Network calculus (also known as real-time calculus) and event stream theory are two theories designed to compute upper bounds on response time for real-time systems. Both generalise the common periodic request arrival to any kind of workload using cumulative functions. In network calculus, a data flow is modelled by its cumulative curve, A(t), representing the amount of data sent by the flow up to time t, whereas the event stream theory counts the number of packets sent up to t, denoted by E(t). Of course, the size of packets is a link between both functions. This work presents a formal model embedding both A, E, the packet functions, and their basic relations.