Antialiasing with Line Samples
Antialiasing is a necessary component of any high quality renderer. An antialiased image is produced by convolving the scene with an antialiasing filter and sampling the result, or equivalently by solving the antialiasing integral at each pixel. Though methods for analytically computing this integral exist, they require the continuous two-dimensional result of visible-surface computations. Because these computations are expensive, most Tenderers use supersampling, a discontinuous approximation to the integral. We present a new algorithm, line sampling, combining a continuous approximation to the integral with a simple visible-surface algorithm. Line sampling provides high quality antialiasing at significantly lower cost than analytic methods while avoiding the visual artifacts caused by supersampling’s discontinuous nature.
A line sample is a line segment in the image plane, centered at a pixel and spanning the footprint of the antialiasing filter. The segment is intersected with scene polygons, visible subsegments are determined, and the antialiasing integral is computed with those subsegments and a one-dimensional reparameterization of the integral.
On simple scenes where edge directions can be precomputed, one correctly oriented line sample per pixel suffices for antialiasing. Complex scenes can be antialiased by combining multiple line samples weighted according to the orientation of the edges they intersect.
KeywordsLine Sample Visible Surface Edge Orientation Stochastic Sampling IEEE Computer Graphic
Unable to display preview. Download preview PDF.
- 1.Greg Abram, Lee Westover, and Turner Whitted. Efficient Alias-free Rendering Using Bit-Masks and Look-up Tables. Proceedings of SIGGRAPH 85, pages 53–59, July 1985.Google Scholar
- 2.James F. Blinn. Jim Blinn’s Corner: Hyperbolic Interpolation. IEEE Computer Graphics & Applications, 12(4), July 1992.Google Scholar
- 3.J. Bloomenthal. Edge Inference with Applications to Antialiasing. Proceedings of SIGGRAPH 83, pages 157–162, July 1983.Google Scholar
- 4.Loren Carpenter. The A-buffer, an Antialiased Hidden Surface Method. Proceedings of SIGGRAPH 84, pages 103–108, July 1984.Google Scholar
- 5.E. Catmull. A hidden-surface algorithm with anti-aliasing. Proceedings of SIGGRAPH 78, pages 6–11, August 1978.Google Scholar
- 6.Edwin Catmull. An Analytic Visible Surface Algorithm for Independent Pixel Processing. Proceedings of SIGGRAPH 84, pages 109–115, July 1984.Google Scholar
- 9.Mark A. Z. Dippé and Erling Henry Wold. Antialiasing Through Stochastic Sampling. Proceedings of SIGGRAPH 85, pages 69–78, 1985.Google Scholar
- 10.T. Duff. Polygon Scan Conversion by Exact Convolution. In Raster Imaging and Digital Typography, pages 154–168. 1989. ISBN 0521374901.Google Scholar
- 11.E. A. Feibush, Marc Levoy, and Robert L. Cook. Synthetic Texturing Using Digital Filters. Proceedings of SIGGRAPH 80, pages 294–301, July 1980.Google Scholar
- 12.James D. Foley, Andries van Dam, Steven K. Feiner, and John F. Hughes. Computer Graphics, Principles and Practice, Second Edition. Addison-Wesley, 1990.Google Scholar
- 15.Norman P. Jouppi and Chun-Fa Chang. Z3: An Economical Hardware Technique for High-Quality Antialiasing and Transparency. In Proceedings of the 1999 Eurograph-ics/SIGGRAPH Workshop on Graphics Hardware, pages 85–93. 1999.Google Scholar
- 16.Kenneth I. Joy, Charles W. Grant, Nelson L. Max, and Lansing Hatfield, editors. Tutorial: Computer Graphics: Image Synthesis. Computer Society Press, 1988.Google Scholar
- 17.Mark E. Lee, Richard A. Redner, and Samuel P. Uselton. Statistically Optimized Sampling for Distributed Ray Tracing. Proceedings of SIGGRAPH 85, pages 61–67, July 1985.Google Scholar
- 19.Michael D. McCool. Analytic Antialiasing With Prism Splines. Proceedings of SIGGRAPH 95, pages 429–436, August 1995.Google Scholar
- 20.Don P. Mitchell. Consequences of Stratified Sampling in Graphics. Proceedings of SIGGRAPH 96, pages 277–280, August 1996.Google Scholar
- 21.Don P. Mitchell. Generating Antialiased Images at Low Sampling Densities. Proceedings of SIGGRAPH 87, pages 65–72, July 1987.Google Scholar
- 22.Steven Molnar. Efficient Supersampling Antialiasing for High-performance Architectures. Technical Report 91–023, University of North Carolina, 1991.Google Scholar
- 24.Juan Pineda. A Parallel Algorithm for Polygon Rasterization. Proceedings of SIGGRAPH 88, pages 17–20, August 1988.Google Scholar
- 25.Toshimitsu Tanaka and Tokiichiro Takahashi. Cross Scanline Algorithm. Proceedings of EUROGRAPHICS ’90, pages 63–74, 1990.Google Scholar