In this paper we propose a fast procedure--stimulation paradigm, data analysis, data presentation--that permits the study of frequency selective enhancements or suppressions of the cortical responses in selected frequency windows. These responses can be used to classify the underlying activity. This novel procedure combines, in one single test, multichannel detection, finely frequency scanned stimulation, a paradigm that permits to study the effect of the stimulation on the underlying activity (during the non-stimulated periods intermingled with the burst of stimuli) and a compact presentation to appreciate the whole rate/frequency/topographical dependency of the spectral responses. In particular we applied this procedure to non-invasively study for the first time the frequency responsiveness within the extended alpha band (6 Hz to 14 Hz) under separate visual and somatosensory stimulation. The test proved robust and the responses quite stationary. We discriminated the lower and higher alpha band through their different topographical and frequency responsiveness features and suggested them to be the same sub-bands recently discriminated via cognitive experiments on the basis of their functional correlates. We measured a rate dependent alpha suppression (both in the somatosensory and visual modality) that is suggested to be linked to the decoding of the stimulus rate. A slight decrease in the frequency of the spontaneous activity following stimulations at any rate is suggested to be connected to attentional load. We stressed the importance of our simple sensorial discrimination of those activities already evidenced on the basis of their functional correlates, as well as the possible clinical uses of the test: on epilepsy, Alzheimer's disease (recently described to have effects on the responsiveness of the cortex to sensory stimuli at different frequencies) as well as other disabling pathologies.