A sound arriving off-centre reaches the nearer ear first. This interaural time difference (ITD) is the brain's primary cue for locating low-frequency sound in the horizontal plane. For a head of radius a and sound speed c = 343 m/s, the Woodworth–Kuhn diffraction formula gives:
ITD(θ) = (a/c) · (θ + sin θ) [θ in radians]
The Jeffress model (1948) proposes that the medial superior olive (MSO) decodes θ with an array of coincidence-detector neurons fed by two axonal delay lines — one per ear — of opposite, systematically increasing conduction delay. A neuron fires maximally only when its own built-in delay exactly cancels the ITD, so spikes from both ears arrive together:
left-ear pulse: x_L(t) = -X + v·(t - t_left)
right-ear pulse: x_R(t) = X - v·(t - t_right)
coincidence at: x_c = -(v/2)·ITD
Each frame the source emits a synchronized click at both ears with the correct ITD; two spikes race along the delay lines drawn above and below the neuron row, and the neuron nearest their meeting point flashes. Averaged over many clicks this builds a population "tuning curve" (the bars) whose peak — a specific place in the neural array — is the brain's estimate of direction: a genuine place code for space built entirely from timing.
- Azimuth slider — moves the sound source; recomputes the true ITD via Woodworth's formula.
- Click rate — how often the source fires a synchronized click pair (like a train of transient clicks used in real ITD-tuning experiments).
- Head radius — a bigger head means a larger maximum ITD (~± 690 µs at 8.75 cm), stretching the map across the same 21 neurons and changing spatial resolution.
- Auto-sweep — sweeps the source back and forth so you can watch the coincidence point — and the decoded azimuth — track it live.