Soleimani, M., Piruz, I., Mann, J. and Hubral, P., 2009. Common-Reflection-Surface stack: accounting for conflicting dip situations by considering all possible dips. Journal of Seismic Exploration, 18: 271-288. The common-reflection-surface stack was originally introduced as a data-driven method to simulate a zero-offset section from 2D seismic reflection prestack data. The principle of the CRS stack is to sum along a surface of contributions from an entire segment of a reflector instead of a reflection point. The aim of the CRS stack is not only to provide a well-simulated zero-offset section but also to determine certain attributes of hypothetical wavefronts at the acquisition surface. The CRS stack is independent of explicit velocity information and only requires the near-surface velocity in case a geometrical interpretation of its stacking parameters is desired. An important aspect of the method is that the estimated parameters are kinematic wavefield attributes which provide significant information on subsurface structures. The pragmatic search strategy of the original CRS stack implementation consists of three one-parameter searches. This implementation determines only one optimum stacking operator for each ZO sample to be simulated. Consequently, conflicting dip situations are not taken into account but only the most prominent event contributes to a particular stack sample. In one of the efforts to overcome this problem, the strategy has been extended in order to take into account up to five conflicting dips at each sample. Here, we propose a strategy which explicitly considers all possible angles and, thus, accounts for all the conflicting dips that may exist at each zero-offset sample to be simulated. This new strategy offers some advantages, e.g., it improves the continuity of events, reflections as well as diffractions, in conflicting dip situations. It also generally emphasizes diffraction events in the stacked section so that we can call it common-diffraction-surface stack. This method has the drawback that it does not give any section of wavefield attributes or coherence as, e.g., required for CRS-based tomography. Here, we processed the Sigsbee 2A synthetic data and also a real land data set with the new method and observe enhanced diffraction events and resolved conflicting dip situations. For the real data set, the definition of faults after poststack migration significantly improves.