In the present investigation, a theoretical approach and experimental models are discussed for estimating the experimental thermal time of an optical digital interferometry experiment. The investigation deals with coupling between heat and mass transfer phenomena. Mach-Zehnder Interferometry has been found to be an accurate and precise experimental tool for visualizing the thermodiffusion phenomenon inside a parallelepiped cavity, when a thermal gradient is applied to its two opposite sides. Processing the results of this experiment requires distinguishing the two experimental phases: 1) refractive index changes due to change of temperature and 2) refractive index changes due to the separation of the components. We can separate these two phases according to the thermal time of the liquid mixture. In previous studies L-2/chi (L = distance between hot and cold sides, chi - thermal diffusivity of the mixture) was used as the thermal time of the liquid. However, due to a high separation rate at the beginning of the thermodiffusion process, an accurate estimation of the thermal time can significantly affect the final result of the interferometry measurements. Here, a theoretical approach and an experimental method were developed to estimate the proper thermal time for this experiment. We also discuss the importance of precisely estimating the thermal time on the experimental results. This study shows that assuming thermal time equals L-2/chi may cause noticeable underestimation when measuring the Soret coefficients and maximum separation values.