@nikki What I don't get is how can such a complex sound (with multiple instruments and vocals) be captured this way. It feels as though a wiggly groove should only be able to record little fluctuations in a single tone.
Our ears worry me for the same reason. It's just a vibrating membrane and 3 tiny bones, but their wiggles are enough to capture multiple different sounds, not just a simple 'is there sound here and how loud'.
How can that 2D waveform encode many different frequencies?
That would involve mister Fourier and his famous transform (in digital), or some physical aspect of length and stiffness of the small hairs we have floating in water inside the inner ear or something similar (which can still be modeled like a band-pass filter of some sort), each different length/stiffness selects a different frequency.
Short answer: yes.
Long one: I'm a computer scientist, not a biologist, so I'm not super-sure about details of the inner ear, the two functions might be in different parts of the ear with different specialization of those cells, but… basically yes, I'd say.