NOMN

NOMN EAR is a search for aural phenomena that have not yet been described. Its instrument is a synthetic ear assembled from models that auditory researchers have published and shared since the 1980s and from our own early attempts at new models. These cover the mechanics of the cochlea, the auditory nerve and the comparison of signals between the two ears, and they include networks trained by machine learning to perform listening tasks.

As humans we listen with an organ that remains under-theorized, with many unknowns. While there has been a tremendous amount of scholarship on “listening” in recent years, there are still few new listening objects. The nonlinear mechanics of the ear and the neural comparisons in the brainstem leave a sound with many perceptual homes.

A long list of distortion products and psychoacoustic effects has accumulated since Tartini described his third tone in the eighteenth century. We are looking for what lies outside this list, along with the synthesis methods needed to produce new listening objects.

A model hears as a microphone hears. Its response is a prediction, and listening remains inseparable from memory and the imaginary, so in the end, this research relies on what we can perceive and make plastic for creative use. We are running about 10,000 simulations per day and hope to return results and their associated tools in late 2027.