Abstract: A method for performing pixel addressing operations for a tiled image. The method maintains separate X and Y components of a pixel address in a tiled image. Each component includes a tile and pixel index spaced apart by blank bit fields such that when combined, the X and Y components form a pixel address that is directly dereferenced to read or write a pixel in the tiled image. To decrement or increment the X or Y component, the method introduces a mask such that any carry bit can be moved across the blank bit field between the tile and pixel indices. This enables the method to use the same efficient add/subtract and mask operations whether or not a pattern of pixel accesses crosses a tile boundary. The X and Y components are represented as virtual X and Y coordinates in a two-dimensional array of pixels to hide the details of the add/subtract and mask operations from the programmer.