3The Electric Dynamic Range: A Few Decibels to Work With

FThreshold to comfort: a razor-thin window

In acoustic hearing the gap between the faintest detectable sound and the level that becomes uncomfortably loud is enormous, roughly 120 dB at the most sensitive frequencies, a ratio of about a trillion in intensity. In electric hearing that window collapses. Between the threshold of detection, which audiologists call the T level, and the maximum comfortable or M/C level, the behavioural range is often only about 10 to 20 dB, and on some electrodes just a few decibels.

Because the window is so small, loudness grows very steeply as current rises: a tiny increase in stimulation can move the percept from barely audible to too loud. This is the defining clinical fact of electric hearing and the reason an implant cannot simply pass the wide range of everyday sound straight to the electrodes.[1994][1983]

TThe right units: current and charge per phase

Acoustic stimuli are described in decibels of sound pressure, but an electrode delivers electric charge, so electric hearing is described in different units. The amplitude of stimulation is a current, measured in microamps, and because implants use brief biphasic pulses, what matters to the nerve is the charge delivered in each phase, the product of current and phase duration, measured in nanocoulombs. Threshold and comfort are most meaningfully expressed as charge per phase rather than current alone.

Charge is the relevant quantity because an excitable membrane integrates current over time. A short, strong pulse and a longer, weaker one can deliver the same charge and reach threshold similarly, within limits. This is also why the electrical dynamic range looks even more compressed when expressed in decibels of current: small absolute changes in microamps span the whole usable range.[1983][1996]

TPhase duration and rate trade against current

The threshold current is not fixed; it depends on how long each pulse lasts. The strength-duration relationship captures this: as phase duration shortens, more current is needed to reach threshold, approaching a steep rise at very short durations; as it lengthens, the required current falls toward a floor called the rheobase. The pulse duration at which threshold is twice the rheobase is the chronaxie, a few hundred microseconds for auditory nerve fibres. A processor can thus trade phase duration against current to reach a target loudness within compliance limits.

Pulse rate matters too. Higher stimulation rates allow temporal integration and facilitation across pulses, generally lowering the current needed for threshold and altering loudness growth, but they also consume more of the device’s power and time budget across channels. Designers juggle current, phase duration and rate together, all within the constraint that the electrode and battery can only supply so much charge before reaching voltage compliance.[1996][2016]

CWhy the range is compressed, and what it means for coding

Several factors conspire to compress the electrical range. Surviving neurons fire over only 7 to 10 dB individually, the spiral ganglion population is often depleted and shrunken, and broad current spread recruits fibres in a narrow band of levels. The result is a behavioural range a fraction of the acoustic one, with steep loudness growth on top.

The practical consequence drives the rest of the implant. A processor must take the 60 to 80 dB range of everyday sound and map it into a few decibels of electric range, so heavy compression, usually a logarithmic loudness-growth function between each electrode’s T and C levels, is mandatory rather than optional. Getting these T and C levels right for each electrode is the core of programming, and explains why even small mapping errors are so audible to the user. Everything downstream, from coding strategy to clinical fitting, is shaped by this narrow window.[1994][2004]