Browse Prior Art Database

Method for dynamic pixel sampling on a zone renderer using vertex transformations

IP.com Disclosure Number: IPCOM000006502D
Publication Date: 2002-Jan-09
Document File: 5 page(s) / 1K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for dynamic pixel sampling on a zone renderer using vertex transformations. Benefits include a method of tuning performance and quality to whichever setting most benefits the customer without sacrificing the extraordinary performance required for supersampling.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 47% of the total text.

Method for dynamic pixel sampling on a zone renderer using vertex transformations

Disclosed is a method for dynamic pixel sampling on a zone renderer using vertex transformations. Benefits include a method of tuning performance and quality to whichever setting most benefits the customer without sacrificing the extraordinary performance required for supersampling.

              The present invention disclosure describes a method for dynamic pixel sampling on a zone renderer using vertex transformations. Pixel sampling using more than one pixel is defined as multisampling. Multisampling is defined by both supersampling and subsampling. Supersampling is the approach of taking more than one sample for each pixel and combining them. Subsampling is the approach of expanding pixels from a single sample. This method allows dynamically increasing quality from supersampling and dynamically increasing performance from subsampling on a hardware-based zone renderer using hardware-based vertex coordinate transform prescaling.

              Supersampling-based antialiasing increases the quality of the image. Antialiasing is the application of techniques that reduce or eliminate jagged edges of primitives. This method involves samples from surrounding pixels and combines them for each pixel (see Figure 1). This technique reduces rendering performance because each pixel requires multiple samples from surrounding pixels. Using subsampling increases performance while sacrificing quality. This method does the opposite of supersampling by replicating pixels and reducing the number of samples per frame. The disclosed method enables dynamic change between supersampling and subsampling so that the quality and performance can be optimized.

              A zone renderer is graphics hardware that decomposes geometric primitives (for example, triangles and lines) into screen space rectangles called zones. Each primitive is drawn in multiple passes, once for each zone it covers. For each pass, however, only those pixels covered by the primitive and contained in the zone are drawn. The zone renderer processes each zone in turn, drawing all the primitives that intersect the zone. The primary rationale for using the zoned renderer is to improve rendering performance through caching. Because the graphics hardware is only working on a small portion of the screen at a time, the renderer is able to hold the frame buffer contents for the entire zone in an internal cache (see Figure 2). Using this cache significantly reduces the memory traffic and improves performance relative to a traditional renderer that draws each primitive completely before continuing to the next one.

              Hardware-based vertex coordinate transformation is a graphics command that enables scaling and translating of vertex coordinates during hardware primitive processing. Scaling means the size of an object is changed. Scaling is accomplished by multiplying the vertex coordinates by a scale factor. Translating means an object is moved. Translating is accom...