A method, system, and computer programs for improving the image quality of one or more predicted frames in a video image compression system, where each frame comprises a plurality of pixels. A picture region or macroblock of certain types of frames can be encoded by reference to one or more referenceable frames in some cases, and by reference to two or more referenceable frames in other cases. Such encoding may include interpolation, such as an unequal weighting. The DC value or AC pixel values of a picture region may be interpolated as well, with or without weighting. A code pattern of such frames having a variable number of bidirectional predicted frames can be dynamically determined. Frames can be transmitted from an encoder to a decoder in a delivery order different from a display order. Sharpening and/or softening filters can be applied to a picture region of certain frames during motion vector compensated prediction.