Pixel Storage for HDR Frame Buffers Using Lossy Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current pixel formats for High Dynamic Range (HDR) graphics systems fail to meet the requirements of packing efficiency, random access, and real-time access, making it challenging to implement a cost-effective HDR graphics system that maintains high visual quality without introducing quantization artifacts.
Innovation Solution
A lossy compression process is employed for storing pixels in HDR graphics systems, utilizing a targeted number of bits per pixel to represent statistical redundancy, which allows for efficient packing and real-time access while maintaining sufficient visual quality, using formats like RGBA10 or FP16 that balance bit usage and storage efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high bit depth (10, 12, 16, 24, 32 bits per sample) is used for HDR pixel storage, then the dynamic range and visual quality are improved, but the memory bandwidth and storage requirements increase significantly
Solution Approach 1:
The pixel data is segmented into multiple components (e.g., YCbCr color space separation) where only the luminance component requires high precision for HDR, while chrominance components can use lower precision. This segmentation allows selective application of high bit depth only where necessary for visual quality, reducing overall memory bandwidth requirements.
Solution Approach 2:
The patent changes the parameter representation by using floating-point formats (e.g., FP16) instead of fixed-point integer formats. This allows dynamic allocation of precision where needed while using fewer bits overall, optimizing the balance between visual quality and memory bandwidth for HDR content.
2Measurement precision
If high bit depth pixel formats are used for HDR, then the visual quality and dynamic range are improved, but the packing efficiency and storage cost worsen
Solution Approach 1:
By segmenting pixel data into luminance and chrominance components and applying different precision levels to each, the system achieves high visual quality where needed while reducing overall storage requirements, directly lowering storage costs for HDR content.
Solution Approach 2:
The patent employs lossy compression techniques that discard imperceptible data details, effectively using a cheaper storage approach that maintains acceptable visual quality. This trade-off reduces storage costs while preserving the essential HDR visual experience.
3Measurement precision
If lossless compression is used for pixel storage, then the visual quality is maintained, but the compression ratio and real-time access performance worsen
Solution Approach 1:
Instead of applying full lossless compression to all pixel data, the patent applies partial compression by selectively discarding only the least significant bits that are imperceptible to human vision. This partial action maintains real-time access performance while preserving sufficient visual quality for HDR content.
Solution Approach 2:
The patent changes from lossless to lossy compression parameters, accepting controlled data degradation in exchange for significantly improved compression ratios and real-time access performance. This parameter change enables HDR content to be processed in real-time while maintaining acceptable visual quality.
4Quantity of substance
If standard 8-bit per sample format is used, then the memory bandwidth and storage cost are reduced, but the dynamic range and visual quality worsen due to quantization artifacts
Solution Approach 1:
The patent transitions from the traditional 8-bit fixed-point dimension to a floating-point dimension (e.g., FP16), enabling more efficient use of memory bandwidth through dynamic precision allocation. This dimensional change allows HDR content to achieve high visual quality with reduced bandwidth requirements compared to naive 10-bit or 12-bit formats.
Solution Approach 2:
By changing the data representation parameter from fixed-point integers to floating-point numbers, the system achieves better precision-to-bandwidth ratio. Floating-point formats allow the same number of bits to represent both very bright and very dark regions accurately, maintaining visual quality while reducing memory bandwidth requirements for HDR content.
Data Source
AI summary
A device implementing the subject pixel storage for graphical frame buffers may include at least one processor configured to obtain a plurality of data units containing a plurality of pixels stored in memory, each of the plurality of data units including a first pixel of the plurality of pixels packed in succession with at least a portion of a second pixel of the plurality of pixels, in which the plurality of pixels is represented by a number of bits, obtain a group of pixels from the plurality of pixels, and store the group of pixels using a targeted number of bits. A method and computer program product implementing the subject pixel storage for graphical frame buffers is also provided.