lightcuts: a scalable approach to illumination bruce walter, sebastian fernandez, adam arbree, mike...

Lightcuts: A Scalable Approach to Illumination

Lightcuts: A Scalable Approach to Illumination

Bruce Walter, Sebastian Fernandez, Adam Arbree, Mike Donikian,

Kavita Bala, Donald Greenberg

Program of Computer Graphics, Cornell University

LIGHTCUTS

SIGGRAPH 2005

2

Lightcuts Lightcuts

• Efficient, accurate complex illumination

Environment map lighting & indirectTime 111s

Textured area lights & indirectTime 98s

(640x480, Anti-aliased, Glossy materials)

LIGHTCUTS

SIGGRAPH 2005

3

ScalableScalable

• Scalable solution for many point lights

– Thousands to millions

– Sub-linear cost

0

100

200

300

400

500

600

0 1000 2000 3000 4000

Number of Point Lights

Tim

e (s

ecs)

Standard

Ward

Lightcut

Tableau Scene

LIGHTCUTS

SIGGRAPH 2005

4

Complex LightingComplex Lighting

• Simulate complex illumination using point lights

– Area lights

– HDR environment maps

– Sun & sky light

– Indirect illumination

• Unifies illumination

– Enables tradeoffs between components

Area lights + Sun/sky + Indirect

LIGHTCUTS

SIGGRAPH 2005

5

Related WorkRelated Work• Hierarchical techniques

– Hierarchical radiosity [eg, Hanrahan et al. 91, Smits et al. 94]

– Light hierarchy [Paquette et al. 98]

• Many lights

– [eg, Teller & Hanrahan 93, Ward 94, Shirley et al. 96, Fernandez et al. 2002, Wald et al. 2003]

• Illumination coherence

– [eg, Kok & Jensen 92, Ward 92, Scheel et al. 2002, Krivanek et al. 2005]

• Env map illumination

– [Debevec 98, Agarwal et al. 2003, Kollig & Keller 2003, Ostromoukhov et al. 2004]

• Instant Radiosity

– [Keller 97, Wald et al. 2002]

LIGHTCUTS

SIGGRAPH 2005

6

Talk OverviewTalk Overview

• Lightcuts

– Scalable accurate solution for complex illumination

• Reconstruction cuts

– Builds on lightcuts

– Use smart interpolation to further reduce cost

LIGHTCUTS

SIGGRAPH 2005

7

Lightcuts ProblemLightcuts Problem

Visiblesurface

LIGHTCUTS

SIGGRAPH 2005

8


LIGHTCUTS

SIGGRAPH 2005

9


Camera

LIGHTCUTS

SIGGRAPH 2005

10

Key ConceptsKey Concepts

• Light Cluster

– Approximate many lights by a single brighter light (the representative light)

LIGHTCUTS

SIGGRAPH 2005

11


• Light Cluster

• Light Tree

– Binary tree of lights and clusters

Clusters

IndividualLights

LIGHTCUTS

SIGGRAPH 2005

12


• Light Cluster

• Light Tree

• A Cut

– A set of nodes that partitions the lights into clusters

LIGHTCUTS

SIGGRAPH 2005

13

Simple ExampleSimple Example

#1 #2 #3 #4

1 2 3 4

1 4

Light Tree

Clusters

IndividualLights

RepresentativeLight

4

LIGHTCUTS

SIGGRAPH 2005

14

Three Example CutsThree Example Cuts

1 2 3 4

1 4

4

1 2 3 4

1 4

4

1 2 3 4

1 4

4

Three Cuts

#1 #2 #4 #1 #3 #4 #1 #4

LIGHTCUTS

SIGGRAPH 2005

15


1 2 3 4

1 4

4

1 2 3 4

1 4

4

1 2 3 4

1 4

4

Three Cuts

#1 #2 #4 #1 #3 #4 #1 #4

Good Bad Bad

LIGHTCUTS

SIGGRAPH 2005

16


1 2 3 4

1 4

4

1 2 3 4

1 4

4

1 2 3 4

1 4

4

Three Cuts

#1 #2 #4 #1 #3 #4 #1 #4

Bad Good Bad

LIGHTCUTS

SIGGRAPH 2005

17


1 2 3 4

1 4

4

1 2 3 4

1 4

4

1 2 3 4

1 4

4

Three Cuts

#1 #2 #4 #1 #3 #4 #1 #4

Good Good Good

LIGHTCUTS

SIGGRAPH 2005

18

Algorithm OverviewAlgorithm Overview

• Pre-process

– Convert illumination to point lights

– Build light tree

• For each eye ray

– Choose a cut to approximate the illumination

LIGHTCUTS

SIGGRAPH 2005

19

Convert IlluminationConvert Illumination

• HDR environment map

– Apply captured light to scene

– Convert to directional point lightsusing [Agarwal et al. 2003]

• Indirect Illumination

– Convert indirect to direct illuminationusing Instant Radiosity [Keller 97]

• Caveats: no caustics, clamping, etc.

– More lights = more indirect detail

LIGHTCUTS

SIGGRAPH 2005

20

Algorithm OverviewAlgorithm Overview

• Pre-process

– Convert illumination to point lights

– Build light tree

• For each eye ray

– Choose a cut to approximate the local illumination

• Cost vs. accuracy

• Avoid visible transition artifacts

LIGHTCUTS

SIGGRAPH 2005

21

Perceptual MetricPerceptual 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

LIGHTCUTS

SIGGRAPH 2005

22

Illumination EquationIllumination Equation

Material term

result = Mi Gi Vi Iilights

Geometric term

Visibility term

Light intensity

Currently support diffuse, phong, and Ward

LIGHTCUTS

SIGGRAPH 2005

23


Material term


Geometric term

Visibility term

Light intensity

LIGHTCUTS

SIGGRAPH 2005

24


Material term


Geometric term

Visibility term

Light intensity

LIGHTCUTS

SIGGRAPH 2005

25

Cluster ApproximationCluster Approximation

Cluster

result ~ Mj Gj Vj Iilights

~

j is the representative light

LIGHTCUTS

SIGGRAPH 2005

26

error Mub Gub Vub Ii

Cluster Error BoundCluster Error Bound

Cluster

lights

• Bound each term

– Visibility <= 1 (trivial)

– Intensity is known

– Bound material and geometric terms using cluster bounding volume

ub == upper bound

LIGHTCUTS

SIGGRAPH 2005

27

Cut Selection AlgorithmCut Selection Algorithm

Cut

• Start with coarse cut (eg, root node)

LIGHTCUTS

SIGGRAPH 2005

28


Cut

• Select cluster with largest error bound

LIGHTCUTS

SIGGRAPH 2005

29


Cut

• Refine if error bound > 2% of total

LIGHTCUTS

SIGGRAPH 2005

30


Cut

LIGHTCUTS

SIGGRAPH 2005

31


Cut

LIGHTCUTS

SIGGRAPH 2005

32


Cut

LIGHTCUTS

SIGGRAPH 2005

33


Cut

• Repeat until cut obeys 2% threshold

Lightcuts (128s) Reference (1096s)

Kitchen, 388K polygons, 4608 lights (72 area sources)

Lightcuts (128s) Reference (1096s)

Error Error x16

Kitchen, 388K polygons, 4608 lights (72 area sources)

LIGHTCUTS

SIGGRAPH 2005

36

Combined IlluminationCombined Illumination

Lightcuts 128s

4 608 Lights(Area lights only)

Lightcuts 290s

59 672 Lights(Area + Sun/sky + Indirect)

LIGHTCUTS

SIGGRAPH 2005

37

Combined IlluminationCombined 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)

LIGHTCUTS

SIGGRAPH 2005

38

Lightcuts RecapLightcuts Recap

• Unified illumination handling

• Scalable solution for many lights

– Locally adaptive representation (the cut)

• Analytic cluster error bounds

– Most important lights always sampled

• Perceptual visibility metric

Lightcuts implementation sketch, Petree Hall C, ~4:30pm

LIGHTCUTS

SIGGRAPH 2005

39

Talk OverviewTalk Overview

• Lightcuts

– Scalable accurate solution for complex illumination


– Builds on lightcuts

– Use smart interpolation to further reduce cost

LIGHTCUTS

SIGGRAPH 2005

40

Reconstruction CutsReconstruction Cuts

• Subdivide image into blocks

– Generate samples at corners

• Within blocks

– Interpolate smooth illumination

– Use shadow rays when needed to preserve features

• Shadow boundaries, glossy highlights, etc.

• Anti-aliasing

– (5-50 samples per pixel)

LIGHTCUTS

SIGGRAPH 2005

41

Image SubdivisionImage Subdivision

• Divide into max block size (4x4 blocks)

4x4 block

LIGHTCUTS

SIGGRAPH 2005

42



• Trace multiple eye rays per pixel

• Subdivide blocks if needed

– Based on material, surface normal,and local shadowing configuration

LIGHTCUTS

SIGGRAPH 2005

43






• Compute samples at corners

Samples

LIGHTCUTS

SIGGRAPH 2005

44






• Compute samples at corners

• Shade eye rays using reconstruction cuts

Samples Reconstructioncut

LIGHTCUTS

SIGGRAPH 2005

45

Sample ConstructionSample Construction

• Compute a lightcut at each sample

• For each node on or above the cut

– Create impostor light (directional light)

– Reproduce cluster’s effect at sample

Cluster

ImpostorDirectional light

LIGHTCUTS

SIGGRAPH 2005

46

Reconstruction CutReconstruction Cut

• Top-down traversal of light tree

– Comparing impostors from nearby samples

Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

47


• Recurse if samples differ significantly

Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

48


• Discard if cluster occluded at all samples

Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

49


Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

50


Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

51


• Interpolate if sample impostors are similar

Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

52


• If cluster contribution small enough, shoot shadow ray to representative light

– Lightcut-style evaluation

Not visited

Recurse

Occluded

Interpolate

Shadow ray

LIGHTCUTS

SIGGRAPH 2005

53


Not visited

Recurse

Occluded

Interpolate

Shadow ray

Temple, 2.1M polygons, 505 064 lights, (Sun/sky+Indirect)

Temple, reconstruction cut block size

LIGHTCUTS

SIGGRAPH 2005

56

Result StatisticsResult Statistics

• Temple model (2.1M polys, 505 064 lights)

Image algorithmAvg. eye rays

per pixelImage time

Lightcuts only 1 225s

Combined (anti-aliased) 5.5 189s

Cut type Avg. shadow rays per cut

Lightcut 373

Reconstruction cut 9.4

Grand Central, 1.46M polygons, 143 464 lights, (Area+Sun/sky+Indirect)

Avg. shadow rays per eye ray 46 (0.03%)

Tableau, 630K polygons, 13 000 lights, (EnvMap+Indirect)


Bigscreen, 628K polygons, 639 528 lights, (Area+Indirect)


LIGHTCUTS

SIGGRAPH 2005

60

ConclusionsConclusions

• Lightcuts

– Scalable, unified framework for complex illumination

– Analytic cluster error bounds & perceptual visibility metric


– Exploits coherence

– High-resolution, anti-aliased images

0

100

200

300

400

500

600

0 1000 2000 3000 4000


Tim

e (s

ecs)

Standard

Ward

Lightcut

LIGHTCUTS

SIGGRAPH 2005

61

Future WorkFuture Work

• Visibility bounds

• More light types

– Spot lights etc.

• More BRDF types

– Need cheap tight bounds

• Other illumination types

– Eg, caustics

LIGHTCUTS

SIGGRAPH 2005

62

AcknowledgementsAcknowledgements

• National Science Foundation grant ACI-0205438

• Intel corporation for support and equipment

• The modelers– Kitchen: Jeremiah Fairbanks

– Bigscreen: Will Stokes

– Grand Central: Moreno Piccolotto, Yasemin Kologlu, Anne Briggs, Dana Gettman

– Temple: Veronica Sundstedt, Patrick Ledda, and the graphics group at University of Bristol

– Stanford and Georgia Tech for Buddha and Horse geometry

LIGHTCUTS

SIGGRAPH 2005

63

The EndThe End

• Questions?

Lightcuts implementation sketch, Petree Hall C, ~4:30pm

LIGHTCUTS

SIGGRAPH 2005

64

ScalableScalable

• Scalable solution for many point lights

– Thousands to millions

– Sub-linear cost

0

100

200

300

400

500

600

0 1000 2000 3000 4000 5000


Tim

e (s

ecs)

Standard

Ward

Lightcuts

0

100

200

300

400

500

600

0 1000 2000 3000 4000


Tim

e (s

ecs)

Standard

Ward

Lightcut

Tableau Scene Kitchen Scene

Lightcuts Reference

Error x 16Cut size

Kitchen, 388K polygons, 59,672 Lights

Kitchen, shadow ray false color

0 750 1500

Tableau, shadow ray false color

0 750 1500

Kitchen with sample locations marked

LIGHTCUTS

SIGGRAPH 2005

70

Types of Point LightsTypes of Point Lights

• Omni

– Spherical lights

• Oriented

– Area lights, indirect lights

• Directional

– HDR env maps, sun&sky

lightcuts: a scalable approach to illumination bruce walter, sebastian fernandez, adam arbree, mike...

Documents

lightcuts siggraph

illumination slide

lightcuts problem slide

lightcuts efficient

lightcuts problem camera

representative light

cost slide

badgoodbad slide