Cochlear Implant Atlas
CI Atlas · Hearing Music Through an Implant · Module 07

7Many Notes at Once: Harmony and Polyphony

A single sung line can be followed; a chord, a choir, or a band tends to collapse into a single rough texture. With only a handful of broad frequency channels, the implant cannot pull simultaneous pitches apart, so harmony blurs and the melody hides inside the mix.

TWhy simultaneous pitches collapse

Resolving two notes sounding together requires placing their harmonics on separate, well-defined channels; an implant offers only a dozen or so broad, overlapping channels, so concurrent pitches land on the same electrodes and merge. When several notes overlap, the device transmits the combined envelope on each channel rather than the individual partials, so the listener hears one fused, roughened sound instead of distinct simultaneous pitches. Place-pitch cues become ambiguous with simultaneous tones because current spread from neighbouring electrodes overlaps, and the temporal-envelope pitch cue that survives is weak and easily masked. The result is that intervals and chords that a normal-hearing listener perceives as consonant or dissonant are largely indistinguishable through the implant.[2008][2008][2009]

Two notes at once: a normal cochlea vs an implant

cochlear place (low → high pitch)normal: two sharp peaksimplant: blurredimplant: notes merge into one blur

A normal cochlea has about 3500 inner hair cells driving roughly 30000 nerve fibres, so two notes a small interval apart sit on two narrow, separate peaks (green). A cochlear implant has only 12–22 electrodes and, after current spread, about 8 effective channels; each note becomes a broad hill (red) and nearby notes fuse. Only when the notes are pushed far apart do the implant peaks separate — which is why chords and harmony are so hard to hear electrically. Schematic.

TThe musical cocktail-party problem

Segregating a melody line from its accompaniment, or a singer from a backing band, is a musical version of the cocktail-party problem: the listener must group some partials into a foreground stream and reject the rest. Normal hearing uses fine spectral detail, harmonicity and pitch differences to form these streams; the implant delivers little of this information, so foreground and background fuse. Lyrics carry linguistic cues the brain can latch onto, which is why a sung melody with words is often easier to follow than the same melody played on an instrument inside an ensemble. Adding an accompaniment typically lowers melody-recognition scores sharply compared with the same melody presented alone.[2009][2008][2005]

Pulling a melody out of its accompaniment

melody (foreground)accompanimentbeats →
Melody recognised31%

Presented on its own, a melody is recognised by implant users around 78% of the time. Add an accompaniment and recognition falls to roughly 31%, because the implant cannot deliver the spectral differences a normal-hearing brain uses to segregatethe tune from the chords. Foreground and background fuse into one stream — the melody is buried rather than blended. Illustrative.

What music survives the implant: easiest (bottom) to hardest (top)

harder ↑Solo drum / pure rhythm90%Single unaccompanied voice68%Simple song (voice + light backing)48%Small ensemble / pop track32%Dense orchestral / choral14%

Solo drum / pure rhythmCarried entirely by timing, which electric hearing preserves well — the most accessible music of all.

The ladder follows what the implant keeps and what it loses. Rhythmis preserved, so a solo drum sits at the bottom as the most accessible music; a single voice and a simple song remain followable. As pitches and timbres pile up — small ensembles, then dense orchestral or choralworks — the few effective channels are overwhelmed and accessibility collapses. Schematic.

CWhat stays accessible: rhythm and single lines

Rhythm and tempo ride on the temporal envelope, which the implant transmits well, so beat, meter and rhythmic patterns are perceived almost normally and remain a reliable source of musical enjoyment. Simple, single-line (monophonic), rhythmically strong music is the most accessible: a solo voice, a single instrument, a strong drum line, or familiar songs with clear lyrics. Dense orchestral writing, choral harmony, jazz and harmonically rich pop are the least accessible because many pitches and timbres overlap at once. Practical counselling: encourage recipients to start with rhythmic, sparsely arranged, familiar pieces and to use lyrics and visual cues, rather than expecting to enjoy a full orchestra early on.[2008][2008]

Case 29.7 · Many Notes at Once
A 58-year-old post-lingually deafened CI user, six months after activation, reports that he can follow and enjoy a solo singer-songwriter recording and tap along to dance music, but a symphony 'sounds like noise' and he can no longer tell whether a chord is major or minor.

Which single factor best explains the pattern of preserved and lost musical abilities he describes?

Self-assessment — Module 73 questions
Question 1

Why are chords and polyphonic textures especially hard to perceive through a cochlear implant?

Question 2

Which type of music is generally MOST accessible to a cochlear implant user?

Question 3

Separating a singer's melody from a backing band through an implant is best described as:

Tracked locally in your browser — see /progress for the dashboard.