Synthetically modelled performances and commercial recordings of orchestral minuets were used in three rating tests and a synchronisation test to investigate two issues. The first one addresses the impression of authenticity and plausibility, i.e., the impression of a human-rendered performance, questioning whether smooth mathematical functions are adequate means of modelling timing and dynamics, or whether it is the deviations from smooth functions observed in human performances that contain hidden information contributing to the impression of authenticity and plausibility. The second question investigated whether the subjective impression of plausibility is also reflected in the anticipation of musical events, such as that required in a synchronisation task. In the plausibility tests, there was no significant difference between participants’ ratings of synthetic performances modelled with smooth functions and those including hidden information . In the synchronisation test, participants showed a higher standard deviation – proportionally to the mean beat length – in the tap-onset difference when tapping to the beat of a synthetic performance that had been rated implausible in the previous tests.