7412533520

Y CORTEX

RAIN)

%

Fig. 7. Some connections in th« atcending pathways o( the auditory nerroui system.

pinpoint. If the phase rclations for toncs 1500 cps and below, which finally reach the car by way of the speakers, are not at all those which the ears would hear if immersed in the original sound, there is no reason to expect to re-create the exact iliusion. The most one could expect would be a "diffuse-ness” to the sound so that it might be hard to say. if blindfolded, just where the sound source is in the room. And indeed this is about all that some Systems and recordings achieve. It is not neccssary, however. to be satisfied with this extcnsion of simple monaural prin-ciples.

To the auditory scientist, of course, the compelling ąuestion is not how one shall create the stereophonic iliusion but how is it possible that the binaural auditory system can be so extremely sensitive to temporal and intensive pat-terns. After all, the speed of the nerv-ous impulse and delays in neurone-to-neurone connections are on the order of milliseconds—how then can the ears be sensitive to difTerences on the order of microseconds?

Auditory Nervous System

The answer to this question seems to lie in the layout of the auditory nervous system. (See Fig. 7.) The in-ner ear, though coiled, can be con-sidered unrolled as a piano harp with a regular progression from the high fre-ąuencies to the Iow. The hair cells in this progression attach to and "trig-ger" the individual auditory nerve fibers. This impulse travels by several ‘‘relay stages” to the cortex, or new brain. Unlike a telephone wire through a switchboard, these fibers do not go straight to the cortex. They emerge from the cochlea and go a few milli-meters to the brain stem where each nerve fiber splits and sends one fibril to the dorsal and one to the ventral cochlear nucleus. At this point some of the fibers terminate on another fiber (the juncture, callcd a "synapsę.” in-troduces a time delay) whereupon that fiber goes up the brain stem. On the other hand. some of the fibers from the cochlea continue straight through the dorsal and ventral nuclei with no synaptic delay and proceed up the brain stem.

Notę that we are by no means up to the cortex at this point. It is very important to undcrstand that some fibers at the level of the first nucleus cross over to the other side so that even at the first "relay” station we have the possibility of inter-aural effects.

Now the crossed and uncrossed fibers, some synaptically delayed and some not. go up the brain stem to another left-right pair of nuclei callcd the “colliculi.” Herc, again. in each colli-culus we have three possibilities. A "relay” can cross from the other side, or can go to the other side, and the colliculus can be bypassed by a fiber going straight through to the genic-ulate body in the mid-brain, just below the cortex.

What we have here is an ideał system for a comparison of stimuli impinging on the two ears. The oppor-tunity for comparing the inputs to the two ears is good—so much so that the neural representation from the left ear is actually greater on the right side of the nervous system. There are not just a few fibers which cross. Further-more the system is not analogous to the visual nervous system where 50 per-cent of the fibers also cross. There, the retina is divided into left and right halves, the fibers from the nasal half of each eye Crossing in the optic chias-ma—the fibers from the temporal half going back, uncrossed. to the visual cortex. But this is a simple and un-complicated layout compared to the "relay” systems, "delay lines,” and "switchboard” opportunities in hear-ing.

For the mathematically minded, the ladder system as we see it here allows for two types of correlation. First a running auto-correlation of one ear on itself. Built-in "delay lines” at the synapses and arising from different fiber lengths make this possible—from a single half-millisecond click we may have nerve impulses reaching a "switchboard” over a period of 5. 10. or 20 milliseconds. Thus a multiple look can be had, within each of several successive "switchboards.” at a single monaural input.

How much morę, then, is the oppor-tunity laid down for a second type of correlation—a running correlation be-tween the two ears. At every "station" from the brain stem to the cortex a multiplex and indeed superb look can be had of the two inputs simultane-ously. It is in the course of these repeated observations that the nervous system shapes and refines the binaural sensation.

Opening a "second ear" to the world results in a dramatic sharpening of auditory experience and in worthwhile improvement in orientation and nat-uralness. We have scen how this is madę possible on the basis of time and intensity cues, with a nervous system dcsigned, as if by a superb architect, for the fuli utilization of those cues. Futurę electro-acoustic systems will undoubtedly exploit to the fuli these capabilities of our binaural sense.

-w-

40

ELECTRONICS WORLD