Acting in and adapting to a dynamically changing environment necessitates to precisely encode the timing of sensory events, and to time our own (re-)actions to them. Cerebellar (CE) and basal ganglia (BG) circuitries play fundamental and complementary roles in timing processes. While the CE seems to use precise timing (when an event occurs) and temporal intervals to generate temporal predictions (when a next event occurs), the BG uses relative timing to extract the beat in rhythmic sequences. As it is generally difficult to record data from respective patient groups in parallel, CE and BG contributions to timing processes are rarely investigated in combination. Here, we let healthy controls and patients with CE or BG lesions listen to isochronous auditory sequences while their EEG was recorded. We assessed intra- and inter-individual variabilities, as well as group differences, using event-related potentials (ERP), delta-band inter-trial phase-coherence, and acceleration dynamics while tuning to the stimulation frequency (Sf). CE and BG lesions increased variability in ERP latency and reduced the coherence of delta-band activity. CE but not BG lesions further impacted the stability of delta-band oscillations while tuning to the Sf . These findings show a causal link between subcortical lesions and the capacity to encode and synchronize ongoing neural activity with temporal regularities in the acoustic environment. While most standard metrics of neural entrainment do not dissociate specific contributions of BG and CE to sound processing in isochronous sequences, the newly introduced ‘stability’ metric isolated distinct changes in delta-band tuning dynamics in CE patients. This observation highlights the fundamental role of the CE in generating and maintaining stable neural representations of event onsets in the sensory environment.

Dynamic tactile perception and discrimination of textures require the ability to encode and differentiate complex vibration patterns elicited at the level of the skin when sliding against a surface. Whether the primary somatosensory cortex (S1) can encode the fine-grained spectrotemporal features distinguishing textures remains debated. To address this question, electroencephalography (EEG) frequency-tagging approach was used to characterize responses to vibrotactile oddball contrasts between two textures. In a first session designed to identify the topographical distribution of responses originating from the hand and foot representations in S1, standard and deviant stimuli were pure sinusoidal vibrations differing in frequency and intensity. In a second session, standard and deviant stimuli were two different snippets of bandpass-filtered white noise matched in terms of intensity and average frequency content, but differing in terms of their complex spectrotemporal content. Using the S1 functional localizer, we showed that oddball responses to a spectrotemporal contrast follow the somatotopical organization of S1. Our results suggest that the encoding of fine-grained spectrotemporal features associated with different vibration patterns involves S1.