hello,
i am using iqmath v213 for c64+ DSP and would like to use one of the following functions _IQNisqrt and/or _IQNdiv
I need to use specifc Q format since I've wide dynamic range on my input data
I do get the following vectors:
short int nomVec[SIZE]
unsigned int denomVec[SIZE]
the IQmath shall provide me the abiliy to compute;
float outVector[SIZE] = nomVec/sqrt(denomVec)
the denomVec vector is 32 bits unsigned with dynamic range from 0-2^32-1
when attempting to use _IQ0isqrt (or _IQ0div) for large numbers (say 2^32-1e6) the output results is saturated (i do get 0)
Same results i get to other Qs such as _IQ1isqrt, _IQ2sqrt etc. The sqrt functions on the contrary do not saturate, but when applying div it saturates again
for smaller Qs i do get good results
Can someone help on this issue?
thanks in advance
ilan