This chapter deals with the signals produced by particles that are being absorbed in a calorimeter. The calorimeter response is defined as the average signal produced per unit energy deposited in this absorption process, for example in terms of picoCoulombs per GeV. Defined in this way, a linear calorimeter has a constant response. Typically, the response of the calorimeter depends on the type of particle absorbed in it. Also, most calorimeters are non-linear for hadronic shower detection. This is the essence of the so-called non-compensation problem, which has in practice major consequences for the performance of calorimeters. The origins of this problem, and its possible solutions are described. The roles of the sampling fraction, the sampling frequency, the signal integration time and the choice of the absorber and active materials are examined in detail. Important parameters, such as the e/mip and e/h values, are defined and methods to determine their value are described.