Electronic ears

A few years ago I had my first mobile phone. Unlike home, he often had to communicate in noisy places. And in some cases, I heard the background even better than the interlocutor, or rather the interlocutor heard worse than the background. But if you put me in the place of his phone, I will hear him perfectly! And the point is clearly not in the quality of communication.

And the fact is that the human hearing aid is capable of separating sounds in range. Despite the fact that the whisper of the interlocutor is objectively quieter than a salute fluttering in the distance, you hear it perfectly, and moreover, a person is able to determine the approximate distance to any sound. Listen now to your surroundings - there are so many sounds that you have not noticed, and most importantly - you can guess from what distance they sound.

I do not know exactly how this trick is implemented in a person, but I want to talk about how you can make such smart ears for a telephone or other listening device.

So, the task: how to make a device that hears only the nearest sounds and is deaf to the background noise?
')
For definiteness, I will call the background noise the background, and the closest sound we need is the voice.

How does the background differ from the voice? In general, only the volume and distance. Volume, we obviously do not consider. The distance remains.

How does the listener find out which sound is next and which is closer? What is the difference? Hmm ... at first glance nothing more. But still I found a difference. Sound, like other waves, fades in proportion to the square of the distance. That is, the further away from the sound source, the less noticeable the drop in volume through each step.

That is, in order to find out the distance to the sound, we must measure its volume at two points, at different distances from it. The greater the difference in volume, the closer the sound.

I want to put two microphones - one near the sound source (for example, as a microphone in the phone), and the second - a little further (for example, on the back of the phone). And take the final signal simply as the difference between the two microphones. Thus, the background noise will be taken away from itself, since it is equally audible in both microphones, and a clear voice will remain.

I do not see fundamental glitches in this method, theoretically it should work. Appeal to knowledgeable people - I wrote nonsense? If not, is this approach used somewhere?