Abstract: How can performers communicate emotions to listeners? What means do they use to accomplish this task? These are issues of paramount importance to musical behavior. The article introduces the Lens Model, a theoretical framework to illustrate how performers use a number of probabilistic (uncertain) but partly redundant cues (e.g. tempo, sound level, articulation) to express certain emotions, and how listeners use these cues to arrive at reliable judgements of the expression. Lens Model Equation (LME) is also introduced, by which regression models of performers and listeners can be mathematically related to each other. Based on the Lens Model and LME, a useful method of improving emotional communication called cognitive feedback (CFB) is put forward, that is, giving performers a chance to compare their own cue utilization to an optimal model for emotional communication on the basis of listeners cue utilization. We assume that except for improving musical performers’ skill, we can also use this method to improve listeners’ level of enjoying music

Key words: the Lens Model, Lens Model Equation, cognitive feedback