Multilayer Translucent Shadow Maps for Cinematic Lighting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for computing soft shadows in computer graphics are either computationally expensive or produce artifacts like darkened shadows and light leaks, especially when dealing with complex geometry and area lights, limiting their practicality in production lighting.

Innovation Solution

The use of multilayer translucent shadow maps (MTSMs) that store multiple depth layers and opacity values, allowing for efficient ray tracing of shadows by exploiting spatial coherence and reducing computational costs, while supporting detailed geometry like fur and volume objects.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If distributed ray tracing is used to compute soft shadows, then shadow quality is improved, but computational cost increases significantly

Engineering Contradiction:
Improveshadow qualityVSAvoidcomputational cost
Core Design Contradiction:
Manufacturing precisionVSPower

Solution Approach 1:

The patent segments the shadow computation problem by separating shadow map generation from shadow rendering. Shadow maps are pre-computed from light source views and stored in a database, allowing shadow quality to be maintained while reducing real-time computational cost. The shadow database stores depth information from multiple light source perspectives, enabling efficient shadow determination without performing full distributed ray tracing during rendering.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs shadow computation in advance by generating shadow maps from light source views before the main rendering process. These pre-computed shadow maps are stored in a database and reused during rendering, eliminating the need for expensive real-time distributed ray tracing while maintaining shadow quality.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If object space based soft shadow algorithms are used, then computational speed is improved, but accuracy degrades with geometric complexity

Engineering Contradiction:
Improvecomputational speedVSAvoidshadow accuracy
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The patent transitions from object space computation to image space computation by using shadow maps. Instead of computing shadows in 3D object space which degrades with geometric complexity, the solution projects shadow information into 2D image space from light source views, maintaining accuracy while improving computational efficiency.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Power

If traditional shadow maps are used, then computational cost is reduced, but shadow quality suffers from bias and abasing artifacts

Engineering Contradiction:
Improvecomputational costVSAvoidshadow quality
Core Design Contradiction:
PowerVSManufacturing precision

Solution Approach 1:

The patent segments shadow information into multiple depth layers stored in a database, each representing different depth ranges. This segmentation allows for more accurate shadow determination by selecting appropriate depth layers, reducing bias and abasing artifacts while maintaining computational efficiency compared to traditional single-layer shadow maps.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2261863B1Soft shadows for cinematic lighting for computer graphics
Publication Date: 2011.11.16 PACIFIC DATA IMAGES LLC
  • EP2261863B1 patent drawingFigure 1
  • EP2261863B1 patent drawingFigure 2a~3d
  • EP2261863B1 patent drawingFigure 4~6

AI summary

A method and apparatus for high quality soft shadows for area lights in cinematic lighting for use in computer graphics, such as computer enabled animation. The method is an extension of traditional shadow maps, so it has the advantage of image based shadow methods; the complexity of the present method is independent of geometric complexity. It uses multilayer translucent shadow maps, which can be used to produce high quality soft shadows for scenes with extremely complex geometry, fur, and volume objects. Instead of the traditional sampling and filtering of shadow maps, the present method computes the shadow factor by ray tracing the multilayer translucent shadow map. The result is soft edged shadows of quality similar to that achieved by conventional stochastic ray tracing, but at much lower computational cost.