We have developed a Brain-Computer Interface (BCI) experience combining electroencephalogram (EEG) signals with Virtual Reality (VR). In our proposal, the users can attend a VR live concert system while participating in it throughout the generation of visual effects (VE) based on their brain activity. The VE are generated according to the beta and alpha waves of the EEG signal. When the beta-alpha ratio exceeds specific values, a threshold triggers the generation of different VE. A comparison between fixed threshold values and personalized thresholds for each subject is conducted. When conducting VR lives, individual differences on the beta-alpha ratio sometimes prevent the generation of VE based on constant thresholds, causing bigger differences on the experience between users. Thus, towards conducting VR live concert with many users, automatic adjustment of the thresholds for compensating the individual differences is required. In this paper, we proposed a method for adjusting thresholds based on percentiles of the distribution of beta-alpha ratio values and implemented it to our VR live system. We show results of experiments for validating the proposed methods.