Output Impedance Checker

An amplifier is not a perfect voltage source. It has a resistance of its own, in series with whatever you plug in, and that resistance and your headphones form a voltage divider. The rule of thumb is that the source impedance should be under one eighth of the headphone impedance. Below that line the interaction stays under about a decibel and you can forget about it. Above it, the source starts acting as a tone control.

What the number actually does

Two things, and only one of them is the one people worry about.

The first is level. Some of the voltage is dropped across the amplifier instead of across your headphones, so you lose a little loudness. At the one-eighth line that loss is 1.02 dB — inaudible, and you would turn the volume up without noticing. On its own it is not worth a page.

The second is frequency response, and it only appears because headphone impedance is not one number. The figure on the box is nominal; the real impedance moves with frequency, sharply around the bass resonance of a dynamic driver and all over the place in a multi-driver balanced-armature in-ear with a passive crossover. Wherever the impedance rises, the divider gives more voltage to the headphone, so that part of the response rises too. The source impedance turns the headphone's own impedance curve into an equaliser curve.

That is why the ratio matters rather than the raw ohms. If a headphone's impedance swings between a minimum and a maximum, here is how much of that swing ends up in the response:

At one eighth — the fourth row — even a fourfold impedance swing moves the response by about 0.8 dB. At one half it becomes 2.5 dB, which is more than most people's EQ presets. The full interaction, curve and all, is here.

Where the cliff is

This database holds 567 non-electrostatic headphones. Run the rule against all of them at once and the drop-off is not gradual:

One ohm is fine for 97% of everything here. Five ohms is fine for 28%. Nothing between those two figures is a small difference — the database's in-ears sit at a median of 18Ω, and once the source is at five ohms most of them are inside the range where the response moves.

What makers actually publish

169 of the 280 sources here publish an output impedance that can be traced to their own document. The median is 0.7Ω. 122 of them are at or below one ohm and 13 claim a flat zero, which is a rounding statement rather than a measurement. At the other end, 34 are at eight ohms or more and 16 at thirty-two or more — mostly audio interfaces and valve amplifiers, where a series resistor is a deliberate design choice rather than an oversight.

For reference, all of these come from the makers' own published figures: Benchmark HPA4 0Ω, Chord Mojo 2 0.06Ω, Rupert Neve RNHP 0.08Ω, Topping A90 0.1Ω, Schiit Magni Heresy 0.25Ω, JDS Labs Atom+ (2nd gen) 0.7Ω, FiiO K7 1Ω, Schiit Vali 2+ 1.8Ω, RME Babyface Pro FS 10Ω, Focusrite Scarlett 2i2 4th gen 50Ω. The Bottlehead Crack (OTL kit) at 120Ω is the highest figure in the database, and at that impedance the one-eighth rule is satisfied by none of the 567 models here — it would want a 960Ω headphone, and the highest impedance in this database is 600Ω. It is an output-transformerless valve kit designed around 300Ω headphones, and it sounds the way it does partly because of this number.

Why it is nearly always an audio interface

34 sources here publish an output impedance of eight ohms or more, and 28 of them are audio interfaces. That is not a coincidence and it is not sloppiness: a series resistor on a headphone output protects the amplifier from a shorted cable in a live room, keeps several pairs of headphones from loading each other on a shared bus, and costs nothing. In a studio full of 250Ω monitoring headphones it has almost no effect. Plug a 16Ω in-ear into the same socket and it is a tone control.

Focusrite's Scarlett Solo and 2i2 (4th generation) and seven Audient units all state fifty ohms; Tascam's US-2x2HR states sixty-six. Fifty ohms wants a 400Ω headphone to satisfy the rule, and six of the 567 models here are that high. None of these makers is hiding anything — they print the figure, which is more than 111 other sources in this database manage.

The figures that are not there

111 sources in this database have no output impedance at all. That is the single most useful number a headphone output can publish and it costs nothing to measure, so its absence is worth noticing rather than shrugging at.

On 29 of those the maker does print an ohm figure somewhere on the page — it just belongs to the line output, or it is the range of headphones the output is meant to drive, or nobody says which jack it describes. Each of those product pages quotes the figure and says what it is, so you can check the spec sheet against this site and see the same thing. The whole count is on the spec audit.

If you have a measured figure you trust for a source that does not publish one, the checker above takes it. This site will not record it as a specification, but the arithmetic does not care where the number came from.

Every published figure, lowest first

The ten above are a sample. Here is the whole set: all 169 of the 280 sources in this database whose own documentation states an output impedance for the headphone jack, in order. The right-hand column is the part no spec sheet gives you — the lowest headphone impedance that still satisfies the one-eighth rule at that figure, and how many of the 567 non-electrostatic models here that covers.

13 of them print a flat zero, which is a rounding statement rather than a measurement; 122 are at or below one ohm, and 55 are low enough to satisfy the rule on every model in the database. At the other end 34 sit at eight ohms or more, 28 of those being audio interfaces. The highest is the Bottlehead Crack (OTL kit) at 120 Ω, which would want a 960 Ω headphone and therefore satisfies the rule on none of them. Interfaces are not uniformly bad about this: the Merging Anubis publishes 0.035 Ω.

Every figure here is the maker’s own. The 169 entries are not a ranking of sound quality and a higher figure is not a fault — on 250 Ω studio headphones, fifty ohms is inaudible. It becomes a tone control the moment a multi-driver in-ear goes into the same socket, and that is the whole of what this column is for. The 111 sources that publish nothing at all are not on this list and nothing here implies they are good or bad; how many there are and what they print instead is its own page.