Determining the Effective Time Filter 
The system determines the area in which a query displays data on a non-cumulative InfoCube as follows:
In general, the times are projected onto the most detailed time characteristic that occurs in the query.
The effective time filter is determined. The system calculates the set of all the time points from the static filter of the query. Global filters and structure elements that contain non-cumulative key figures are included here. The resulting set of time points from the static filter intersects the set of time points from the dynamic filter.
The validities are still taken into consideration. The query only displays non-cumulative values for time points, which are within the validity area for at least one cell in the query.
Separate validities are calculated for cells without selection and for cells with constant selection. A query can contain a maximum of two validity areas: one for cells without constant selection and one for cells with constant selection. The validity area for cells without constant is not influenced by the cells with constant selection. However, the validity area for cells with constant selection is influenced by the validity of cells without constant selection. This is due to two reasons:
The constant selection does not create objects. This means that constant selection cannot create new cells in the result. Therefore the validity of cells with constant selection cannot influence cells without constant selection.
Constant selection is often chosen to display a comparison value. Therefore the validity of cells without constant selection also applies to cells with constant selection.
More information: Constant Selection