Research standards

How ToneTwin researches a guitar tone

ToneTwin separates three things that are often blurred together: what a source documents, what the recording itself supports, and what a player can try on a different rig. That separation is the core editorial rule.

Researched and maintained by the ToneTwin Research Team · Last reviewed July 20, 2026

The short version

• Reported guitars, amps, pedals and studio details need a traceable source.
• Primary interviews, producer or engineer accounts, session documentation and manufacturer archives outrank summaries.
• Conflicting or era-incompatible claims are lowered in confidence or omitted.
• Exact historical knob positions are not published unless a direct source documents them.
• Settings for a user's own rig are modeled recommendations, not claims about the original control panel.

1. Define the exact recording and part

Research starts with the artist, canonical song title, original release era, instrument and section. A verse clean part and a solo can use different chains. Live rigs, later tours, remasters and signature products do not automatically prove what was used during the original session.

2. Follow an evidence hierarchy

Tone-recipe sites, preset marketplaces and competing tone-matching services are excluded as research sources. They can repeat the same unsupported guess and make it look corroborated when it is not.

3. Check identity, chronology and consistency

• Identity: model, player, studio version and part must match.
• Chronology: gear released after the session cannot be part of that recording.
• Consistency: named gear must appear in the chain, and prose cannot silently introduce extra equipment.
• Uncertainty: “reported,” “attributed” and confidence notes are retained instead of converted into false certainty.

4. Why original amp knob values are usually absent

Most sessions did not publish a readable control-panel photograph or an engineer's exact settings. Even when numbers exist, two amps with the same labels can have different tapers, circuits, speakers, voltage and volume. ToneTwin therefore does not turn an AI estimate into a historical claim. Public tone pages focus on the reported rig, pickup, gain character, effects, chain order, tuning and playing context.

After a player selects their own guitar, amp and pedals, ToneTwin can model a practical starting point for that equipment. Those values are recommendations to test by ear. They are not described as the artist's original knob positions.

5. How AI is used

AI helps search, extract candidate facts, normalize gear names and adapt a tonal target to a user's controls. It is not treated as a source. URLs are preserved for inspection, blocked source domains are removed before storage and display, and historical settings are suppressed unless directly documented. Automation can still make mistakes, which is why uncertainty and corrections remain part of the product.

6. Dataset and aggregate-study rules

The vetted July 20, 2026 catalog snapshot contains 508 public tone records representing 508 unique normalized song–artist pairs, 233 artists and 1,659 retained citation links. Legacy V1 research and duplicate variants are excluded. Aggregate reporting excludes knob values, states the snapshot date and describes catalog composition. The collection is broad and curated; it is not a popularity-weighted census of every recording ever made.

Read the 508-tone aggregate study for the current findings and limitations.

Corrections and accountability

Every public tone page includes a correction control. A useful correction identifies the exact claim, supplies a direct source and explains why it applies to that recording rather than a later live rig. General product feedback can also be sent through the feedback page.