The effects of active and passive processing of interval events on prospective and retrospective time estimates