Music is the aesthetical arrangement of sounds over time. When musical tones are concatenated in rhythms new melodies arise, governed by metrical and tonal-harmonic structures. The metrical structure is fundamental for our perception of rhythm: it allows us to process the temporal events of music and dance. Our rhythmic cognition is based on two distinct processes (Fitch, 2013): beat extraction (the perception of periodic points over time) and meter induction (the hierarchical organization of isochronous beats into sequences of strong and weak patterns). The neural correlates of these timing mechanisms can be captured with electroencephalographic recordings, because the neural activity increases at the frequencies corresponding to the perceived beat and the induced meter (Nozaradan et al., 2011; Nozaradan, 2014). Our main research goal is to understand how the human brain processes beat and meter across modalities, and attest the role of attention and formal training in music on the perception of rhythms. Since several species have been found to effectively use timing mechanisms in both the auditory and the visual modalities (Repp, 2005; Repp and Su, 2013; Patel and Iversen, 2014; Ravignani et al., 2014; Merchant et al., 2015), our first study explores meter induction in vision. We asked musicians to project an internal metrical structure onto auditory and visual stimuli, and they succeed in mentally grouping the isochronous visual flashes. The second study focuses on the spatial dimension of sound. We observed that auditory spatial cues are reliable to build up a metrical structure when attention is placed on the auditory modality. After comparing the neural entrainment of musicians and non-musicians, we also observed stronger neural responses to beat and meter in musicians. This reinforces the idea that the neural processing of beat and meter is facilitated by the long periods of formal training in music. These results suggest that the processing of beat and meter is less restricted to the auditory modality than previously believed. In fact, the existence of amodal timing mechanisms, partially present in other primates, may have allowed our species to master music, dance, language, and other synchronized activities that require precise timing of actions across modalities. Finally, to shed light onto the evolutionary bases of beat perception, we compared a non-vocal learning species, rats, and a vocal learning one, humans. After training them to distinguish between isochronous and non-isochronous beats, they were tested with new stimuli and both species succeeded in discriminating the regular beats from the irregular ones. This suggests that the perceptual timing mechanisms underlying regularity detection are not limited to humans, but also shared with a non-vocal learning species like rats. To sum up, we found that meter induction can occur beyond the auditory modality and that isochrony discrimination is shared with other animals.