Hi,
does anybody have an example of an efficient implementation of a Q.15 vector multiplication for the C55x DSP? I am thinking about something like the function "mul32" of DSPLIB, but that one is for Q31 vectors.
Or is there an efficient way to use the mul32 function for Q.15 vectors? In my understanding I would have to copy my Q.15 values into a "long" vector and change the content of the sign bit from bit #15 to bit #31. I cannot imagine that this is woth the effort.
Regards,
Raphael