One of the major science goals of the Visible Emission Line Coronagraph (VELC) payload aboard the Aditya-L1 mission is to map the coronal magnetic field topology and quantitative estimation of longitudinal magnetic field on a routine basis. The infrared channel of VELC is equipped with a polarimeter to carry out full Stokes spectropolarimetric observations in the Fe xiii line at 1074.7 nm. The polarimeter is in a dual-beam setup with a continuously rotating wave plate as the polarization modulator. Detection of circular polarization due to the Zeeman effect and depolarization of linear polarization in the presence of a magnetic field due to the saturated Hanle effect in the Fe xiii line require a high signal-to-noise ratio (SNR). Due to the limited number of photons, long integration times are expected to build the required SNR. In other words, signals from a large number of modulation cycles are to be averaged to achieve the required SNR. This poses several difficulties. One is the increase in data volume and the other is the change in the modulation matrix in successive modulation cycles. The latter effect arises due to a mismatch between the retarder's rotation period and the length of the signal detection time in the case of the VELC spectropolarimeter. It is shown in this paper that by appropriately choosing the number of samples per half rotation, the data volume can be optimized. A potential solution is suggested to account for modulation matrix variation from one cycle to another.