Film Grain Synthesis Using Programmable Templates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digitally recorded images and videos lack the realistic film grain effect of film-recorded media, leading to artificial appearances when digital film grain is added, and this effect is difficult to preserve during encoding.
Innovation Solution
The system uses image processing hardware to generate programmable templates that simulate film grain patterns, pseudo-randomly sample these templates to fetch film grain values, scale them, and combine them with pixel values, employing either a parametric method based on autoregressive models or a spectral method with target spatial frequency spectra and correlation structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If digital film grain is added to digitally recorded images or videos, then the cinematic appearance is improved, but the appearance becomes artificial or unrealistic
Solution Approach 1:
The patent creates a master grain template that captures authentic film grain patterns, which is then reused and pseudo-randomly sampled across multiple frames. This copying approach preserves the authentic film grain characteristics while efficiently applying them to digital content, avoiding the artificial appearance of procedurally generated grain.
Solution Approach 2:
The grain template is generated in advance before the actual video processing. This preliminary creation of the grain pattern allows for consistent, realistic grain to be applied across all frames without requiring complex real-time generation, thereby maintaining realistic appearance while improving processing efficiency.
2Reliability
If film grain is added to video, then the cinematic effect is improved, but the bit rate increases
Solution Approach 1:
Instead of storing actual film grain images for every frame, the patent stores a single master grain template that is then pseudo-randomly sampled and applied to multiple frames. This dramatically reduces the storage requirements and bit rate while maintaining the cinematic film grain effect across the entire video sequence.
Solution Approach 2:
The single grain template serves multiple functions across many video frames. By designing the template to be reusable and adaptable through pseudo-random sampling, the system achieves cinematic grain effects throughout the video without the bit rate overhead of storing separate grain data for each frame.
3Reliability
If film grain is added after decoding, then the grain effect is preserved, but the processing complexity increases
Solution Approach 1:
The grain template is designed to be self-contained with embedded randomness seeds that enable it to self-generate the appropriate grain patterns for each frame through pseudo-random sampling. This self-service capability reduces the need for complex external processing while preserving authentic grain characteristics.
Solution Approach 2:
The patent introduces a grain template as an intermediary element between the decoded video and the final output. This template acts as a mediator that adds grain effects in a standardized, efficient manner, simplifying the overall processing pipeline compared to frame-by-frame grain generation or manipulation.
4Reliability
If film grain is removed before encoding, then the grain can be readded, but the spatiotemporal randomness is difficult to preserve
Solution Approach 1:
The master grain template captures the essential spatiotemporal randomness characteristics of film grain in a compressed form. By copying and pseudo-randomly sampling from this template rather than attempting to preserve and readd actual grain frames, the system maintains the natural randomness of film grain while enabling efficient removal and readdition around encoding operations.
Data Source
AI summary
Systems and methods are provided for adding digital film grain to images and videos displayed by a display of an electronic device. An image processing circuitry of the electronic device may include hardware, such as display pipeline hardware, memory-to-memory scaler and rotator (MSR) hardware, and/or other possible hardware, that enables generation of film grain templates based on characteristics of a target film grain, pseudo-randomly samples the programmable template to fetch film grain values, scales the film grain values, and combines the scaled film grain values with values of pixels in an image frame.