Home/How it works

How this works

Three steps from your favorite song to your rig. We do the listening and map the chain; you get the numbers.

  1. We chase the tone

    Every recipe starts with the historical record: producer interviews, gear lists, period photos, live footage, and the session log when one exists. Our AI research pipeline reads all of it and works out what was plugged into what, with sources cited and inferred settings marked as inferred. When the record is thin, the recipe says so.

  2. We map the chain

    Every block, every knob, every tap-tempo. The full signal path (guitar, drives, amp, cab, mic placement, post effects) is documented like a service manual. If the original tone uses snapshot switching, parallel routing, or a specific footswitch assignment, the recipe spells it out.

  3. You get the numbers

    The chain gets translated into the block names and parameter values your modeler uses: Helix, Quad Cortex, TONEX, Fractal, Kemper, Katana, or a physical pedalboard. dB is dB, Hz is Hz, time is ms. Helix, Quad Cortex, and Katana players download a preset; everyone else dials the numbers in by hand.

What you can count on

Sources cited

Every recipe cites its sources: producer interviews, Equipboard listings, gear photos, period-correct mods. When we don't know, we say so. When we get one wrong, the fix ships as a public commit.

Era-correct gear

We don't list signature pickups on a recording from 1969. The Hetfield Master of Puppets recipe uses the Seymour Duncan Invaders in his 1985 King V, not the EMG 81/60 set he switched to two years later.

Real units on every knob

dB on the cab level, Hz on the high-pass, milliseconds on the delay. A generic 0–10 mark is impossible to repeat.

Cab + mic placement

When the original tone is dual-mic'd (SRV's 4x10 Vibroverb with a Sennheiser 421 close and a Royer 121 ribbon further back), the second mic is built into the same Helix patch.

Ready to play

Pick a song from the archive and dial it in. The blog goes deeper on why the settings work.

Curious who's behind this? Mostly an AI, and we say so. How we work explains the machine, and the experiment shows the running record.