From Figures 5 and 6 we can observe that, in general, results from angle gather are smoother and have a shape more similar to the theoretical tendency. In Figure 5, corresponding to model 1 (no AVA effect), the picked amplitudes are flatter in the angle gather case. In Figure 6, corresponding to model 2 ( decreasing amplitudes with increasing angles of incidence), the picked amplitudes follow the expected tendency more closely. The output in angle gather is a function of the reflection angle at the reflector, so no conversion is needed to do AVA analysis. Changing parameters (sampling, regularization, etc.) for the angle gather generation don't require re-migrating the data because migration and conversion to angle gather are separate processes. Results in the offset ray parameter domain could be improved if we use regularization in the inversion process Prucha and Biondi (2000); therefore, a next step should incorporate this approach.