EGGH01: SIGGRAPH/Eurographics Workshop on Graphics Hardware 2001

Permanent URI for this collection

https://diglib7.eg.org/handle/10.2312/334

Browse

Now showing 1 - 14 of 14

Compiling to a VLIW Fragment Pipeline
(The Eurographics Association, 2001) Mark, William R.; Proudfoot, Kekoa; Kurt Akeley and Ulrich Neumann
The latest generation of graphics hardware supports fully programmable vertex and pixel/fragment operations, but programming this hardware at a low level is difficult and time consuming. To address this problem, we have developed a complete real-time procedural shading system that compiles a high-level shading language to programmable vertex and fragment hardware, as described in a separate publication. In this paper, we describe in detail the algorithms used by this system to generate and optimize fragment code for NVIDIAs register combiner architecture and show that our compiler generates efficient code. The register combiner architecture has some similarities to WIW CPU architectures, so we compare our compilation algorithms to those described in the literature for VLIW CPU architectures. We also discuss some of the lessons we leamed from building and using this compiler that may be useful to the designers of future programmable graphics hardware.
The F-Buffer: A Rasterization-Order FIFO Buffer for Multi-Pass Rendering
(The Eurographics Association, 2001) Mark, William R.; Proudfoot, Kekoa; Kurt Akeley and Ulrich Neumann
Multi-pass rendering is a common method of virtualizing graphics hardware to overcome limited resources. Most current multi-pass rendering techniques use the RGBA framebuffer to store intermediate results between each pass. This method of storing intermediate results makes it difficult to conectly render partially-transparent surfaces, and reduces the performance of shaders that need to preserve more than one intermediate result between passes. We propose an alternative approach to storing intermediate results that solves these problems. This approach stores intermediate colors (or other values) that are generated by a rendering pass in a FIFO buffer as the values exit the fragment pipeline. On a subsequent pass, the contents of the FIFO buffer are fed into the top of the fragment pipeline. We refer to this FIFO buffer as a fragment-stream buffer (or F-buffer), because this approach has the effect of associating intermediate results with particular rasterization fragments, rather than with an (x,y) location in the framebuffer. Implementing an F-buffer requires some changes to current mainstream graphics architectures, but these changes can be minor. We describe the designs pace associated with implementing an F-buffer, and compare the F-buffer to recirculating pipeline designs. We implement F-buffers in the Mesa software renderer, and demonstrate our programmable-shading system running on top of this renderer.
Fast Volumetric Deformation On General Purpose Hardware
(The Eurographics Association, 2001) Rezk-Salama, C.; Scheuering, M.; Soza, G.; Greiner, G.; Kurt Akeley and Ulrich Neumann
High performance deformation of volumetric objects is a common problem in computer graphics that has not yet been handled sufficiently. As a supplement to 3D texture based volume rendering, a novel approach is presented, which adaptively subdivides the volume into piecewise linear patches. An appropriate mathematical model based on trilinear interpolation and its approximations is proposed. New optimizations are introduced in this paper which are especially tailored to an efficient implementation using general purpose rasterization hardware, including new technologies, such as vertex programs and pixel shaders. Additionally, a high performance model for local illumination calculation is introduced, which meets the aesthetic requirements of visual arts and entertainment. The results demonstrate the significant performance benefit and allow for time-critical applications, such as computer assisted surgery.
Hardware Support for Adaptive Subdivision Surface Rendering
(The Eurographics Association, 2001) Boo, M.; Amor, M.; Doggett, M.; Hirche, J.; Strasser, W.; Kurt Akeley and Ulrich Neumann
Adaptive subdivision of triangular meshes is highly desirable for surface generation algorithms including adaptive displacement mapping in which a highly detailed model can be constructed from a coarse triangle mesh and a displacement map. The communication requirements between the CPU and the graphics pipeline can be reduced if more detailed and complex surfaces are generated, as in displacement mapping, by an adaptive tessellation unit which is part of the graphics pipeline. Generating subdivision surfaces requires a large amount of memory in which multiple arbitrary accesses are required to neighbouring vertices to calculate the new vertices. In this paper we present a meshing scheme and new architecture for the implementation of adaptive subdivision of triangular meshes that allows for quick access using a small memory making it feasible in hardware, while at the same time allowing for new vertices to be adaptively inserted. The architecutre is regular and characterized by an efficient data management that minimizes the data storage and avoids the wait cycles that would be associated with them ultiple data accesses equired for traditional subdivision. This architecture is presented as an improvement for adaptive displacement mapping algorithms, but could also be used for adaptive subdivision surface generation in hardware.
Hardware Support for Non-photorealistic Rendering
(The Eurographics Association, 2001) Raskar, Ramesh; Kurt Akeley and Ulrich Neumann
Special features such as ridges, valleys and silhouettes, of a polygonal scene are usually displayed by explicitly identifying and then rendering 'edges' for the corresponding geometry. The candidate edges are identified using the connectivity information, which requires preprocessing of the data. We present a nonobvious but surprisingly simple to implement technique to render such features without connectivity information or preprocessing. At the hardware level, based only on the vertices of a given flat polygon, we introduce new polygons, with appropriate color, shape and orientation, so that they eventually appear as special features.
High-Quality Pre-lntegrated Volume Rendering
(The Eurographics Association, 2001) Engel, Klaus; Kraus, Martin; Ertl, Thomas; Kurt Akeley and Ulrich Neumann
We introduce a novel texture-based volume rendering approach that achieves the image quality of the best post-shading approaches with far less slices. It is suitable for new flexible consumer graphics hardware and provides high image quality even for low-resolution volume data and nonlinear transfer functions with high frequencies, without the performance overhead caused by rendering additional interpolated slices. This is especially useful for volumetric effects in computer games and professional scientific volume visualization, which heavily depend on memory bandwidth and rasterization power. We present an implementation of the algorithm on current programmable consumer graphics hardware using multi-textues with advanced texture fetch and pixel shading operations. We implemented direct volume rendering, volume shading, arbitrary number of isosurfaces, and mixed moder endering. The performance does neither depend on the number of isosurfaces nor the definition of the transfer functions, and is therefore suited for interactive highquality volume graphics.
Incremental and Hierarchical Hilbert Order Edge Equation Polygon Rasterization
(The Eurographics Association, 2001) McCool, Michael D.; Wales, Cluis; Moule, Kevin; Kurt Akeley and Ulrich Neumann
A rasterization algorithm must efficiently generate pixel fragments from geometric descriptions of primitives. ln order to accomplish per-pixel shading, shading parameters must also be interpolated across the primitive in a perspective-correct manner. lf some of these parameters are to be interpreted in later stages of the pipeline directly or indirectly as texture coordinates, then translating spatial and parametric coherence into temporal coherence will improve texture cache performance. Finally, if framebuffer access is also organized around cached blocks, then organizing rasterization so fragments are generated in block-sequential order will maximize framebuffer cache performance. Hilbert-order rasterization accomplishes these goals, and also permits efficient incrementale valuation of edge and interpolation equations.
Perlin Noise Pixel Shaders
(The Eurographics Association, 2001) Hart, John C.; Kurt Akeley and Ulrich Neumann
While working on a method for supporting real-time procedural solid texturing, we developed a general purpose multipass pixel shader to generate the Perlin noise function. We implemented this algorithm on SGI workstations using accelerated OpenGL PixelMap and PixelTransfer operations, achieving a rate of 2.5 Hz for a 256x256 image. We also implemented the noise algorithm on the NVidia GeForce2 using register combiners. Our register combiner implementation required 375 passes, but ran at 1.3 Hz. This exercise illustrated a variety of abilities and shortcomings of current graphics hardware. The paper concludes with an exploration of directions for expanding pixel shading hardware to further support iterative multipass pixel-shader applications.
Quasi-Linear Depth Buffers With Variable Resolution
(The Eurographics Association, 2001) Lapidous, Eugene; Jiao, Guofang; Zhang, Jianbo; Wilson, Timothy; Kurt Akeley and Ulrich Neumann
In this paper we present new class of variable-resolution depth buffers, providing a flexible trade-off between depth precision in the distant areas of the view volume and performance. These depth buffers can be implemented using linear or quasi-linear mapping function of the distance to the camera to the depth in the screens pace. In particular, the complementary Z buffer algorithm combines simplicity of implementation with significant bandwidth savings. A variable-resolution depth buffer saves bandwidth by changing size of the per-pixel depth access from 24 bits to 16 bits when distance to the pixel from the camera becomes larger than a given threshold, This distance is selected in order to keep the resulting resolution equal or larger than the resolution of the 24-bit screen Z buffer. For dynamic ratios of the distances between far and near planes 500 and above, bandwidth savings may exceed 20%. Quasi-linear depth floating-point depth buffers are best at high dynamic ratios; 3D hardware should support per-frame setting of the optimal depth buffer type and format. Per-frame adjustment of the resolution switch distance allows balancing performance with depth precision and should be exposed in the graphics A PI.
R-Buffer: A Pointerless A-Buffer Hardware Architecture
(The Eurographics Association, 2001) Wittenbrink, Craig M.; Kurt Akeley and Ulrich Neumann
We present a graphics hardware architecture that implements Carpenter's A-buffer. The A-buffer is a software renderer that uses pointer based linked lists. Our pointerless approach computes order independent transparency for any number of layers with minimal hardware complexity. Statistics are shown for a variety of different scenes using a trace based methodology, with an instrumented Mesa OpenGL implementation. The architecture is shown to require from 2.1 to 3.6 times more memory than traditional Z-buffering. A detailed hardware design is provided, Order independent transparency is computed without application sorting and without artifacts. The architecture can also be used for antialiasing, and an example of Carpenter's classical A-buffer antialiasing is shown.
Real-Time Bump Map Synthesis
(The Eurographics Association, 2001) Kautz, Jan; Heidrich, Wolfgang; Seidel, Hans-Peter; Kurt Akeley and Ulrich Neumann
In this paper we present a method that automatically synthesizes bump maps at arbitrary levels of detail in real-time. The only input data we require is a normal density function; the bump map is generated according to that function. It is also used to shade the generated bump map. The technique allows to infinitely zoom into the surface, because more (consistent) detail can be created on the fly. The shading of such a surface is consistent when displayed at different distances to the viewer (assuming that the surface structure is self-similar). The bump map generation and the shading algorithm can also be used separately.
SPAF: Sub-texel Precision Anisotropic Filtering
(The Eurographics Association, 2001) Shin, Hyun-Chul; Lee, Jin-Aeon; Kim, Lee-Sup; Kurt Akeley and Ulrich Neumann
Texture mapping is a technique which most effectively improves the realism of computer-generated scenes in 3D Graphics. Trilinear filtering of the mip-mapped textue has been popular as a texture filtering method but it blurs images on the surface of objects angled obliquely away from the viewer in a scene. Various anisotropic filtering methods like footprint assembly, Feline, and fast footprint mip-mapping have been proposed to satisfy the desire for the high quality image [7]. In spite of the increase of the memory bandwidth, tie memory bandwidth limit is still the bottleneck of the texture filtering hardware. Moreover, it is very important to keep the quality of rendered image good. In this paper, we propose Sub-texel Precision Anisonopic Filtering (SPAF) which filters texels in a region that covers a quadrilateral footprint with the weights. The weight plays a key role in effective filtering to render the image of high quality with the restricted number of texels loaded from memory for real-time filtering. First, the area coverage based texel filtering scheme is introduced to obtain the footprint's coverage for each texel on the sub-texel precision leading to the small weight table size. Second, the Gaussian weight is applied to this footprinfs coverage for each texel to reduce the artifacts. Therefore, the quality of rendered images is superior to other anisotropic filtering methods in the same restricted number of texels. And the size ofthis weight table is several hundred KBytes which is much smaller than fast footprint mip-mapping. This small ROM table size enables the SPAF to be implemented at feasible hardware costs.
Vertex-based Anisotropic Texturing
(The Eurographics Association, 2001) Olano, Marc; Mukherjee, Shrijeet; Dorbie, Angus; Kurt Akeley and Ulrich Neumann
MIP mapping is a common method used by graphics hardware to avoid texture aliasing. In many situations, MIP mapping over-blurs in one direction to prevent aliasing in another. Anisotropic texturing reduces this blurring by allowing differing degrees of filtering in different directions, but is not as common in hardware due to the implementation complexity of current techniques. We present a new algorithm that enables anisotropic texturing on any current MIP map graphics hardware supporting MIP level biasing, available in OpenGL 7.2 or through the GLEXT-texture-lod-bias or GL-SGIX-texture-lod-bias OpenGL extensions. The new algorithm computes anisotropic filter footprint parameters per vertex. It constructs the anisotropic filter out of several MIP map texturing passes or multitexture lookups. Each lookup uses MIP level bias and perturbed texture coordinates to place one probe used to construct the more complex filter profile.
Watertight Tessellation using Forward Differencing
(The Eurographics Association, 2001) Moreton, Henry; Kurt Akeley and Ulrich Neumann
In this paper we describe an algorithm and hardware for the tessellation of polynomial surfaces. While conventional forward difference-based tessellation is subject to round off error and cracking, our algorithm produces a bit-for-bit consistent triangle mesh across multiple independently tessellated patches. We present tessellation patterns that exploit the efficiency of iterative evaluation techniques while delivering a defect free adaptive tessellation with continuous level-of-detail. We also report the rendering performance of the resulting physical hardware implementation.

Browse

Browsing EGGH01: SIGGRAPH/Eurographics Workshop on Graphics Hardware 2001 by Title

Results Per Page

Sort Options