Brain chemistry is more important than EEG studies in cerebral mechanism research for endogenous nature. Following brain science study paradigm in self stimulus condition of calm, smile, truth and lie, this paper grasps brain chemistry complexity through neurotransmitters’ super-slow spectrums and their allometric power law features. First, one practical microsystem is constructed based on ARM7-CPU embedded into db4 wavelet analysis in tendem of multy-fractal extracting. Then five subjects are invited and tested in processing near infrared 940 nm otopoint-signals. Nonlinear power law model was extracted in form of Ps = Pc∧a with base-line of calm state, wherein a1 fitted with wavelet analysis signature, and a2 with wavelet analysis in tendem of multy-fractal feature. The interesting results include that a2 means are decreasing basicly in self stimulus condition of smile, truth and lie versus calm. The preliminary conclusion suggests that automatic scorings are feasible for emotion and cognition such as smile, truth and lie versus base-line of calm, in reason of more details mapping brain neurotransmitters and fractal synaptics in vitro.