We resolve a misunderstanding on the general observation of AE releases from mode-I and mixed-mode fracture toughness tests, and from compression tests on rock core or prismatic blocks. The fracture toughness tests expect few early AE while the compression test expects many. We investigate the detailed AE energy release organization from two novel approaches. From a system’s perspective, the former (mode-I and mixed-mode) undergoes a process of continuous entropy increase from the well-organized global condition regardless of having continuous energy input; the latter (rock core compression) displays a process of continuous entropy decrease until macroscale rock fracture formation. From perspective of elementary-damage-accumulation, the dynamic building process of the AE topology network from the mode-I or mixed-mode test displays a significantly stronger triggering effect compared to the rock core compression—especially in the early stages of these tests. Observations from both aspects consistently suggest the two types of rock fracture processes are of different physical processes. The analogue of the mode-I and mixed-mode fracture toughness tests is the pervasive process, while that of the rock core or prismatic block compression is the site percolation process. The pervasive source of the mode-I and mixed-mode tests provides a significantly lower energy barrier for the rock fracture formation than the rock core or prismatic block compression. Consequently, its early-stage AE releases will be suppressed or eliminated. Investigating the scaling features of the AE releases suggests the observational features for a notch-induced rock fracture process are universal across scale.