The b-value, pivotal for analyzing acoustic emissions (AE) and seismic events, is often underestimated without considering amplitude spatial attenuation. This study presents a refined b-value determination methodology that integrates initial acquisition amplitude and accounts for geometrical and physical attenuation components. Our numerical simulations demonstrate that neglecting attenuation can lead to substantial inaccuracies in b-value estimations. To address this issue, we conducted uniaxial compression tests on various rocks with AE monitoring to capture subsurface damage. The results indicate that by correcting for amplitude decay, we obtain a more precise depiction of b-value evolution. This adjusted b-value, termed bgp, correlates more closely with the rock’s loading stages and fracturing processes than the traditional b-value, bt. For instance, granodiorite shows minimal b-value fluctuations due to fewer, larger-scale fractures during the initial compaction stage. The b-value fluctuates upwards in the elastic stage and early fracturing stage and maintains serrated fluctuations during the pre-peak softening stage. Upon further expansion of large fractures, the b-value in the peak softening stage exhibits a downward trend, typically reaching its nadir before the stress drop. A continuous decline in b-value following a serrated fluctuation at a high level signals the rock’s transition to a critical failure state. We also observed that b-value curves vary due to stress concentrations arising from specific structural features or end-face friction in rock samples, which closely track fracturing dynamics. This study underscores the necessity of amplitude attenuation correction for establishing reliable b-value evolution characteristics in seismic events across various scales.