Dynamic Metadata Transformation for HDR Tone Mapping

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

High dynamic range (HDR) content mastered in one format often requires significant memory storage, longer creation time, and increased complexity for rendering on devices with different color volume transformation (CVT) models, making efficient format transitions challenging.

Innovation Solution

A method to transform metadata from a source CVT model to a target CVT model by computing and mapping a first transfer curve to a second transfer curve, generating content metadata for accurate or perceptually reasonable rendering across different CVT models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If content is archived in each HDR format to be supported, then rendering accuracy across different CVT models is improved, but memory storage requirements increase

Engineering Contradiction:
Improverendering accuracyVSAvoidmemory storage
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts the essential transfer curve characteristics from the source CVT model metadata and maps them to the target CVT model, rather than storing complete content in every format. This extraction approach retains the core rendering information while eliminating redundant data storage.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms metadata parameters from one CVT model to another by computing transfer curves and mapping them between different color volume transformation models. This parameter transformation enables accurate rendering across formats without archiving content in each format.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If content is archived in each HDR format to be supported, then rendering accuracy across different CVT models is improved, but content creation time increases

Engineering Contradiction:
Improverendering accuracyVSAvoidcontent creation time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary transfer curve computation and metadata transformation during content creation, storing the transformed metadata for quick retrieval. This preliminary action eliminates the need for time-consuming format conversion during playback.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If content is archived in each HDR format to be supported, then rendering accuracy across different CVT models is improved, but system complexity increases

Engineering Contradiction:
Improverendering accuracyVSAvoidcontent delivery complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent creates a universal metadata transformation system that can handle multiple CVT models through a common transfer curve mapping approach. This multi-functional system replaces the need for separate content archives for each format, simplifying the content delivery infrastructure.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Adaptability or versatility

If metadata transformation between different CVT models is implemented, then adaptability across HDR formats is improved, but computational complexity increases

Engineering Contradiction:
Improveformat compatibilityVSAvoidcomputational complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent introduces transfer curves as an intermediary representation between different CVT models. This intermediary enables systematic metadata transformation while managing computational complexity through established curve mapping techniques.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3488614B1Transformation of dynamic metadata to support alternate tone rendering
Publication Date: 2022.01.19 DOLBY LABORATORIES LICENSING CORP
  • EP3488614B1 patent drawingFigure 1A
  • EP3488614B1 patent drawingFigure 1B
  • EP3488614B1 patent drawingFigure 2A~2C

AI summary

An existing metadata set that is specific to a color volume transformation model is transformed to a metadata set that is specific to a distinctly different color volume transformation model. For example, source content metadata for a first color volume transformation model is received. This source metadata determines a specific color volume transformation, such as a sigmoidal tone map curve. The specific color volume transformation is mapped to a color volume transformation of a second color volume transformation model, e.g., a Bézier tone map curve. Mapping can be a best fit curve, or a reasonable approximation. Mapping results in metadata values used for the second color volume transformation model (e.g., one or more Bézier curve knee points and anchors). Thus, devices configured for the second color volume transformation model can reasonably render source content according to received source content metadata of the first color volume transformation model.