I am unable to understand that why we go for horizontal or vertical scan at
all while decoding MPEG-4 bitstream? I guess Zig-Zag scan is the best then
why not only ZIG-ZIg. Is there any more compression by horizontal or
vertical scan? According to mpeg-4 standard zig-zag scan is done only when
ac pred flag is set, Why?
This message was sent using the Comp.DSP web interface on
Reply by Correlious●July 16, 20052005-07-16
Depends on how the coefficents appear, somtimes your matrix doesn't
turn out w/ zeros increasing as you get closer to one corner (where
ZigZag encoding would be most efficient)!