A time interval reproduction task involving the parallel timing of two sensory signals presented either in the same modality or in different modalities was conducted. A memory component was tested by manipulating the delay separating the presentation of the target intervals and the moment when the reproduction began. Results show that there is more variance when only visually marked intervals are presented, and this effect is exacerbated with longer retention delays; with only one interval to process, encoding the interval with signals delivered from two modalities contribute to reduce variance. The results suggest the existence of modality-specific clock components and memory mechanisms.