NF ACOUS - NM20 - Blue

How the match is computed: each measured curve is interpolated linearly in log-frequency onto a shared log-spaced grid of 40 points, 20 Hz-10 kHz, then level-aligned by its mean over the grid's 200 Hz-1 kHz slice (12 points, ~186 Hz-~1.07 kHz), so rig level offsets don't count. The root-mean-square (RMS) of the 40 dB differences at those same frequencies is the distance: similarity % = 100 - 6 × RMS dB, rounded, clamped to 0-100 - every 1 dB of RMS difference costs exactly 6 points. Labels: under 1 dB "Near identical", under 2.5 "Very close", under 4 "Close". Matches past 4 dB RMS are never shown - past that, treble differences are clearly audible and calling two tunings alike isn't defensible, so this section never appears without a match under 4 dB. Only real measurements enter the index (20+ raw points per curve); matches pair the same form factor only (in-ear with in-ear, headphone with headphone - rigs are never mixed), and curves under 0.5 dB apart count as the same-tuning re-listing, never matched - so a match is always a different model, and each card is one distinct tuning: a model's variant listings (its Red / Grey / Open-Box SKUs, all under 0.5 dB from each other) fold into the one card, shown at whichever variant carries a tracked price. Matches are shown most similar first. A measurement-space fact, not a promise two ears agree.

A match covers tuning only: two products can measure near-identical yet differ in everything a frequency-response curve cannot see - build and shell material (metal vs plastic), durability, driver matching and QC unit-to-unit, comfort, fit and isolation, included cable and accessories, warranty and after-sales service, resale value. That's usually what the pricier one charges for - weigh those against the price gap before paying more for the "same" sound.