Motion (2.2x)

More Related