Human mental tasks are influenced by some variables, such as the sound stimulation which is used for
emotional therapy by physicians. Although human perceptions toward sound stimulation are different
from each other, generally we can find or develop sound sources which can cause certain mental task. For
therapy evaluation, we need EEG as the measurement instrument and identification and classification
system that has developed in this research, based on power spectral and apriory knowledge about the EEG
signal. Alpha wave will appear at relax condition, theta wave when sleepy or at emotional stress, and beta
wave relates to non-relax condition. The EEG signal is evaluated toward 12 sound stimulations. Each
sound source was given to 3 subjects repeatedly. As a validation, we observed facial expression and
knowledge about the sound stimulation.
The research showed that power spectral analysis can be used as a mental task classification system
model. From 12 sound stimulations, about 41% gave sleepy, 36% relax, and 23% non-relax condition.
Sleepy condition was dominant at fast beat music and the end of measurement. The result gives that EEG
spectral of symmetric channels are asymmetric.