“…It is obvious that nowadays, detailed seismic processing requires algorithms and techniques that are created in an innovative way using a combination of well-known seismic procedures and newly adapted techniques from different branches of science and other industries-like image processing (Zareba and Danek 2018a, b;Krasnov et al 2018;Li et al 2009), big data analysis (Gupta et al 2014;Rawat 2014), deep learning neural networks (Lewis and Vigh 2017), and the stock market (Bedi and Toshniwal 2018;Kyoung-jae 2003). There are plenty of established filters for seismic data pre-stack or post-stack filtering, enhancing, and analyzing based on a statistical approach (Wang et al 2016;Hoeber et al 2003). While these filters are useful, we noticed that in regions of complicated geology like the Carpathians (please note we will only refer to the Polish Carpathians including Central Carpathians, Outer Carpathians, and Carpathian Foredeep region in general), problematic residual noise occurs when techniques based solely on statistical approach are used, especially when they are directly applied to the parts of the seismic dataset where high-value dips or complicated fault systems are present.…”