Disclosed herein are systems and methods that help to optimize the sound quality produced when a sound processing mode is enabled in an audio device, such as a hearing prosthesis. When the sound processing mode is enabled, a sound processor in the audio device will create a candidate transformed signal based on a default set of parameters associated with the sound processing mode (i.e. a candidate signal processing function). Further, the sound processor will create a reference transformed signal based on the same enabled sound processing mode, but with a varied set of parameters (i.e. a modified signal processing function). The sound processor will then compare the candidate transformed signal to the reference transformed signal to determine which signal has a higher quality. The device then creates an output based on the higher quality signal.