Fast Fourier transform (FFT) computing may improved using two ways: the first way consist in packing two separate real functions into the complex input array in such a way that their individual transforms can be separated from the result; the second method used to improve the FFT computation consists in packing the real input of length N array cleverly, without extra zeros, into a complex array having half of its length. In this paper an analysis of a combination of these two methods is performed and also a numerical simulation in order to detect the right conditions in which this method is applicable.