1Why Music Is the Implant's Hardest Sound

FThe speech-music asymmetry

The cochlear implant was engineered to restore speech, and in a well-selected recipient it does so remarkably well, often supporting conversation over the telephone with no lip-reading. Music is the everyday sound the same device handles worst, so a recipient who follows a phone call may still fail to name a tune they have known since childhood. This asymmetry is the organising fact of the chapter: the implant gives back the spoken word far more faithfully than it gives back the song. Recipients consistently report that music after implantation sounds thin, harsh or unmusical, and most listen to it far less than they did before losing their hearing.[2014][2009][2004]

TSpeech leans on cues the implant keeps

Speech intelligibility rides largely on the slowly varying amplitude envelope of the signal, exactly the cue an envelope-based implant is designed to extract and transmit. A classic demonstration showed that speech remains highly intelligible when divided into just a few frequency bands carrying only envelope information, proving how little fine spectral detail speech actually needs. Speech is also rich in redundancy and linguistic context: a listener who mishears a phoneme can recover it from surrounding words, grammar and meaning. Because the cues speech depends on are the ones the implant preserves, the device can reconstruct an intelligible message from a coarse signal.[1995][2004]

TMusic leans on cues the implant discards

Music is carried by precise pitch and detailed spectral structure rather than by the slow envelope, so it depends on exactly the information electric hearing represents most poorly. Music is frequently polyphonic, with several instruments and voices sounding at once, demanding the spectral separation that a handful of broad implant channels cannot provide. Music offers no linguistic redundancy: a wrong note has no grammar or vocabulary to fall back on, so degraded pitch information cannot be guessed back into place. The result is a predictable split that frames the rest of the chapter: rhythm and tempo survive the implant well, while pitch, melody and timbre are largely lost.[2014][2004][2004]