Challenge 3: Why a Recording's Own Noise Profile Beats a Generic One — Solution Walkthrough

The answer:

A noise profile is a fingerprint of exactly what the unwanted
background sound looks like in this specific recording — this
specific fan, this specific room's own hum, this specific
microphone's own faint hiss. Sampling it directly from a silent
stretch of the actual recording means the fingerprint matches the
actual noise that needs removing, sample for sample.

Why a generic, pre-made profile wouldn't match as well:

A different recording, made in a different room, with different
equipment, would have its own different background noise — a
different fan, a different hum frequency, a different hiss
character. A one-size-fits-all profile built from some other
recording's own noise wouldn't line up with this recording's actual
background sound, so noise reduction using it would remove less of
the real noise, more of the wanted signal, or both.

Why this ties back to Chapter 2:

This is exactly why Chapter 2 emphasized identifying a genuinely
silent stretch by eye on the waveform first — that specific,
recording-matched silence is what makes the profile sampled from it
actually effective.

WHY THIS WORKS AS AN ANSWER
------------------------------
This exercise checks that the effectiveness of noise reduction is
correctly traced back to the profile being sampled from the same
recording it's later applied to, not from an unrelated, generic
source.