I need some information on fixed point arithmetic operation
like if we multiply two Q23.19 format numers what will be the resultant format and how do you reprsent it in 32bit etc
give some document on this topic...
hi, how can u represent any number in Q23.19, it should be Q31.0 to q0.31 format
ok any how
if u multiply two q23.19 number i think u get q46.38, now u have to shift right by 23 times to get back the same q23.19
Hi I think u r going to implement DCT in Fixed Point Arithmatic. For that u need to multiply ur floating point data (e.g. DCT Matrix ) with 2^23 and den round,Store this matrix in an array. Now on each multiplication round 22 bit and shift down by 23.I can send u Some Lecture slides if u want