In many industrial situations, bearing failure can lead to serious consequence on the overall process. A bearing's fault progressive character raises the question of finding the right moment to perform replacement at the cost of stopping the machine. The study done in this paper deals with mathematical modeling of the bearing's rolling element with a local defect on its fixed outer ring, based on a mass-spring-damper archetype system. A simulation of the vibratory behavior is performed, and its impact on ball-defect coincidences during shaft rotation under different working conditions is analyzed. The paper suggests applying advanced pre-processing techniques such as Singular Spectrum Analysis (SSA) and Envelope Analysis (EA) before extracting statistical indicators. Some wellknown time-domain indicators such as the Root Mean Square (RMS), kurtosis, and Energy around Ball Pass Frequency Outerring (EBPFO) are used on the raw and processed signals to highlight the defect evolution. The results carried out show that, applying the EA with SSA on the raw time-series signal at the pre-processing level, before statistical analysis can significantly improve the detection, thus an excellent diagnosis to incipient defects in bearings.