Jittered Filter Taps for Real-Time Ray-Traced Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ray-tracing techniques require a large number of samples for accurate rendering, leading to high computational demands that hinder real-time applications, and existing spatial filters fail to account for the directional nature of light, resulting in noisy renders and visual artifacts.
Innovation Solution
Implementing temporally and spatially filtered approaches using jittered filter taps and anisotropic spatial filters, which are sparsely sampled and applied as separable filters to reduce computational requirements and enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of ray-traced samples are used for each pixel, then the accuracy of lighting conditions is improved, but the computational resources and rendering time increase significantly
Solution Approach 1:
The patent applies partial sampling by using a reduced set of ray-traced samples combined with spatial filtering. Instead of fully sampling every pixel with numerous rays, the system sparsely samples pixels and uses spatial filters to reconstruct the full image, achieving acceptable accuracy with fewer computational resources.
Solution Approach 2:
The patent introduces spatial filters as an intermediary between sparse ray-traced samples and the final rendered image. These filters act as mediators that interpolate and smooth the sparsely sampled data to produce visually pleasing results without requiring dense sampling at every pixel.
2Device complexity
If conventional spatial filters are used that assume axis-aligned and isotropic filtering, then the filtering process is simplified, but the directional nature of light is not accounted for resulting in lower quality renders
Solution Approach 1:
The patent applies local quality by using anisotropic spatial filters that adapt to the directional characteristics of light at each pixel location. Instead of using uniform axis-aligned filters throughout the image, the system employs filters with varying orientations and shapes that match the local lighting geometry, such as using elongated filters along reflection directions.
3Measurement precision
If a filter kernel that does not follow a true Gaussian distribution is used to enhance spatial filter accuracy, then the filtering precision is improved, but the separable filtering optimization is no longer valid
Solution Approach 1:
The patent applies partial sampling to the filter kernel itself by using a subset of tap locations rather than evaluating all points in a continuous Gaussian kernel. This discrete, sparsely sampled approach approximates the desired filtering effect while maintaining computational efficiency through separable filtering operations.
Data Source
AI summary
Various approaches are disclosed to temporally and spatially filter noisy image data—generated using one or more ray-tracing effects—in a graphically rendered image. Rather than fully sampling data values using spatial filters, the data values may be sparsely sampled using filter taps within the spatial filters. To account for the sparse sampling, locations of filter taps may be jittered spatially and/or temporally. For filtering efficiency, a size of a spatial filter may be reduced when historical data values are used to temporally filter pixels. Further, data values filtered using a temporal filter may be clamped to avoid ghosting. For further filtering efficiency, a spatial filter may be applied as a separable filter in which the filtering for a filter direction may be performed over multiple iterations using reducing filter widths, decreasing the chance of visual artifacts when the spatial filter does not follow a true Gaussian distribution.