Motion prediction in video coding

An apparatus comprising:

• at least one processor and at least one memory including computer program code, the at least one memory and computer program code configured to, with the processor, cause the apparatus to:
• determine a block of pixels of a video representation encoded in a bitstream, values of said pixels having a first precision, wherein the first precision indicates the number of bits needed to represent values of said pixels;
• determine a type of the block;
• wherein if the determining indicates that the block is a block predicted by using two or more reference blocks, the reference blocks having said first precision:
• determine a first reference pixel location in a first reference block and a second reference pixel location in a second reference block;
• use said first reference pixel location to obtain a first prediction by a fractional pixel sample interpolation process, said first prediction having a second precision, which is higher than said first precision;
• use said second reference pixel location to obtain a second prediction, said second prediction having the second precision, which is higher than said first precision, wherein the second precision indicates the number of bits needed to represent values of said first prediction and values of said second prediction;
• add said first prediction having the second precision and said second prediction having the second precision and a rounding value to obtain a combined prediction;
• and
• after adding the first prediction and the second prediction and the rounding value, decrease the precision of said combined prediction to said first precision by shifting bits of the combined prediction to the right.

A computer program product comprising at least one non-transitory computer readable storage medium having, computer executable program code portions stored therein, the computer executable program code portions comprising program code instructions configured to:

• determine a block of pixels of a video representation encoded in a bitstream, values of said pixels having a first precision, wherein the first precision indicates the number of bits needed to represent values of said pixels;
• determine a type of the block;
• if the determining indicates that the block is a block predicted by using two or more reference blocks, the reference blocks having said first precision:
• determine a first reference pixel location in a first reference block and a second reference pixel location in a second reference block;
• use said first reference pixel location to obtain a first prediction by a fractional pixel sample interpolation process, said first prediction having a second precision, which is higher than said first precision;
• use said second reference pixel location to obtain a second prediction, said second prediction having the second precision, which is higher than said first precision, wherein the second precision indicates the number of bits needed to represent values of said first prediction and values of said second prediction:
• add said first prediction having the second precision and said second prediction having the second precision and a rounding value to obtain a combined prediction;
• and
• after adding the first prediction and the second prediction and the rounding value, decrease the precision of said combined prediction to said first precision by shifting bits of the combined prediction to the right.

An apparatus comprising:

• an input to determine a block of pixels of a video representation encoded in a bitstream, values of said pixels having a first precision, wherein the first precision indicates the number of bits needed to represent values of said pixels;
• a determinator to determine a type of the block;
• wherein if the determining indicates that the block is a block predicted by using two or more reference blocks with the reference blocks having said first precision, said determinator further to determine a first reference pixel location in a first reference block and a second reference pixel location in a second reference block a first predictor to use said first reference pixel location to obtain a first prediction by a fractional pixel sample interpolation process, said first prediction having a second precision, which is higher than said first precision;
• a second predictor to use said second reference pixel location to obtain a second prediction, said second prediction having the second precision, which is higher than said first precision, wherein the second precision indicates the number of bits needed to represent values of said first prediction and values of said second prediction;
• a combiner to add said first prediction having the second precision and said second prediction having the second precision and a rounding value to obtain a combined prediction;
• and
• after adding the first prediction and the second prediction and the rounding value, a shifter to decrease the precision of said combined prediction to said first precision by shifting bits of the combined prediction to the right.

An apparatus comprising;

• means for determining a block of pixels of a video representation encoded in a bitstream, values of said pixels having a first precision, wherein the first precision indicates the number of bits needed to represent values of said pixels;
• means for determining a type of the block;
• means for determining a first reference pixel location in a first reference block and a second reference pixel location in a second reference block, if the determining indicates that the block is a block predicted by using two or more reference blocks with the reference blocks having said first precision;
• means for using said first reference pixel location to obtain a first prediction by a fractional pixel sample interpolation process, said first prediction having a second precision, which is higher than said first precision;
• means for using said second reference pixel location to obtain a second prediction, said second prediction having the second precision, which is higher than said first precision, wherein the second precision indicates the number of bits needed to represent values of said first prediction and values of said second prediction;
• means for adding said first prediction having the second precision and said second prediction having the second precision and a rounding value to obtain a combined prediction;
• and
• after adding the first prediction and the second prediction and the rounding value, means for decreasing the precision of said combined prediction to said first precision by shifting bits of the combined prediction to the right.