I Simulated 29 Jazz Bass Pickups So I Wouldn't Have to Buy Them All

Pickups cost €80-150 each. You need two. You can’t return them soldered. So I built a simulator: Python, measured pickup data, the actual circuit the signal passes through. Twenty-nine pickups on the same axes. No vibes, no forum opinions. Just math.

Reading the curves

All curves normalized to 0dB at 200Hz. The differences live above 1kHz.

The resonant peak. The bump in the upper mids. Where it sits defines the pickup’s character. Higher and later = brighter, more string attack. Lower and earlier = thicker, warmer.

Rolloff. How fast the curve drops after the peak. Steep = warm and filtered. Gentle = open and airy. All pickups roll off monotonically after. They have to, it’s a passive circuit.

One dimension of tone. Says nothing about dynamics, touch response, or feel.

The full landscape

Architecture matters, but less than I expected. The extremes hold: brightest are single-coils, darkest are stacked. But in the middle (where the buying decisions actually live), it’s a mix. Some split-coils are brighter than some single-coils.

Single-coil. Widest variation. The lightest-wound vintage specs (CS ’60s, V-Mod II) sit bright. The overwound models (SJB-3, SJB-2) are darker than most hum-cancelling pickups. The Fender Pure Vintage ‘74 (full GuitarNutz2 measurements: L=3.65H avg, C=140pF) lands mid-pack despite its reputation. The “70s brightness” comes from 70s bridge pickup placement (10mm closer to the bridge), not the electrical response. But they all hum when soloed. In a studio, that 60Hz is always there.

Split-coil (side-by-side). Each coil senses two strings, wired in hum-cancelling series. Huge spread: EMG JV at -1.4dB to Bartolini 9J at -7.2dB. The Bartolini 9J — ceramic, steel blades — sits dark despite the lowest DCR in the group. High-permeability cores push inductance far above what resistance suggests. Inductance drives the ranking more than the architecture label.

Stacked humbucker. The Wilde J-45N (Bill Lawrence coaxial “Noisefree” design) has the lowest Q in the sim (8.2) — its 20kΩ DCR damps the resonance so hard there is no peak at all, just flat upper mids into rolloff. Stacking two coils vertically raises effective inductance and capacitance, which is why the SJB-5 and Ultra Noiseless come out the darkest hum-cancelling pickups in the sim.

The stock ceramic proof

The Fender Stock Ceramic (MIM/Squier) measures only 5.74kΩ DCR, the lowest in the entire lineup. But its steel slug pole pieces push inductance to 4.08H, higher than the Alnico CS ’60s at 2.66H. It’s darker than pickups with twice its resistance.

DCR doesn’t predict tone. Inductance does. Steel slugs are cheap and high-permeability; Alnico poles are expensive and low-permeability. The spec sheet says “5.7kΩ” and you think “bright.” The bode plot says otherwise.

DCR vs brightness

If DC resistance determined tone, all the dots would sit on one diagonal line. They don’t.

The Stock Ceramic (5.74kΩ, measured 4.08H) is darker than the Lollar (8.65kΩ, measured 4.19H). Three kilohms less resistance, darker output. Steel slugs vs Alnico poles. The SJB-3 Quarter Pound (13.6kΩ, measured 6.72H) is the darkest in the sim, its quarter-inch Alnico 5 poles and massive capacitance (244pF measured) pulling the resonant peak down to 3.85kHz.

L and C do the work. Resistance is just the easiest number to print on the box.

Measured data as anchor points. The seven measured pickups (solid borders) calibrate the estimates. If a measured single-coil at 3.65H sits at -4.1dB, an estimated one at 3.8H should sit around -4.5dB.

The Lollar question

Three favourites always come up: Fralin, Nordstrand NJ4, Lollar. The Fralins are split-coil (already in the database). The Nordstrand has no published inductance. Pure estimate.

Lollar publishes inductance for their entire lineup (the only major pickup maker that does). The Echoes of Mars database has independently measured the bridge: R=9.69kΩ, L=4.881H, C=131pF. Significantly higher than Lollar’s published spec (R=8.5kΩ, L=4.0H). Sample variance, or a 2014-era spec. Using the average of measured bridge and published neck values: L=4.19H, C=131pF, 5kHz level = -5.6dB.

The Lollar lands in “warm vintage” territory. Almost identical to the Fender PV ‘74 in character. Both sit about 5dB darker than the CS ’60s at 5kHz. The PV ‘74 gets there with more inductance from heavier winding; the Lollar through both winding and higher capacitance.

The BYO Lightning ($60/set) has vendor-published specs almost identical to the Lollar published specs. Same L, same estimated C. The difference is build consistency, magnet quality, and winding precision. The sim can’t see any of it.

The full ranking

The hypothesis: brighter pickups capture more harmonic information for studio DI. You can EQ darker. You can’t recover what rolled off before the converter.

Ranked by 5kHz level. Not a quality ranking: a brightness ranking viewed through that hypothesis. Measured = full RLC data. Adjacent positions within 1dB are within margin of error.

Top two are single-coils with the lowest inductance. Least wire on the bobbin. The CS ’60s has full measured data and confirms the model: measured resonant peaks (3.51kHz bridge, 3.81kHz neck with 470pF load) track the sim within 5%.

The first hum-cancelling pickup at #3 (EMG JV). Its inductance is a rough estimate. DiMarzio’s lineup shows the range within one manufacturer:

Both are split-coil side-by-side designs. DiMarzio tech support confirmed: “All of our Jazz Bass pickups are split humbuckers.” The Area guitar pickups are stacked; the J-bass version is not.

The Ultra Jazz has the highest DCR but also the highest treble rating. DiMarzio’s own copy: “even with a relatively high DC resistance figure, the Ultra Jazz has very strong, percussive highs.” Its hybrid magnet (Alnico 5 rods + ceramic bar) reduces inductance per ohm. The Area J runs Alnico 2 and sits darker at a 4.5 treble rating. Magnet type, not architecture, sorts this family: Alnico 2 versus hybrid.

The Bartolini 9J at #23 is the ceramic/steel proof case. Only 6.4kΩ — the lowest DCR of any hum-cancelling pickup — but -7.2dB at 5kHz, darker than the Fralin at nearly double the resistance. Ceramic + steel blades push L/R to ~0.69 H/kΩ, giving 4.4H despite low winding resistance. The Stock Ceramic at #20: same story, single-coil side. 5.7kΩ (lowest in the table) but -6.0dB. Both disprove “low ohms = bright.” Ceramic + steel = high inductance per ohm.

The three darkest are all over 6H. At that inductance, rolloff starts in the upper mids. Designed for a specific sound, not maximum capture.

Same pickups, different circuit

Cable capacitance loads the pickup node directly, in parallel, regardless of pot value. A 500K volume pot doesn’t shield the pickup from the cable. For the Ultra Noiseless at 5.2H, the difference between cable and no cable is 13dB at 5kHz — though on a bass, that 5kHz lives far above the strings. The Volume·Q·Blend harness — 500K volume, no tone pot, a Q-filter in place of the tone control — changes the loading, but the upgrade happens in two distinct steps.

Left: stock circuit. Centre: passive optimized (500K vol, no tone pot, cable still present). Right: buffered (same pots, the ring-hot buffer driving the cable from a low impedance). The pots alone give a modest, uneven improvement. The buffer is where the resonant peak comes back — though on a bass that peak sits above where the instrument plays, so the practical win is impedance immunity, not sparkle.

What the pots buy you

The cable is still there, 500pF at the pickup node regardless. For high-inductance pickups, that’s the dominant load, and pots don’t touch it.

The passive optimized spread is actually wider than stock (12.8dB vs 11.4dB). The low-inductance pickups benefit more from the pot upgrade proportionally, while the high-inductance ones are still crushed by cable loading. The Ultra Noiseless gains only +0.5dB from the pot upgrade. The stock Fender gains +1.6dB. The pots help the pickups that needed help least.

The passive optimized ranking

Same 29 pickups through the passive optimized circuit: 500K volume (single-coils get 250K), no tone pot, 500pF cable still present. The ranking barely changes from stock.

The buffer

A phantom-powered ring-hot buffer drives the cable from a low impedance. Tip stays the passive bass, always; the ring carries a buffered, powered copy — a compound 2N5457 → 2N3904 follower on half-rail bias, fed 18V in the studio, 9V for jamming. The cable capacitance no longer loads the pickup.

The gold bars are the pot upgrade (stock to passive). The cyan bars are the buffer (passive to buffered). For single-coils, the pot upgrade is the bigger share: they were never heavily loaded by cable. For stacked humbuckers, the buffer dominates. The SJB-5 Stack gains +0.1dB from pots, then +15.2dB from the buffer. The Ultra Noiseless: +0.5dB from pots, +15.2dB from the buffer. Their distributed capacitance was resonating with the cable, double-punishing them. Remove the cable and the peak comes back.

It lands at 5kHz, two octaves above the highest note on the instrument. The Fralin Split Jazz gains +1.0dB from pots, then +10.7dB from the buffer — all of it in a band the bass never excites. Its high inductance (4.8H) made it the most cable-sensitive of the four shortlisted hum-cancelling pickups, which is why the buffer matters: not because it adds treble to a note, but because it makes the pickup behave the same regardless of what cable hangs off it.

The buffered ranking

Run the same 29 through the buffered circuit — 500K volume (single-coils get 250K), no tone pot, the ring-hot buffer driving the cable so the pickup never sees it.

The Bartolini 9J: #23 stock, #23 passive, #1 buffered. It takes the top spot only when the cable is gone. Its ceramic + steel blades and low DCR gave it the highest inductance-to-resistance ratio in the database. Pot upgrades alone gained it +0.9dB. The buffer gained it +15.1dB. The Fralin Split Jazz: #22 stock, #22 passive — the pot upgrade doesn’t move it — then #6 buffered. Every hum-cancelling pickup that was penalised by cable loading sits in the top half after the buffer, not after the pot upgrade.

The single-coils compress into a tight cluster around +2 to +3dB with the buffer. They were never that far apart. The pot upgrade already closed most of the gap. The overwound SJB-2 and SJB-3 still sit last. At 6.7H+, they’re rolling off in the upper mids regardless of loading.

The thickness shift

Spectral balance tells a different story. In the stock circuit, the high-inductance pickups (Fralin, Bartolini 9J) sit thick. The cable was flattening their peaks, spreading energy across the low-mids. Remove the cable (buffered circuit) and every pickup shifts thinner. The Bartolini 9J drops from -0.6 to -4.8. The Fralin from 0.0 to -2.7. Their warmth was borrowed from cable loading.

The buffered circuit changes the perceived character: pickups that sounded warm through cable-loaded stock circuits read leaner and more focused with the cable gone. On a bass most of that shift sits above where the strings put energy, so what you actually get is consistency — the pickup sounds the same every day, into any cable. EQ adds warmth. The treble was never there to come back.

Why some pickups sound thicker

The ranking reshuffled. But numbers at 5kHz don’t tell you what a pickup sounds like in a mix. The PV74 sits nicely either side of the critical midrange guitar/vocal region, its treble peak giving useful transient definition to a solid, even bottom. A ceramic/steel pickup like the Bartolini 9J does the opposite: the peak sits lower, the midrange clanks, and it reads as aggressive rather than dark — more string energy than snap.

That’s not bright vs dark. That’s perceived thickness. Audible even at matched output levels. If the response below resonance is flat for all pickups (it is; the RLC model predicts this), where does the thickness come from?

The resonant peak’s sharpness determines how the pickup’s energy is distributed across the spectrum.

Same bulb wattage. A tight beam concentrates all the energy in one spot, and everything outside the cone goes dark.

Q factor is the beam width. A high-Q pickup concentrates its resonant energy in a narrow band. Everything outside that peak goes quieter. A low-Q pickup spreads it. The bass hasn’t changed, the string energy is the same. Q just changes what your ear notices.

The Bartolini 9J runs a tall, narrow peak — ceramic, steel blades, energy concentrated in a tight band. That should make it sound thinner, and in the spectral balance, it does. But its peak sits so low (4.4H pushes resonance down into the upper mids) that the concentrated energy lands where your ear reads it as aggressive midrange presence, not brightness. That’s the “clank.”

The PV74 at Q=21.0 is still sharp but broader. The peak sits higher, giving “snap” (attack transients) while the low-mids sit undisturbed. Ultra Jazz at Q=14.5 spreads the most: broad, even, no single frequency dominates.

To quantify this, the sim computes spectral balance: energy in the 80-400Hz band vs the 1.5-6kHz peak region. Closer to zero = thicker. More negative = thinner:

The overwound pickups (SJB-2, SJB-3, SJB-5) cluster top-left: low Q but thick, because their peaks are so low in frequency they don’t steal energy from the fundamentals. The vintage single-coils: moderate Q, thin. The ceramic/steel pickups sit far right, high Q, but their peaks land in the mids, not the treble, so they read balanced rather than thin. The Fralin is the thickest of the shortlisted hum-cancelling pickups (balance ≈ 0).

One pickup, the whole database

Stock: #24. Passive optimized: #24. Buffered: #2. The pots bought it +0.5dB. The buffer bought it +15.2dB. At +7.3dB through the buffered circuit, it’s the second-brightest hum-cancelling pickup in the database (behind the Bartolini 9J). Measured RLC, silent in every position.

With the buffer in place, a small cap between hot and ground at the blend pot input moves the resonant peak — down in frequency as you add capacitance, without touching the coil. The range in the chart below requires the buffer; without it, the Ultra Noiseless is stuck at -7.9dB regardless of cap value. This is the seed of the Q-filter in the build: make that cap-and-resistor a knob and you dial the peak instead of soldering it.

Buffered baseline: +7.3dB at 5kHz. Add 100pF and it drops into CS ’60s / Fralin territory (+4.3). Add 200pF and it’s flat at the top (−0.3). Add 500pF and it’s darker than everything in the stock ranking except the SJB-5 Stack (−11.6), Jazz Bomb (−11.6), SJB-2 (−10.8), Big J-Blade (−9.0), and SJB-3 (−11.9) — and brighter than its own stock value (−8.4). 15dB of range. The entire database on one set of coils.

One caveat the chart won’t tell you: all of this lives above ~1kHz, and a bass barely reaches it. Moving a peak that sits in the dead zone above the strings changes the harmonics and the noise floor, not the fundamental. The audible payoff comes from boosting those harmonics deliberately — which is what the Q-filter does, where the bass actually plays.

Methodology

Every pickup is an RLC circuit. Second-order resonant lowpass: flat through bass and mids, a peak in the upper harmonics, then monotonic rolloff. The simulator models the full signal chain: pickup, blend pot, volume pot, tone circuit, cable capacitance, load. Same circuit for all 29 pickups. All charts: 250K pots, tone wide open, 500pF cable (typical 3m), 1MΩ load.

Load: 1MΩ. This matches the UCX II Hi-Z input impedance (1MΩ). With 250K–500K pots in the signal path, the interface impedance is negligible. The pots dominate the pickup’s load. A 1MΩ load in parallel with a 500K pot is effectively 333kΩ, but the pot is already doing the loading. Cable capacitance loads the pickup node directly, in parallel, independent of pot value. The three-tier circuit comparison (stock / passive optimized / buffered) isolates this effect. The sim load matches the real interface.

Data tiers. Seven pickups have confirmed R and L values: five from independent Echoes of Mars / GuitarNutz2 measurements (CS ’60s, PV74, Stock Ceramic, Lollar, SJB-3), two from Fender and manufacturer data (Cobalt Chrome, Gen 4 Noiseless). Wilde J-45N R/L confirmed by Bill Lawrence directly. Three have vendor-published R and L. The remaining eighteen: R from spec, L estimated from architecture and L/R ratio (~0.41 H/kΩ for Alnico single-coils, higher for ceramic and stacked). Estimates carry ±15-20% uncertainty.

How published and measured diverge. Lollar’s published bridge: 8.5kΩ / 4.0H. GuitarNutz2 measurement of a 2014 bridge: 9.69kΩ / 4.881H. 14% higher R, 22% higher L. The sim uses measured values where available.

How reliable are the estimates? Alnico: ~0.41 H/kΩ (measured range 0.37–0.49). Ceramic + steel: ~0.69 H/kΩ, calibrated from the DiMarzio Chopper (0.668) and Stock Ceramic (0.710). Estimates could be ±0.5H off, enough to shift rankings by 1-2 positions, not enough to change architecture-level conclusions.

Spectral balance. Perceived “thickness” as energy ratio: 80-400Hz low-mid band vs 1.5-6kHz peak region. Captures what 5kHz brightness alone misses. A pickup can be “dark” but still sound thin depending on where its peak sits. The ceramic/steel pickups show it: a high-Q peak sitting low creates midrange aggression, not brightness.

What’s missing. String/body resonance, eddy current damping, magnetic field geometry, nonlinear effects, subjective character. The sim captures the passive electronics. Measurable isn’t everything.

I narrowed the field to four, then simulated every circuit variable around them. Four pickups, every circuit.