Abstract: Presented herein is an automatic direct memory access engine. In one embodiment, a decoder system for decoding video data, comprises a video decoder and a direct memory access engine. The video decoder decodes portions of the video data and comprises a local buffer and an extractor. The local buffer stores the portions of the video data. The extractor transmits a signal indicating that a portion of the local buffer is available to store another portion of the video data. The direct memory access engine provides the another portion of the video data to the portion of the local buffer, responsive to receiving the signal from the extractor.

Abstract: Presented herein are systems and methods for pixel reordering and selection. A decoded frame is stored in a frame buffer with a particular pixel order and byte order. A pixel feeder fetches portions of the decoded frame and stores portions of the frame in a double buffer with the same pixel order and byte order. An endian swizzle converts the byte ordering to a predetermined format, as needed. Reordering logic changes the pixel order to a predetermined order. Selection logic selects luma and chroma pixels from fetched pixels and provides the luma pixels to a luma pixel register, chroma Cr pixels to a chroma Cr pixel register, and chroma Cb pixels to a chroma Cb pixel register.

Abstract: Presented herein is a system for storing macroblocks for concatenated frames. A decoder system comprises a frame buffer. The frame buffer comprises one or more rows. A particular one of the rows stores macroblocks from a plurality of frames.

Abstract: The present invention is directed to a line address computer for calculating the starting line addresses for lines of a decoded frame. The starting addresses for a display frame are provided to the line address computer by a host processor. The line address computer determines the starting line addresses for subsequent lines by appropriately incrementing the line addresses of previous lines.

Abstract: Presented herein is a line address computer for providing chroma coefficients to a chroma filter. At each horizontal synchronization pulse, the line address computer provides a set of interpolation weights to a chroma filter. The chroma filter uses the provided set of weights to interpolate pixels in chroma pixel positions in a display format from chroma pixels in another format.

Abstract: Methods of setting numerically controlled delay lines using step sizes based on a delay locked loop lock value are presented herein. In one embodiment, a method may comprise, for example, one or more of the following: calculating an offset value for at least one NCDL; and interpolating a new offset value for the at least one NCDL, based on a change in a delay locked loop (DLL) output value from a previous DLL output value to a new DLL output value.