Knowledge hub / MusicMuse / gear
The Rig Mechanism Index
Music knowledge / Guides and practical tools
About this reference
From the music-learning collection, adapted for Philojain Music Muse. Referenced sources remain credited in the article.
134 things that change what a guitar rig sounds like — what each one does, and what it does to the sound. Searchable, grouped by where in the chain it acts.
One hundred and thirty-four things that change what a guitar rig sounds like, each with what it actually does and what it does to the sound. They are grouped by where in the chain they act — the instrument itself, then gain, filtering, time, modulation, dynamics, the amp, the speaker and the routing. Filter by category or search the whole set.
Pickup resonant peakthe RLC peakInstrument
A pickup is an RLC circuit peaking at f₀ = 1/(2π√LC). A directly measured Strat single coil — 1.58 H, 222 pF — self-resonates at 8,500 Hz, and the C in that equation includes everything downstream.
Sounds like: where the guitar's character sits: below 1 kHz reads dull and hollow, 2.0–2.5 kHz is the singing humbucker region, 3.0–5.0 kHz the bright metallic Strat and Tele region
Cable capacitancethe first tone controlInstrument
Real guitar cable measures 15.8–48.8 pF per foot. An 18-foot run takes a Strat pickup's peak from 8,500 Hz to about 4.6 kHz; 35 feet takes it to about 3.6 kHz.
Sounds like: duller and more mid-peaky at the same time — the top gone and a resonance pulled down into the midrange
Capacitive vs resistive loadingthe two tone sucksInstrument
Capacitance shifts the peak down in frequency — cable, plus roughly 15 pF per true-bypass pedal even switched off. Resistance separately damps the peak's amplitude, lowering its Q.
Sounds like: one moves the sparkle down in pitch, the other flattens it out; a healthy Q is 2–5, below 2 shallow and above 5 edgy
Pot value loading250k / 500k / 1MInstrument
A lower pot value loads the circuit harder and damps more treble. Zollner measured a tone pot drifting 250 kΩ→350 kΩ as audibly “more brilliant” and →150 kΩ as “duller” — while calling the difference “not really dramatic.”
Sounds like: a small, real brightness shift — except on a 1 MΩ Jazzmaster volume, which kills treble radically the moment you turn down at all
Pickup position comb filterInstrument
A pickup samples the string at one fixed point, so it nulls any harmonic with a node underneath it: V = sin(π·X·F/(L·Fopen)). On a Strat, neck pickup 6.375 in from the bridge: first peak 165 Hz, first null 330 Hz.
Sounds like: neck pickups deep and bridge pickups bright because of geometry, not because of winding
Two pickups blendedthe quackInstrument
Blending adds a cosine term that nulls at odd multiples only — a different notch pattern rather than a volume blend.
Sounds like: the hollow, scooped, slightly out-of-phase Strat position-two and -four sound
Pickup inductanceInstrument
The dominant electrical parameter, setting resonant frequency and correlating with output. A Strat single coil measures about 1.7 H bare and 2.2 H with magnets; the range across types is roughly 1–10 H, and inductance scales with turns as N2.14.
Sounds like: higher inductance reading as darker and fatter, lower as brighter and thinner
DC resistanceDCRInstrument
A poor proxy for output. Pull the magnet out and DCR is unchanged while output collapses to near zero; two 7,500-turn Strat pickups wound in AWG 42 and 43 read 5.5 kΩ and 7.0 kΩ with nearly identical output voltage.
Sounds like: nothing reliable at all — the number on the spec sheet does not predict what you will hear
Active pickup preampEMG-styleInstrument
An onboard preamp buffers the coil to a fixed low output impedance before the cable. A measured EMG81: 10 kΩ out, 1.25 V average, resonance fixed at 2.25 kHz, noise floor −91 dBV.
Sounds like: the same tone regardless of cable length or what follows, with a very low noise floor and no interaction with anything downstream
Coil splitInstrument
One coil of a humbucker is shorted to ground, leaving the other active. Output drops and hum cancellation is lost. Not the same thing as a coil tap.
Sounds like: thinner, brighter and noisier, with mains hum returning — single-coil-ish rather than an actual single coil
Coil tapInstrument
A wire brought out partway along one coil's own winding, giving a lower-turn-count output with nothing disconnected. Rare on humbuckers; the word is almost always used to mean a split.
Sounds like: a lower-output version of the same voice rather than a different one
Scale lengthInstrument
T = μ(2Lf)², so the same gauge at the same pitch runs about 6% more tension on a 25.5″ Fender scale than on a 24.75″ Gibson scale.
Sounds like: longer scale reading as tighter and more articulate under the fingers, shorter as slinkier and easier to bend
Multiscalefanned fretInstrument
Each string gets its own scale length — commonly 26.5″ to 28″ on an eight-string — with the perpendicular fret usually between the seventh and ninth, so the low strings gain length without over-tensioning the trebles.
Sounds like: low strings that stay tight and defined at pitches where a standard scale goes floppy
Inharmonicitystring stiffnessInstrument
B = π³Ea⁴/(4TL²), with overtones following fn = n·f₀√(1+Bn²) and curving sharp. B rises with the fourth power of string diameter and falls with tension.
Sounds like: thick strings at low tension sounding progressively out of tune with themselves as the overtones drift sharp
Neck dead spotInstrument
A note whose frequency coincides with one of the neck's own resonances while its fret sits near that mode's antinode, so string energy drives the neck instead of staying in the string. Fleischer documents a Strat dead spot at the fourth fret of the D string with a measured 430 Hz neck resonance.
Sounds like: one specific note on one specific fret dying far faster than its neighbours, for no reason you can hear coming
Downtuning tension lossInstrument
Since tension scales with the square of frequency, dropping a whole step at fixed gauge leaves about 79% of standard tension — and lower tension raises inharmonicity at the same time.
Sounds like: floppy, buzzing, progressively out-of-tune low strings unless gauge or scale compensates
String alloyInstrument
Pickups sense the disturbance a ferromagnetic string makes in a static field, so permeability governs coupling. Pure nickel couples least, nickel-plated steel more, stainless most.
Sounds like: pure nickel warmer and rounder, nickel-plated steel brighter and more balanced, stainless brightest and hottest
Action over scaleInstrument
A peer-reviewed study of string bending found that fretboard action height influences the force required far more than vibrating string length does.
Sounds like: how hard the guitar feels to play being set by the setup, not by the scale on the spec sheet
Soft clippingdiodes in the loopGain
Clipping diodes sit inside the op-amp's feedback loop, so gain is actively reduced as the signal approaches threshold and the waveform rounds instead of squaring. A Tube Screamer: silicon diodes at about 1 V, input ~447 kΩ, output ~1.2 kΩ, gain 12× to 118×.
Sounds like: a smooth, rounded, mid-forward overdrive that compresses gently rather than tearing
Hard clippingdiodes to groundGain
Diodes sit after the amplifying stage, shunted to ground rather than in the feedback path, so the op-amp slams into a fixed ceiling. A ProCo RAT's LM308 swings toward full gain — up to 2,305×, 67 dB — before the clamp.
Sounds like: a much squarer wave with denser harmonics: aggressive and flat-topped rather than rounded
Cascaded clippingGain
Diodes in the feedback loops of two cascaded transistor stages at about 0.6 V each. A Big Muff's measured stage gains are 23 dB and 25 dB — nominally soft clipping, audibly not, because cascading matters as much as topology.
Sounds like: an enormous sustaining wall of fuzz with almost no dynamic response left
Transistor saturationno diodesGain
A Fuzz Face has no clipping diodes anywhere. Distortion is the transistor's own operating point, deliberately mis-biased with the collector near −1.6 V rather than the theoretical −4.5 V, so small signals clip asymmetrically and large ones drive both half-cycles.
Sounds like: spitting, unstable, violently touch-sensitive fuzz that changes character with how hard you pick
Shunt-feedback input gainGain
The first stage's gain is approximately Rf/Ri, where Ri is whatever source impedance is in front of it. Lower the source impedance and the gain rises and flattens across frequency. This is a gain change, not tone suck.
Sounds like: a fuzz that gets louder, brighter and harsher and loses its cleanup the moment something buffered is placed in front of it
Low input impedanceGain
Simulated at component level a Fuzz Face's first stage works out around 8 kΩ, and the feedback network drops it to roughly 5 kΩ, varying 5.2–8.4 kΩ with the Fuzz control. It hears what is in front of it.
Sounds like: a circuit that responds to the guitar itself — roll the volume back and it genuinely cleans up to near-clean
High input impedance driveGain
A Tube Screamer presents about 447 kΩ and a RAT about 494 kΩ — high enough to swamp any pickup interaction, with large fixed gain ahead of a fixed diode threshold.
Sounds like: the same sound regardless of guitar or volume knob position, just quieter when you turn down
Feed-forward clean blendthe Klon architectureGain
The clipped path is summed with two feed-forward clean paths at a summing amplifier, and a charge pump supplies +18 V and −9 V so the op-amp does not slew-limit on transients.
Sounds like: drive that adds grit without losing the dry attack underneath — loud, open, and dynamically intact
Germanium vs siliconGain
Bandgap 0.66 eV against 1.1 eV, so VBE is roughly 0.1–0.3 V against 0.6–0.7 V, and germanium leakage runs 0–10 µA where silicon is orders of magnitude lower. Leakage doubles per +10 °C in both — the coefficient is the same and the baseline is a thousand times higher.
Sounds like: a fuzz whose bias wanders with temperature: gated and farty under stage lights, thin and weak in the cold
hFE selectionGain
R.G. Keen: “a definite sweet spot for musical sounding clipping at transistor gains of about 80–110,” conventionally Q1 near 80 and Q2 near 120. AC128s as delivered ranged from 16 to about 180.
Sounds like: the difference between a fuzz that sings and one that splutters — the same circuit, a different pair of parts
Treble boosterGain
A single-transistor high-pass boost that removes low frequencies before the amplifier clips, so the amp's distortion stays articulate instead of turning to mush.
Sounds like: a cranked amp that suddenly cuts and sings instead of blurring — thin on its own, enormous through the amp
Octave up by rectificationOctaviaGain
A step-up transformer feeds diodes arranged for full-wave rectification, flipping the negative half of each cycle positive and doubling the waveform's repetition rate. Not ring modulation — there is no carrier oscillator.
Sounds like: a screaming upper octave riding on the fuzz, clean on single notes near the twelfth fret and clangy chaos on chords
Octave down by divisionflip-flopGain
The signal is squared by a Schmitt trigger, then a D-type flip-flop toggles at exactly half the input frequency.
Sounds like: a synthetic, square sub-octave under the played line, glitching whenever tracking loses the fundamental
Monophonic tracking limitGain
Both analogue octave types need one unambiguous repeating feature per cycle. Low strings and chords produce dense waveforms with multiple zero-crossings per true cycle, so the detector has nothing to lock onto.
Sounds like: octave effects that work beautifully high on the neck and fall apart into stuttering garbage on low notes and chords
Preamp vs power amp distortionGain
Preamp distortion is voltage gain against the tube's curves with the tone stack alongside, so it is EQ-shapeable and described as more compressed and smoother with higher sustain at low volume. Power amp distortion is current delivery downstream of the tone stack — richer, punchier, less compressed.
Sounds like: preamp gain sounding controlled and tight at any volume; power amp gain blooming and opening up, but only when it is loud
Stacked drivesGain
Two low-gain stages each adding a small amount of soft clipping, rather than one stage hard-clipping. The second is usually set for level rather than for gain.
Sounds like: layered, complex saturation that keeps note definition where a single high-gain pedal would go flat
Frequency-selective clippingFilter
A filter placed before a clipper decides which frequencies distort most; a filter after only reshapes what already exists.
Sounds like: the boosted band saturating hard while everything outside it stays comparatively clean
Pre-clipping high-passthe 720 Hz cornerFilter
Inside a Tube Screamer's feedback loop sits an active high-pass at 720 Hz, so harmonics above it get the full gain of the distortion stage and everything below gets progressively less.
Sounds like: the tightening: palm-muted low strings stop turning to mud because the mud never reaches the gain stage
Post-clipping low-passFilter
A separate passive low-pass at 723 Hz shapes what the clipping generated. Same corner frequency, opposite side of the nonlinearity.
Sounds like: the fizz above the distortion rolled off without touching the distortion itself
Sweeping bandpassthe wahFilter
A swept bandpass boosting one band ahead of a level-dependent clipper means the frequencies in the boosted range distort most — and the whole spectrum is not hitting the clipper at once, which cuts harsh intermodulation.
Sounds like: a vocal sweep that changes what is distorting, not just what is loud
Wah inductor and QFilter
Measured across brands and decades: 507–759 mH, DCR 15.5–30.7 Ω — a two-to-one range. Q tracks DCR. The spread within one nominal type across production years matches the spread between types.
Sounds like: high Q giving a sharp, narrow, vocal peak and low Q a mellow, smooth sweep — set by the part, not by the label on it
Envelope filterauto-wahFilter
The filter's corner is driven by the input's own envelope rather than by a treadle or an LFO.
Sounds like: a filter the picking hand plays: dig in and it opens, lay back and it closes
Passive FMV tone stackFender / MarshallFilter
Bass, mid and treble in one interactive passive network. A 5F6-A Bassman stack loses about 10 dB at maximum mid, Marshall's about 7 dB; bass gives about 20 dB of range at 81 Hz and treble about 26 dB at 6 kHz; the dip centres near 500 Hz. It is cut-only — it cannot boost mids at any setting.
Sounds like: the familiar scooped amp voice, with all three controls fighting each other and all three at zero giving near silence
Tone stack placementFilter
Fender puts the stack after a cathode follower following the first gain stage, isolated from what drives it. Marshall's JTM45 puts it after the second stage, plate-driven and interacting — two stages of grind before the EQ instead of one.
Sounds like: the same three knobs shaping a clean signal in one amp and a distorted one in the other
Vox Cut controlFilter
Not part of the tone stack at all: an adjustable high-frequency filter wired across the two plates of the long-tailed-pair phase inverter, downstream of the whole preamp. Turning it up removes treble by cancellation.
Sounds like: top end coming off the entire amp at once, after everything else has already happened
Baxandall stackactive EQFilter
Frequency-selective components in a gain stage's feedback path, so it can boost as well as cut. Bass and treble both at zero gives a mid boost; both maxed gives a scoop — the inverse of the passive stack's logic.
Sounds like: EQ that actually adds what you ask for instead of only removing what you did not
Tilt EQJames networkFilter
A passive two-band network that reduces attenuation asymmetrically rather than boosting: one way cuts treble relative to bass, the other inverts it, flat at the centre detent, with minimal interaction between controls.
Sounds like: a single knob tipping the whole balance from dark to bright without changing anything else
Slew-limited voicethe LM308Filter
A RAT's tone control sweeps a low-pass from 475 Hz to 32 kHz, and the LM308's 0.3 V/µs slew rate — about forty times slower than a TL071 — imposes a practical ceiling near 5.3 kHz that is itself part of the pedal's voice.
Sounds like: distortion with a natural top-end limit built in: dense and gritty but never brittle
Muff tone notchFilter
A Big Muff's tone stack produces a genuine notch centred at 1 kHz with about 6.5 dB of extra loss there, on top of roughly 7 dB of overall stage loss.
Sounds like: enormous scooped fuzz that disappears completely the moment a band starts playing
Treble bleedFilter
A capacitor of 220–1500 pF, often with a 100–330 kΩ resistor, across the volume pot so treble survives turning down.
Sounds like: a guitar volume knob that gets quieter without getting darker
Bucket-brigade delayBBDTime
An analogue shift register: charge stored on capacitor stages and clocked down the chain. Delay equals stages divided by two, divided by clock frequency. The MN3005 has 4,096 stages; the MN3007 has 1,024, giving 5.12–51.2 ms.
Sounds like: warm, imprecise repeats that are never quite identical to the note that caused them
CompandingTime
A compander compresses before the delay line and expands after, so the line's own noise is pushed down by the same ratio the signal was lifted.
Sounds like: long analogue delays that are usable at all, instead of hissing louder than the guitar
Anti-clock filteringTime
Low-pass filters of 30–36 dB/octave with a corner near 3 kHz sit before and after the BBD to strip clock feedthrough — and every feedback pass goes through them again.
Sounds like: repeats that are already dark on the first one and get progressively darker until they dissolve
Tape echoTime
Head spacing divided by tape speed sets delay time. A Space Echo RE-201 has three playback heads in twelve combinations: head one 69–177 ms, head two 131–337 ms, head three 189–489 ms.
Sounds like: repeats that wobble in pitch, saturate and lose top end on every pass, in rhythmic patterns rather than single echoes
Self-oscillationTime
Feedback past unity gain and the loop runs away. Onset is program-dependent, not a fixed knob position.
Sounds like: the delay taking over and building into a howling, rising wall that has to be pulled back by hand
Echoplex preampTime
A simple JFET gain stage that, with volume at maximum, gets only about 6 dB above unity. A tone conditioner rather than a boost pedal.
Sounds like: a modest lift and a midrange thickening that makes the amp behind it feel bigger than it is
Digital delay headroomTime
No anti-clock filter is needed, so repeats keep full bandwidth — but the ubiquitous PT2399 clips at about 3 V peak-to-peak and distorts above roughly 4 kHz, so “digital means more headroom” is a per-chip fact, not a category fact.
Sounds like: repeats that are exact copies, for better and for worse
SlapbackTime
A single repeat at roughly 80–120 ms, used as a rhythmic thickener rather than as ambience.
Sounds like: a hard doubled attack on every note, tight and dry, with no wash at all
Dotted-eighth delayTime
Repeats land on the off-beats, so a simple arpeggio becomes a continuous sixteenth-note pattern. The part is written for the delay and does not work without it.
Sounds like: one guitarist sounding like two, with notes appearing between the ones being played
Spring reverbTime
A transducer at each end of a helical spring launches and recovers torsional waves. The spring is dispersive — different frequencies travel at different speeds — so a transient smears out by frequency rather than arriving together.
Sounds like: the drip: a metallic, splashy, unmistakably mechanical reverb that no algorithm reproduces
DwellTime
A drive-level control feeding the tank's input transducer, not a decay-time control at all. Drive it harder and the springs themselves distort.
Sounds like: the reverb crashing and overloading rather than simply getting longer
Tank decay ratingTime
Short tanks 1.2–2.0 s, medium 1.75–3.0 s, long 2.75–4.0 s; guitar amps typically use long. Multiple springs in parallel fill in each other's resonant dead spots.
Sounds like: denser and smoother with more springs, boingier and more resonant with fewer
Plate reverbEMT 140Time
A suspended sheet of steel about 2.4 by 1.2 m weighing around 270 kg, driven by a transducer, with decay set by a movable damping pad.
Sounds like: dense, bright, immediate reverb with no early reflections and no sense of a room
Algorithmic vs convolutionTime
Algorithmic reverb synthesises decay from delay lines, comb and all-pass filters tuned to approximate a space. Convolution convolves the input against a recorded impulse response of an actual space.
Sounds like: algorithmic being tweakable and unreal; convolution being fixed and exactly like somewhere
ShimmerTime
A pitch shifter of +12 semitones, sometimes with a second voice at +24, placed inside the reverb's own feedback loop, so every pass rises again.
Sounds like: reverb that climbs upward into a choral, endlessly ascending wash
Reverse reverbTime
Reverb applied to a reversed signal, so the tail arrives before the note.
Sounds like: sound swelling backwards into each note, with the attack buried at the end instead of the start
All-pass stageModulation
Equal gain at every frequency with phase shifting continuously from 0° toward 180°, passing exactly 90° at the stage's own corner frequency.
Sounds like: nothing on its own — audible only once it is mixed back against the dry signal
Stages to notchesModulation
n all-pass stages produce n/2 notches. A Phase 90's four stages give two notches, with calculated rest positions at 58.5 Hz and 340.8 Hz; a Phase 45's two stages give one.
Sounds like: fewer stages sounding gentle and watery, more stages sounding thick and jet-like
Phaser modulator typeModulation
A Phase 90 sweeps four matched JFETs used as voltage-controlled resistors; a Small Stone modulates OTA transconductance directly.
Sounds like: the same effect with a different sweep character — one liquid and even, the other harder-edged
Uni-Vibe staggerModulation
Four phase stages with deliberately mismatched capacitors — 330 pF, 470 pF, 4 nF and 15 nF — all swept by one incandescent bulb driving four photocells, which respond faster to rising light than to falling.
Sounds like: an uneven, lopsided throb with unequally spaced notches, quite unlike an evenly matched phaser
ChorusModulation
A modulated delay of roughly 5–40 ms with no feedback path. The sweeping delay length Doppler-shifts the wet copy, which is the detuning.
Sounds like: two slightly out-of-tune guitars playing the same part
FlangerModulation
A modulated delay under about 20 ms with feedback, producing evenly spaced comb notches that regeneration deepens and narrows.
Sounds like: a metallic jet sweep rather than a detuning — hollow and watery on negative feedback, whooshing on positive
VibratoModulation
Pitch modulation with no dry signal mixed in, so there is no comb filtering at all.
Sounds like: pure pitch wobble, with none of the hollowness of chorus or phase
Optical tremoloModulation
An LFO drives a lamp — a neon needs about 90 V to strike — sensed by a photocell forming part of a divider. The lamp re-striking each cycle is what makes it choppy.
Sounds like: a hard, square, stuttering throb rather than a smooth swell
Bias tremoloModulation
The LFO is ridden onto the power tubes' grid bias, cyclically shifting the output stage's operating point and therefore its gain.
Sounds like: a thicker, gooier pulse that changes the amp's distortion as well as its level
Harmonic tremolobrownfaceModulation
The signal is split by an RC crossover — about 320 Hz on the Fender 6G4 — into two bands whose gains are modulated out of phase with each other, then recombined. As treble rises, bass falls.
Sounds like: a swirling, phase-like sway rather than a simple volume pulse
Rotating baffleLeslie / VibratoneModulation
A real speaker mechanically Doppler-shifted by a spinning baffle or horn.
Sounds like: a three-dimensional wobble with genuine pitch and amplitude movement that no phase-shift circuit reproduces
Whammy flutterModulation
The tremolo arm held constantly so sustained notes never sit still in pitch.
Sounds like: every long note drifting and wobbling slightly, never settling
Glide guitarModulation
The vibrato arm held through strummed chords so the entire chord's pitch bends continuously.
Sounds like: whole chords sliding and warping in pitch as they ring
Ratio and thresholdDynamics
Arithmetic: at 4:1, for every 4 dB above threshold the compressor allows 1 dB through.
Sounds like: the loud notes pulled down toward the quiet ones
Attack timeDynamics
Roughly 0–1 ms squashes the transient and adds perceived sustain, 1–10 ms is moderate, and 10–30 ms preserves the snap by letting the attack through before clamping.
Sounds like: fast attack removing the pick, slow attack exaggerating it
ReleaseDynamics
The control that actually extends perceived sustain, by determining how quickly gain is restored after the note decays.
Sounds like: notes that hang on and bloom instead of dying
OTA compressionDyna Comp / RossDynamics
Built around the CA3080 operational transconductance amplifier.
Sounds like: the classic squashed, popping country compression with an audible bloom on the tail
FET compression1176 lineageDynamics
Lightning-fast FET gain reduction — the studio 1176's mechanism in a pedal.
Sounds like: aggressive pumping and breathing that you hear working
Optical compressionDynamics
A light-dependent resistor turns on quickly and off slowly, so the physics does the envelope rather than a circuit.
Sounds like: a natural, unhurried, musical squeeze that is hard to hear as an effect
Compressor before driveDynamics
Evens the pick attack before the clipper, giving smoother and more sustained distortion — at the cost of lifting the noise floor, which the drive then amplifies and clips along with the notes.
Sounds like: even, singing sustain, and a hiss that arrives with it
Compressor after driveDynamics
Does almost nothing dynamically, because hard clipping is already a brutal compressor pinning the waveform at a fixed voltage. It becomes a level control instead.
Sounds like: a sustain leveller rather than a compressor — useful, but not the thing people think they are buying
Input gateDynamics
Nearly useless on a high-gain rig, because the hiss and hum are mostly generated by gain stages downstream of it.
Sounds like: noise that carries on regardless of how hard the gate is squeezed
Two-gate detectionDynamics
One gate reads the clean input for detection while a second, after the gain block, does the muting. Detection and muting must happen on different signals for accurate tracking, because distortion puts a sustaining note and the noise floor at nearly the same level.
Sounds like: hard, instant silence between chugs, without the chattering a single gate produces
Volume swellDynamics
The attack removed by raising level after the note has already been struck.
Sounds like: notes blooming in from nothing, like a bowed instrument
Power supply sagDynamics
Three causes: rectifier internal resistance, transformer winding resistance and filter capacitance. Worked example — idle current rising from 70 mA to 170 mA through a 200 Ω secondary gives 20 V of plate drop; a rectifier can add as much as 50 V under full load.
Sounds like: the amp ducking on hard attacks and swelling back up behind them
Phase inverterAmp
A long-tailed pair is the most efficient, linear and balanced-sounding splitter; older paraphase designs create smooth, early distortion with lots of compression.
Sounds like: an amp that stays composed right up to the power tubes, or one that softens well before them
Cathode biasAmp
Self-regulating: rising cathode current raises cathode voltage, which increases negative grid bias and throttles the rise. Under signal it shifts progressively deeper into Class AB, which is the actual mechanism behind its compression.
Sounds like: an amp that squashes and gives way on its own as you push it
Fixed biasAmp
Grid voltage held constant from an external supply regardless of signal, with no self-compensation — more headroom and more output from the same tubes.
Sounds like: a stiffer, louder, later-breaking amp with the attack intact
The 70 percent ruleAmp
In Class AB, peak average plate dissipation happens at mid-swing, about 30% above idle. Biasing idle to 70% of maximum rated dissipation leaves exactly enough margin. An EL34 rated 25 W at 400 V works out near 44 mA. It is a fallback for when the plate load is unknown, not a universal target.
Sounds like: not audible on its own — it is the difference between an amp that lasts and one that eats tubes
Class A in practiceAmp
Aiken's AC30 measurements show both output tubes shutting completely off for a considerable portion of the cycle at full output: the positive swing reaches 138 mA where Class A would require −43 mA and the circuit reaches 0 mA. Class A at idle, Class AB the moment it is pushed.
Sounds like: a designation that describes how hard you are playing, not what you bought
Negative feedbackAmp
A small amount of output-transformer secondary signal fed back to the phase inverter — typically 6–10 dB in guitar amps. It flattens response, reduces distortion inside the loop and raises damping factor.
Sounds like: a tighter, cleaner, more controlled amp with less character of its own
Presence and depthAmp
They work by reducing that feedback at high and low frequencies. The maximum boost available equals the amount of feedback in the loop, so a 6 dB feedback amp offers at most about 6 dB of presence — and an amp with no global feedback cannot have one that behaves this way at all.
Sounds like: top end and low end opening up, along with a change in how tightly the amp grips the speaker
Rectifier dropAmp
A GZ34/5AR4 drops about 10 V at 100 mA and 16 V at 200 mA; the range runs from roughly 10 V for a 5AR4 to about 60 V for a 5Y3GT; silicon drops under a volt.
Sounds like: stiff and immediate with solid state, soft and elastic with the saggiest tube types
Push-pull cancellationAmp
The two output tubes' currents flow in opposite directions through a centre-tapped primary, so their signal sums while even-order harmonics cancel, leaving predominantly odd-order content.
Sounds like: a firmer, more austere distortion with the sweetness subtracted
Single-endedAmp
No cancellation mechanism exists, so even-order content passes straight through, and asymmetric clipping under overload emphasises it further.
Sounds like: a warmer, sweeter, softer breakup at a fraction of the power
TransconductanceAmp
An EL34's gm of about 11,000 µmho is roughly double a 6L6GC's 4,700–6,000, so it saturates and compresses on a smaller input signal.
Sounds like: one tube type giving way earlier than the other in the same circuit — a real difference, though far smaller than the genre labels imply
Output transformerAmp
Impedance scales with the square of the turns ratio. Real Fender primaries run 4.3–6.5 kΩ, not tens of kΩ. A measured single-ended unit held within 1 dB from 53 Hz to 25 kHz, then −2 dB at 40 Hz and −6.7 dB at 20 Hz, with 32% THD at 20 Hz — and mostly second harmonic.
Sounds like: the low end going soft and thick rather than harsh when the transformer runs out of core
Preamp tube muAmp
A 12AX7 has an amplification factor of 100, a 12AT7 about 60, a 12AU7 just 17. Swapping an AX7 for an AU7 cuts stage gain by roughly 5.9 times.
Sounds like: a lower-gain, higher-headroom amp from the same circuit
Ghost notesAmp
Non-harmonic tones riding on the fundamental, produced by 120 Hz supply ripple when filtering is inadequate. Too much sag is a defect, not a feature.
Sounds like: a faint, unrelated pitch hanging behind low notes
Amp output impedanceAmp
Tube amps present 1–10 Ω; solid state under 0.1 Ω. Driving a speaker whose impedance rises to five times its rating at resonance from a high source impedance produces a large output rise there.
Sounds like: a mid-forward, resonant voicing from the tube amp that the solid-state amp flattens out
Wattage and headroomAmp
+3 dB doubles acoustic power, +6 dB doubles pressure, and roughly +10 dB doubles perceived loudness — which needs ten times the power. Fifteen watts to a hundred and fifty buys about one perceived doubling.
Sounds like: not how loud it gets, but how loud it stays clean
Impedance curveSpeaker
A measured 8 Ω Eminence reads exactly 8 Ω at 50, 200 and 500 Hz, peaks near 70 Ω just above 100 Hz, and climbs to about 50 Ω by 20 kHz. A V30 rated 8 Ω measures 7.3 Ω at DC.
Sounds like: the load being a curve rather than a number, which is why what the amp does depends on frequency
Band-limitingSpeaker
A guitar speaker is specified roughly 70–5,000 Hz. Distortion generates harmonics far above that which are simply never heard.
Sounds like: no fizz at all through a real cabinet, because the speaker never reproduces it
Mechanical non-linearitySpeaker
Guitar drivers are deliberately non-pistonic above roughly 500 Hz, and at higher excursion produce cone cry and edge yowl.
Sounds like: the speaker adding its own vocal, ragged distortion on top of the amp's
Thermal power compressionSpeaker
Heating a copper voice coil from 20 °C to 200 °C raises its DC resistance about 72% and drops sensitivity 4.7 dB.
Sounds like: the last watts delivering less than the first, so the amp stops getting louder as it heats up
SensitivitySpeaker
Celestion's twelve-inch range spans 96–100 dB at 1 W/1 m, and differences under 2 dB are barely perceptible. Four decibels of sensitivity beats doubling amplifier power.
Sounds like: the speaker choice deciding the volume far more than the wattage on the head does
Published speaker specsSpeaker
G12M Greenback 96 dB / 20 W; Vintage 30 100 dB / 60 W; G12M-65 Creamback 97 dB / 65 W; G12H-75 100 dB / 75 W; Alnico Blue 100 dB / 15 W. All five share a 75 Hz resonance and a 5,000 Hz top.
Sounds like: the real published differences being loudness and how hard it can be driven before it compresses — not frequency range
Open back vs closed backSpeaker
Front and rear waves interfere below the dipole peak, rolling off at about 6 dB/octave. Sealing the back raises the driver's resonance by roughly two and a half semitones.
Sounds like: open backs airy, omnidirectional and mid-punchy; closed backs directional, tighter and deeper
Multiple driversSpeaker
Coherent sources give +3 dB from doubled radiated power alone, and up to +6 dB where spacing is small relative to wavelength.
Sounds like: a four-by-twelve behaving like one very large speaker, with far more weight from the midrange down
Mic placementSpeaker
Moving an SM57 from dust cap to cone edge measures about 5 dB of change below 500 Hz, 8 dB at 1.6 kHz and 6 dB at 3.5 kHz.
Sounds like: an enormous tonal move — brighter at the centre, fuller and darker at the edge — from a few centimetres
Impulse responseIRSpeaker
A linear time-invariant capture: frequency response, phase, inter-speaker cancellation, diffraction and the whole mic and converter chain, frozen into one filter. It categorically cannot capture distortion, dynamic compression or cone breakup.
Sounds like: an exact snapshot of one cabinet at one volume, which never changes no matter how hard you play
Reactive vs resistive loadSpeaker
A reactive load presents a speaker-like varying impedance so the output stage behaves normally; a resistive load presents a flat impedance and, as attenuation rises, makes the tone more compressed and dark.
Sounds like: resistive loads recording far too bright and buzzy; reactive loads keeping the amp's own voice
Attenuator limitsSpeaker
Even a good reactive load cannot reproduce excursion-dependent breakup or thermal compression, because the cone is not actually moving that much air.
Sounds like: the electrical half of loud without the mechanical half
Power scalingSpeaker
Reduces the plate supply voltage delivered to the power tubes before the output transformer, shrinking the whole operating envelope rather than dissipating output as heat.
Sounds like: a small amp's behaviour from a big amp, rather than a big amp turned down
Impedance mismatchSpeaker
Mismatches within double or half the rated impedance are generally considered safe; beyond that, too high a load stresses the transformer and tubes with excess voltage and too low pulls excess current.
Sounds like: a slightly looser or slightly stiffer feel, until it stops being a tone question and becomes a repair
Break-inSpeaker
Klippel measured about 30% total suspension stiffness loss on a complete driver, roughly 85% of it break-in, implying about a 16% drop in resonant frequency. Real and permanent — and concentrated at low frequencies, not in the 1–5 kHz breakup region that gives a guitar speaker its voice.
Sounds like: a modest loosening of the low end, not the transformation the folklore promises
BufferRouting
A low output impedance of a few hundred ohms pushes the downstream RC corner far above the audible range. It does not add treble; it moves the filter out of the way.
Sounds like: the guitar sounding the same through fifty feet of cable as through five
Bridging ratioRouting
A passive pickup's output impedance is roughly 5–20 kΩ and rises with frequency; a guitar amp's input is about 1 MΩ. That deliberate ten-to-one-plus ratio exists solely to avoid loading the instrument — which is why a mixer line input at 10–50 kΩ is, per Sound On Sound, “way too low to allow a guitar's pickups to work properly.”
Sounds like: a guitar plugged into the wrong input sounding thin, dull and lifeless before anything else happens
True bypassRouting
The pedal is switched entirely out of circuit — but its jack capacitance stays and nothing terminates the cable. DC offset above 5 mV across the switched node causes an audible pop; by 20 mV it is intolerable.
Sounds like: no colouration when off, no trails, and a click on every stomp unless the pull-downs are right
Effects loopRouting
Placed after the preamp distortion, so delay and reverb act on a distorted signal rather than being distorted themselves.
Sounds like: repeats that stay clear behind a saturated tone instead of turning to porridge
Four cable method4CMRouting
Drives in front of the preamp, time effects in the amp's loop, using the unit's own internal split. Four cable runs, and it demands good buffering.
Sounds like: one amp doing what three normally would, with every effect in its correct position
True stereo vs ping-pongRouting
True stereo runs independent, decorrelated processing per side; ping-pong bounces one signal between two outputs. Both get marketed as stereo.
Sounds like: genuine width in one case, and a novelty bouncing between speakers in the other
What is actually stereoRouting
Modulation and time-based effects genuinely produce different left and right signals. Distortion, compression and most gain stages are inherently mono processes even inside a box labelled stereo.
Sounds like: width that only ever comes from the modulation and the ambience, never from the drive
Mono-summing cancellationRouting
Modulation deliberately produces out-of-phase content between channels. Summed to mono the effect level drops, and in the documented worst case a stereo tremolo running 180° out of phase can completely erase itself and sound like bypass.
Sounds like: the effect vanishing the moment the front-of-house engineer collapses the rig to mono
Split sumRouting
Duplicating one channel to both outputs rather than summing left plus right, which avoids the cancellation entirely for modulation, while a true sum stays fine for delay and reverb.
Sounds like: a stereo rig that survives mono without disappearing
Wet/dry/wetRouting
Dry to a centre amp for punch and mono reliability, effects split to two amps panned hard.
Sounds like: enormous width around a solid centre, that does not depend on the audience standing in one place
Buffered loop switchingRouting
Each pedal sits in its own buffered loop and is electrically removed when off, so it never loads the chain regardless of how many are installed.
Sounds like: a thirty-pedal rig with the same top end as a guitar straight into the amp
Order as functionthe Keen hierarchyRouting
R.G. Keen's Effects Order is a distortion-centric hierarchy rather than a list of boxes: amplitude effects, pre-distortion EQ, distortion, post-distortion EQ, short time-based effects, long ones, then reverb. Each rule has an electronic reason.
Sounds like: the difference between a chain that clarifies and one that muddies, with no change of pedals at all
Time before driveRouting
Phasers and delays create notches in the frequency response, and distortion placed after them fills in the response notches the time delay created. Delay first also means every repeat is re-distorted and progressively muddier.
Sounds like: modulation that disappears into the distortion, and repeats that get dirtier rather than quieter
Profiling vs modellingRouting
A profiler sends noise into the target amp, mics the cab and analyses the return, capturing a fingerprint of one amp at one setting through one cab and one mic. Component modelling simulates the circuit, so every control responds natively.
Sounds like: a profile that is uncannily right at its captured setting; a model that keeps behaving when you turn the knobs
Nothing matches that. Clear the box or the category to see everything again.
Keep exploring