A conventional processing workflow favors only the specular reflections, reducing or removing other wavefield interactions. These specular reflections are unsuitable for directly imaging sharp corners, such as those in fault zones and pinch outs; therefore, diffractions are used instead in a technique known as diffraction imaging. Plane-wave destruction (PWD) is a well-established method for removing reflections and imaging diffractions. However, this method assumes a gently variable slope; therefore, it fails to remove energy in areas that do not follow this assumption such as curved interfaces. To remove the remnant energy in these areas and thus enhance the overall interpretability of the diffraction images, we have adopted a simple spatial-variable filter in the frequency-wavenumber f-k domain based on the calculated dip field used for PWD, applied post PWD. To demonstrate the method, we have examined this on a range of synthetic data, complex synthetic data, and real data. The created diffraction images have then been interpreted to evidence the benefit of diffraction imaging in seismic interpretation, helping to delineate pinch outs, faults, and rugose surfaces.