Phase I: To devise and evaluate a processing scheme that can be trained by the individual user to select automatically the amplification characteristics that the user would prefer in each listening situation. 

Phase II: To optimise the trainable algorithm.