In developing a model to simulate heart sounds, the effect of the listener needs to be taken into account by including in the model the performance of the human ear. It has been observed that graphs of heart sound waveforms does not correspond to what is heard through a stethoscope. Most modelling would expect to be based on a flat frequency response, but because the sounds are quiet the loudness response of the human ear needs to be considered. A comparison is made between heart sounds and drum sounds so that suitable frequency components can be mixed to produce the final model.