Whaleblue
Southbound
...
It's premised that in this type of holistic listening we use patterns & statistical analysis. For instance, if listening to the sound of applause, it's a wash of sound we hear, not any particular element although some elements pop out into our awareness, at times. Now listen to the sound of rain on a tin roof & we can usually differentiate this sound from applause. Both consist of a mix of random frequencies with random timing & are changing all the time but they have different statistical frequency distributions.
It’s not, I’d suggest, primarily the different statistical frequency distributions, but the different overtones that allow us to distinguish such things.
Just as we can descern a violin from a piano if they play the exact same tune.