Variation of mass supply rate from the companion can be smeared out by viscous processes inside an accretion disk. Hence, by the time the flow reaches the inner edge, the variation in X-rays need not reflect the true variation of the mass supply rate at the outer edge. However, if the viscosity fluctuates around a mean value, one would expect the viscous time scale t visc also to spread around a mean value. In high mass X-ray binaries, which are thought to be primarily windfed, the size of the viscous Keplerian disk is smaller and thus such a spread could be lower as compared to the low mass X-ray binaries which are primarily fed by Roche lobe overflow. If there is an increasing or decreasing trend in viscosity, the interval between enhanced emission would be modified systematically. In the absence of a detailed knowledge about the variation of mass supply rates at the outer edge, we study ideal circumstances where modulation must take place exactly in orbital time scales, such as when there is an ellipticity in the orbit. We study a few compact binaries using long term All Sky monitor (ASM) data (1.5-12 keV) of Rossi X-ray Timing Explorer (RXTE) and all sky survey data (15-50 keV) of Swift satellites by different methods to look for such smearing effects and to infer what these results can tell us about the viscous processes inside the respective disks. We employ three different methods to seek imprints of periodicity on the X-ray variation and found that in all the cases, the location of the peak in the power density spectra is consistent with the orbital frequencies. Interestingly, in high mass X-ray binaries the peaks are sharp with high rms values, consistent with a small Keplerian disk in a wind fed system. However, in low mass X-ray binaries with larger Keplerian disk component, the peaks are spreaded out with much lower rms values. X-ray reflections, or superhump phenomena which may also cause such X-ray modulations would not be affected by the size of the Keplerian disk component. Our result thus confirms different sizes of Keplerian disks in these two important classes of binaries. If the orbital period of any binary system is not known, it may be obtained with reasonable accuracy for HMXBs and with lesser accuracy for LMXBs by our method.