Ati Technologies Inc.

Ati Technologies Ulc

Daniel W Wong

Milivoje M Aleksic

Philip L Swan

Henry Hing Law

Daniel Doswald

Kenneth Man

A video decoding method and apparatus receives a motion compensation shader command, such as a packet, for a programmable shader of a 3D pipeline, such as programmable vertex shaders and pixel shaders, to provide motion compensation for encoded video, and decode the encoded video using the programmable shader of the 3D pipeline. As such, the programmable shader of a 3D pipeline is used to provide motion compensated video decoding as opposed to, for example, dedicated hardware, thereby, among other advantages, eliminating the need for the dedicated hardware.

Programmable shader-based motion compensation apparatus and method

A de-interlacer includes recursive motion history map generating circuitry operative to determine a motion value associated with one or more pixels in interlaced fields based on pixel intensity information from at least two neighboring same polarity fields. The recursive motion history map generating circuitry generates a motion history map containing recursively generated motion history values for use in de-interlacing interlaced fields wherein the recursively generated motion history values are based, at least in part, on a decay function.

Method and de-interlacing apparatus that employs recursively generated motion history maps

A method and method and apparatus for data re-arrangement includes the steps of receiving output pixel coordinates (X, Y) and obtaining an input pixel offset value (&#x394;S, &#x394;T), wherein the output pixel coordinate represents a location for a two dimensional matrix. The input pixel offset value is obtained in reference to initial input pixel coordinates (S, T) which may be received with the output pixel coordinates or calculated based on the input and/or output pixel coordinates. The input pixel offset value may be any type of representation that provides for a delta value, for example, (&#x394;S, &#x394;T) may represent a shift representation for the offset within a matrix array. The method and apparatus for data re-arrangement further includes retrieving an input pixel based on the initial input pixel coordinates and the offset value.

Method and apparatus for data re-arrangement

A method and system for prevention of unauthorized access to multimedia data are disclosed herein. A tamper-resistant system having a software driver, a peripheral device, and a system memory is used to encrypt sensitive routines used by the software driver. The software driver is used to interface between one component of the system, such as a processor, and a peripheral device, such as a graphics chip. The driver incorporates one or more sensitive routines, that if divulged, could possibly allow an unauthorized party access to data processed by the software driver. Accordingly, in one embodiment, the sensitive routines are stored in an encrypted format with the driver. To access a sensitive routine, the driver submits the associated encrypted routine to the peripheral device, as well as a decryption method, if desired, where it is decrypted and stored in a plaintext format in a location, such as system memory, accessible to both the driver and the peripheral device. The driver can then use the plaintext routine to process the data. When finished processing the data, the plaintext routine can be re-encrypted using one or more of a variety of encryption methods and stored with the driver. Any remaining copies of the plaintext routine can be removed from the system. By encrypting the sensitive routines at all times other than when in immediate use, the system can effectively prevent an unauthorized party from accessing data based on knowledge about the sensitive routine. In addition, the use of the hardware of the peripheral device to encrypt/decrypt the sensitive routines provides an additional barrier to an unauthorized party.

System for preventing unauthorized access to sensitive data and a method thereof

A method of deinterlacing interlaced fields of video for display in a progressive display device includes providing at least one candidate motion vector per scan line and determining, at least one final motion vector per scan line of interlaced video, for use in deinterlacing the interlaced fields, by iteratively changing the at least one candidate motion vector per scan line based on pixel intensities from each of a plurality of same polarity fields at locations along a single dimension.

Method and apparatus for deinterlacing interleaved video

A system and method for handling errors is provided. Errors related to the processing and storage of inverse discrete cosine transform (IDCT) image data cause hardware to become stalled. Stalled hardware may result in multiple image frames being dropped or lost during video playback. To avoid the stalling of hardware, the hardware is used to analyze the data being processed. Data being stored may be analyzed to determine if any error-characteristics, such as overflow or underflow, are present in the storing of the data. The data is manipulated to avoid stalling due to the error-characteristics.

System for handling errors related to IDCT data and method thereof

The invention provides a method and apparatus for an anti-aliasing process that allows for super-sampling at a high subpixel resolution, but does not require the process and memory resources typically required for conventional super-sampling at this subpixel resolution. Each pixel is partitioned into an array that provides for a large number of subpixels, and a smaller set of super-samples from this set of subpixels are sampled and used to determine the resultant pixel values. Because the set of super-samples is substantially smaller in number (less than half) than the number of subpixels, the processing and memory requirements are substantially reduced. The set of super-samples are preferably determined so as to provide for a uniform sampling frequency in each of the major axes, and along each diagonal, even though the super-samples may not provide an uniform sampling of each pixel area.

Method and apparatus for antialiasing using a non-uniform pixel sampling pattern

A method and apparatus for rendering an object to have a bump texture begins when object parameters for the object are received. The object parameters include bump mapping coordinates and physical display coordinates. From this information, a first and second axis specific tables are generated to provide a plurality of axis specific bump intensity values. The first axis specific table relates to bump intensity values along a first axis of the bump and the second axis specific table relates to bump intensity values along a second axis of the bump map. In essence, the axis specific tables represent the bump map being mathematically repositioned to be in the same plane, with respect to the fixed coordinates of the display, as that of the object. With the tables generated, the object is rendered on a pixel by pixel basis, wherein the first and second axis specific tables are addressed for each pixel to retrieve a corresponding first and second intensity values (i.e., B,&times;(fu&middot;L), B,&times;(fv&middot;L)). The first and second intensity values are combined to produce a bump-shading component, or bump intensity value, (&Dgr;N&middot;L) for the given pixel. The bump-shading component (&Dgr;N&middot;L) is then combined with the normal shading component (N&middot;L) to produce the shading function for the given pixel. The resulting shading function is then combined with rendered pixel data to produce the resultant display, or pixe,l data for the given pixel.

Method and apparatus for object rendering including bump mapping

A method and apparatus for determining the level of detail for texture mapping applications in video graphic circuits begins by updating first and second linear components of a set of derivatives when rendering a pixel of an object in a given scan line. The first and second linear components are updated in a linear manner based on first and second increment values, respectively. The set of derivatives corresponds the changes of the texture coordinates with the changes of the physical coordinates of an object being rendered, the derivatives include .delta.s/.delta.x, .delta.t/.delta.x, .delta.s/.delta.y, and .delta.t/.delta.y--in general, the set of derivatives equals .delta.a/.delta.b=w.sup.2 *((1/w)*(.delta.(a/w)/.delta.b-(a/w)*(.delta.(1/w)/.delta.b))). When the object is being rendered on a new scan line, third and fourth linear components are updated in a linear manner based on third and fourth increment values, respectively. As the linear components are updated, their magnitudes are compared, and the one having the greatest magnitude is selected. The selected magnitude is then converted into a logarithmic scale to produce a logarithmic magnitude. The logarithmic magnitude of the linear component is then combined with a logarithmic representation of a non-linear component of a set of derivatives to obtain the level of detail for a given pixel.

Method and apparatus for determining level of detail for texture

A method and apparatus for space variant texture filtering begins when texture coordinates for a given pixel of an object being rendered are received. From the texture coordinates, first and second axis derivatives are determined. When the first and second axis derivatives have differing degrees of magnitude, a determination is made as to whether the first access derivative or the second axis derivative corresponds to a major axis. Next, a number of samples along the major axis are determined based on a logarithmic function of a ratio between the first and second derivatives. Having obtained the number of samples, a separation between the samples is determined based on derivatives along the major axis. Having obtained the number of samples and separation, one of a plurality of related texture maps (e.g., MIP maps) is referenced based on the derivatives along a minor access to obtain the samples. Having obtained the samples, they are filtered based on the number of samples to obtain a filtered texel for the given pixel.

Growing community of inventors

Thornhill, Canada

Daniel W Wong