Jeffress Model: Sound Localization by Coincidence Detection

How does a brain with no compass work out where a sound came from? For low-frequency sound the answer is timing: the ear closer to the source hears it first, by a few hundred microseconds. This simulator builds a real Jeffress coincidence-detector circuit — the classic model of the auditory brainstem's medial superior olive — in 3D. Move the sound source and the true interaural time difference is computed from head geometry using the Woodworth–Kuhn diffraction formula; a spike races along each of two opposing delay lines, and the neuron in the array where the two spikes meet lights up. Averaged over repeated clicks, the row of neurons builds a population tuning curve whose peak decodes the source's azimuth from nothing but two arrival times — the same "place code from timing" strategy real binaural neurons use to let you localize sound with your eyes closed.