The band sedimentation experiment in analytical ultracentrifugation (AUC) allows for the performance of a chemical reaction inside the AUC and also offers separation of individual pure components in a sedimentation velocity experiment. Although this experiment offers exciting possibilities for application, it is barely used. This is related to the bad definition of the initial conditions. Both the duration and the time of the solution overlay during rotor acceleration are not known. In this study, we investigate these conditions under the variation of the overlay volume using recording of interference patterns in a continuous mode during the acceleration of the rotor. It was found that the overlay occurs at rotor speeds between 770 and 2000 rpm, which is very low compared to typical experimental rotor speeds from 3 000 to 60 000 rpm and therefore elucidates that the generated reaction products, respectively, overlaid species are subject to the centrifugal force almost from the beginning. Also, the duration of the overlay is less than 1.2 s, which is very fast compared to hours of centrifugation time for an experiment and we demonstrated that the overlay compartment is completely emptied during overlay allowing for the precise calculation of the meniscus using the known sample sector geometry. Our results show that the initial conditions of the experiment are defined and should make an adapted analysis possible if the interdiffusion of the two solvents is taken into account, which lead to a dynamic density gradient.