The paper considers recovery of signals from incomplete observations and a problem of determination of the allowed quantity of missed observations, i.e. the problem of determination of the size of the uniqueness sets for a given data recovery procedures. The paper suggests a way to bypass solution of this uniqueness problem via imposing restrictions investigates possibility of data recovery for classes of finite sequences under a special discretization of the process range. It is shown that these sequences can be dense in the space of all sequences and that the uniqueness sets for them can be singletons. Some robustness with respect to rounding of input data can be achieved via including additional observations.