Measurement libraries
PedalScope's free tier can open, chart, compare, and listen to any measurement — it just can't make hardware measurements without the Measurement Unlock. So here are real devices, already measured: measurement libraries you can download and open in PedalScope at no cost, no purchase needed. Each one was captured under stated conditions, exported from the shipping build exactly as measured, and is published with its size, its SHA-256 and its licence beside the link. Further sets join this page as they are measured.
To open one: unzip, then double-click the
.pedalscope file inside. PedalScope opens it and reports
what arrived. Opening a library adds to yours — nothing
is replaced or removed, and opening the same library twice never
duplicates anything. Every imported record carries an
imported chip saying where its numbers came from, and
its calibration describes the rig that measured it, not yours. The
details are in the guide:
Sharing and opening
libraries.
Reference Box 2026-09-19
| What it is | The project's hardware Reference Box — six circuits (P1–P6) simple enough that their distortion was computed from resistor arithmetic and Fourier theory before anything was plugged in. Built by the project; nobody else's product. |
|---|---|
| Captured | 2026-09-19, one session, one calibration, under the project's publishable-set checklist. |
| What it holds | 37 records under one pedal: six positions × six measurement kinds (Harmonic Distortion, Transfer Curve, Compression, Gain Map, Chord IMD, Waveform Matrix), plus one single-sweep Harmonic Distortion capture of P3 noted as such. Harmonic Distortion at 8 averages; Waveform Matrix 7 × 7 (E2–E6, seven levels); a control-state photo on every position. |
| The rig, in one line | RME Babyface Pro at 96 kHz, instrument input (1 MΩ), through the project's breakout box; drive calibrated in volts at the box's input (0 dBFS = 4.828 V pk, so the standard −26 dBFS drive is 242 mV pk); source impedance 0 Ω; one loopback calibration, 11:18:37, embedded byte-for-byte in every record (latency 1169 samples, flatness +0.10 / −0.63 dB, floor 0.009 %). |
| Download | PedalScope-Reference-Box-2026-09-19.zip — 73.2 MB zipped (158 MB unpacked). Hosted as a release asset of the public repository PhaseDog/pedalscope-library. |
| SHA-256 | e13765baab8973ed46ad7bb23f839e5cfedc62e1b3b389ac776d1c3ea0f9a835 — to check the download, run shasum -a 256 on the zip and compare. |
| Licence | CC BY 4.0 — use it, share it, build on it, with attribution to PedalScope. The PedalScope name and mark are not part of the grant. |
Why this set
Because for this box, the right answers exist independently of the instrument. A famous pedal can only show you what PedalScope says; the Reference Box shows you PedalScope being checked. Every number below is read from the records in the download by the project's own dump tool, never from a screen, and the expected value beside it was computed first.
The one-diode demonstration
The centerpiece is a pair of circuits, P3 and P4, that differ by one diode. P3 clips with a matched pair, one pointing each way — symmetric. P4 uses the same resistor but a single diode one way and two in series the other — asymmetric. Same box, same session, same drive. At 240 Hz, harmonic levels re the input:
| P3 (matched pair) | P4 (one diode) | |
|---|---|---|
| 2nd harmonic | −82.8 dB | −54.1 dB — 28.7 dB more |
| 3rd harmonic | −47.4 dB | −56.3 dB |
| H2 relative to H3 | −35.4 dB (odd-dominant) | +2.2 dB — a 37.6 dB swing |
| Compression knee (220 Hz) | 416 mV pk | 539 mV pk |
| Gain Map, distortion peak | 22.4 % | 15.5 % |
Where to look for the asymmetry is itself part of the lesson. A transfer curve will not show it: every pedal has an output coupling capacitor, and a series capacitor removes an asymmetric clipper's level asymmetry while leaving its harmonic asymmetry intact — driven clear of both thresholds, this box's P4 draws clip levels within 7 % of each other. (At the set's standard drive of 242 mV pk the transfer captures sit below both positions' knees and draw a straight line.) The one diode lives in the even harmonics. Open the pair in Compare and the Harmonic Distortion overlay says it directly.
The arithmetic, published beside the measurement
| Quantity | Computed in advance | PedalScope measured | Difference |
|---|---|---|---|
| P1 divider attenuation (resistor arithmetic, 1 MΩ load included) | −20.16 dB | −20.21 dB, flat from 85 Hz to 6.9 kHz | 0.05 dB |
| P2 one-pole low-pass corner (as-built values into 1 MΩ) | 1606 Hz | 1564 Hz, plateau −0.18 dB | 2.6 % |
| P5 rectifier, H4 relative to H2 (Fourier theory) | −13.98 dB | −13.97 dB | 0.01 dB |
| P5 rectifier, H6 relative to H2 | −21.34 dB | −21.33 dB | 0.01 dB |
| P5 rectifier, H8 relative to H2 | −26.44 dB | −26.42 dB | 0.02 dB |
| P6 crossover stage, distortion against frequency | THD ∝ f | THD ∝ f0.973 over 82 Hz–3 kHz | exponent 0.973 against 1 |
P5 is a precision rectifier: it suppresses its own fundamental by 48.7 dB while producing an even-harmonic ladder whose spacing is pure Fourier theory, and the instrument reads that ladder to a hundredth of a decibel. It is the sharpest single check in the box, and it is in the library for you to open.
What the records carry — and what they don't hide
Every record ships exactly as it was measured, with its own calibration, its measurement floors, and its bounds wording. Nothing was idealised, and no record was re-taken because a nicer one was wanted. Three things the set carries that a reader should know in advance:
- P4's Chord IMD record carries a coherence notice — its check reads 10.8 dB over the bar, as it has on every capture of this box since the check existed. The record ships with the notice it carries; the app states that the capture's floor and product levels are not trustworthy, and the page says so here rather than leaving it for you to find. Its harmonic records are unaffected.
- The bracketing null runs are not in the set — a bare-loop capture was taken before and after the box, and their floors are stated here instead: fundamental gain 0.000 dB on both; harmonics H2–H5 at −119 to −123 dB re input; Chord IMD capture floor −137.4 and −137.3 dBc, coherence 22.2 and 19.8 dB under the bar; Gain Map peak 0.043 % and 0.036 %. The set sits between two clean floors.
- P1's Gain Map peak tile reads 0.28 %, on a 20 dB passive divider, from its two lowest-frequency columns at the quietest level — a corroborated number where an earlier rig read a bound. It is reported as measured; it is not a property of the divider.
The per-position control-state photos and notes were attached to each position's Harmonic Distortion record after the run, from photos taken at the bench before each position was measured; the photo shows on every record of the position.
About averaging. This set's Harmonic Distortion records were averaged over eight sweeps; PedalScope's default is four. Compare a record you measure at the default against one of these and Compare shows its cross-averaging note — neutral wording, and the distance is still computed. That is the instrument stating that the two sides have different resolution limits, not a fault on either side.
A published set is permanent: a later capture of the box is published under its own dated name and arrives as a separate pedal, so the two compare cleanly instead of interleaving. The box itself is documented in the project's bench notes, and an open-hardware release of its schematics and layout is planned — so the check stays checkable: build one, measure it, and compare your library against this one.