2The Cues the Brain Uses: ITD, ILD and Spatial Hearing

TTwo cues, two frequency ranges: the duplex theory

A sound off to one side reaches the nearer ear sooner and louder; the brain reads these as the interaural time difference (ITD) and the interaural level difference (ILD). Lord Rayleigh's duplex theory holds that low frequencies are localized mainly by ITD and high frequencies mainly by ILD, because the two cues are physically reliable in different bands. The maximum natural ITD for an adult human head is only about 680 microseconds, corresponding to a sound arriving from one side, with about 50 microseconds equal to roughly 5 degrees off the midline. Wightman and Kistler showed that when timing and level cues conflict, listeners follow the low-frequency timing cue, confirming the dominance of ITD for broadband sounds that contain low frequencies.[1992][2020]

TWhy ITD owns the lows and ILD owns the highs

Low-frequency sounds have wavelengths far larger than the head, so they bend around it and arrive at both ears at nearly the same level; their useful cue is the phase, or timing, difference. A 500 Hz tone has a wavelength near 68 cm, dwarfing the ~17.5 cm head, which is why level differences are nearly useless below about 1500 Hz. High-frequency sounds are shorter than the head and are reflected by it, casting an acoustic shadow that can make the far ear up to about 10-20 dB quieter, a strong level cue. Above roughly 1500 Hz the rapid waveform cycles so fast that ongoing timing becomes ambiguous, so the brain switches to the head-shadow level cue, exactly as the duplex theory predicts.[2020][1992]

TWhere the brain fuses the two ears

Binaural comparison happens early, in the superior olivary complex of the brainstem, before either ear's signal reaches consciousness. The medial superior olive (MSO) is exquisitely sensitive to interaural time differences and is fed by precisely timed, low-frequency input from both cochleae. The lateral superior olive specializes in interaural level differences, comparing the loudness of high-frequency input across the two sides. Because this circuitry needs accurate, synchronized timing from both ears, it sets the bar that any prosthetic system must meet to deliver true binaural fusion.[2009][2020]

CWhat survives in electric hearing

Cochlear implants relay interaural level differences reasonably well, because loudness maps onto stimulation amplitude and the head shadow remains a physical, acoustic effect at the microphones. Implants convey interaural timing poorly: each processor runs independently with its own clock, envelope-based strategies discard the fine temporal structure, and there is no cross-device synchronization of pulse timing. Most implant users cannot follow temporal detail above a few hundred hertz electrically, precisely the low-frequency timing the MSO needs for ITD, so the dominant natural localization cue is largely lost. The practical result is the asymmetry seen throughout this chapter: bilateral implant users get level-based and head-shadow benefits readily but struggle to obtain the timing-based squelch and fine localization that ITD would provide.[2003][2020]