Hi Arbaz,

I agree with you; That's why I said "average" in the calculation. Also we know about compression and about algotithms for getting the voice. But there are not miracles - without memory there are not good voices, even with sophisticated algorithms. And if we use most of the memory for only a few voices the others will be weaks.

As it is said in my country "without eggs there are no omelets"

Regards
Carlos Rodrigues