Assessment of the area of interest on tectonic overprint is a key aspect during various geoscientific exploration endeavors. Conventionally, 3D reflection imaging is used because it provides the highest resolution of the subsurface image. Tectonic features, e.g., faults or fractures, are imaged indirectly by means of discontinuities of the specular reflections. Specular reflections, however, are only a part of the backscattering wavefield. Geologic heterogeneities in the subsurface can act as scattering points or scattering edges, which both evoke diffracted waves. Thus, diffracted waves are a direct seismic response from subsurface heterogeneities, which have a size comparable with the prevailing wavelength or a curvature growing locally to infinity. We investigate a dedicated processing of diffracted waves, which allows direct imaging of such heterogeneities. The method is based on separation of reflections from diffractions with subsequent diffraction focusing. We apply a combination of frequency-wavenumber filtering and an adaptive-subtraction engine to isolate diffractions. Once the reflections are removed from a data set, diffracted waves are focused into their apexes yielding an image of scatter points, which can be interpreted as an indicator for the extent of the degree of inhomogeneities or can be used to map structural elements or fracture density. We use a modified coherence analysis as a focusing tool. Diffractions from the scattering points have a different phase response than diffractions from the edges. These differences can be distinguished and further guide the geologic interpretation.