How we
review speakers.

The model starts with one input: the measurements every speaker ships with. A frequency response curve, and a Thiele‑Small parameter set. Both are standardised engineering measurements taken under controlled conditions by the manufacturer.

What follows is model-building. Corner frequencies, band-level features, peak magnitudes, damping regimes, compression characteristics — everything derived from the same feed, run through consistent rules, compared across the catalogue. Nothing hand-typed per speaker. Nothing smuggled in from outside.

Validation is the test of whether the model matches how guitarists actually hear speakers. Two checks. First, first principles — does the physics predict what a player hears? Second, vocabulary — does the model’s zone for tight bass line up with what guitarists already call tight? Does piercing fall in the 2 kHz band where players expect it? Where the model and the ear disagree, the model yields.

Guitar speakers come in different efficiencies, different impedances, different measurement conventions. Compared raw, the curves mislead — a 100 dB speaker looks louder than a 94 dB one by a full centimetre of graph, regardless of voice.

Normalisation is the fix. Every frequency-response curve on this site is centred on its own flat passband — the mean SPL across 200–500 Hz, the part of the response doing the least colouring. That becomes the zero. What you see in every plot is shape, not level. Sensitivity differences live separately on the spec card.

From that common zero, everything that sits above or below — bass rolloff, presence peaks, top-end extension — gets compared on the same axis. One speaker against another, apples to apples. Band labels, zone scores, and comparison data share one and the same baseline.

The model scores every speaker on a hierarchy of perceptual scales. Each scale translates a physical measurement into a position on a spectrum, and each position carries a word guitarists already use to describe it. Not a quality judgement. A loose speaker isn’t worse than a tight one; it’s a different character.

Five compound scales roll up into the summary you see on the spec card — the speaker’s headline character at a glance. Nine detail scales sit beneath them, feeding the compounds and surfacing on the lens tag chips when a reader digs in.

Zone thresholds are calibrated against the full catalogue and cross-checked against known reputations — the Vintage 30’s profile aligns with its reputation for tight bass, aggressive presence and high efficiency.

Ask a guitarist what a speaker does and the answer comes out in three dimensions: what it sounds like on a record, what it feels like under the hands, and what happens when the amp works harder. That’s the whole shape of how players experience a speaker.

The three lenses on this site map to those axes directly. Voice is the frequency-domain character — the tonal shape, the peaks, the voicing family. Feel is the time-domain response — pick attack, decay, cone grip, amp dependence. Headroom is what happens under drive — clean ceiling, compression character, breakup onset.

Not three separate tests. Three readings of the same measurement set, translated into the axes a guitarist already uses.

Nobody pays us to review anything. No review units. No brand partnerships. No display advertising.

We offer an Amazon affiliate link if you’d like to purchase a speaker — it’s how we keep the lights on here. If you prefer Sweetwater or Guitar Center, they’re there too; we don’t get a cut. The review says what it says either way.