Amplifier Matcher

Pick your headphones in the calculator above and the three fields below fill themselves in. Every source is then judged on three things at once, not just power: whether it can drive your headphones, whether its output impedance is low enough to leave their frequency response alone, and whether its own noise floor will be audible as hiss.

That last pair is why the answer differs so much between a full-size headphone and an IEM. Full-size cans mostly need power. Sensitive IEMs almost never do — for them, a source that is quiet and low-impedance matters far more than a source that is strong, and plenty of powerful sources fail at exactly that.

Filled in from the calculator above.
Used to predict audible hiss.

Enter your headphone specs above to see matched amplifiers.

The whole curve, not just the top of it

The four products above are one point on a curve. Here is the curve. These are the 70 of the 280 sources in this database that publish all three figures — a power rating at a stated load, an output impedance and a noise figure — ranked by how many of the 567 non-electrostatic models they suit on all three axes at once. That is the same test the tool above runs, computed in advance against the whole catalogue.

The 210 sources that are not here are not here because something is missing from their documentation, not because they failed anything. That is the single most important sentence beside this table, and it is why the count is 70 and not 280.

The median of the 70 suits 546 of the 567 models; the range runs from 567 down to 32. The cheapest product on the list is the JadeAudio JA11 at $20, which suits 545 of them. 54 of the 70 clear 500 models, and the cheapest of those is the same product. At the bottom, the Schiit Valhalla 3 costs $499 and suits 32 — a reminder that price and this number are only loosely related, because the three axes pull in different directions and a design can be built to win one of them.

SourceTypePrice Models it suits on all three Output impedanceNoise
Topping L30 IIdesktop$149567 of 5670.1 Ω0.3 µV
Topping A30 Prodesktop$249567 of 5670.1 Ω0.3 µV
Topping A90desktop$499567 of 5670.1 Ω0.2 µV
Topping A90 Discretedesktop$599567 of 5670.1 Ω0.2 µV
SMSL HO100desktop$160566 of 567<0.1 Ω1.3 µV
SMSL HO200desktop$399566 of 567<0.1 Ω1 µV
RME ADI-2/4 Pro SEdesktop$2499566 of 5670.1 Ω1.09 µV
Schiit Magni Unitydesktop$119565 of 5670.1 Ω1.59 µV
Schiit Midgarddesktop$219565 of 5670.1 Ω1.59 µV
iBasso DX260 MK2player$949565 of 5670.6 Ω1 µV
Chord Hugo TT 2desktop$5695565 of 5670.042 Ω1.7 µV
FiiO BTR13Bluetooth$59564 of 5670.6 Ω1 µV
Soncoz SGA1desktop$600564 of 5670.1 Ω1.8 µV
FiiO KA3dongle$79563 of 5670.5 Ω1.7 µV
FiiO KA5dongle$99563 of 5670.3 Ω1.6 µV
SMSL SP400desktop$529563 of 567<0.1 Ω1.9 µV
FiiO KA11dongle$29561 of 5670.7 Ω1.7 µV
JDS Labs Atom Amp 2desktop$129561 of 5670.7 Ω1.33 µV
iBasso DC03 Prodongle$69559 of 5670.12 Ω0.9 µV
RME ADI-2 DAC FSinterface$1399559 of 5670.1 Ω2.18 µV
RME ADI-2 Pro EXinterface$2499559 of 5670.1 Ω2.18 µV
JDS Labs Atom+ (2nd gen)desktop$99558 of 5670.7 Ω1.99 µV
Benchmark HPA4desktop$3499557 of 567<0.1 Ω2.45 µV
Chord Hugo 2portable$2595556 of 5670.025 Ω2.6 µV
FiiO KA15dongle$109553 of 5670.8 Ω1.7 µV
JDS Labs Element III MK2desktop$449553 of 5670.7 Ω2.51 µV
Chord Mojo 2portable$650553 of 5670.06 Ω2.7 µV
Schiit Vali 3desktop$149550 of 5670.5 Ω2.83 µV
FiiO JM21player$199550 of 5671 Ω1.6 µV
Shanling H7portable$829550 of 5671 Ω1.5 µV
Shanling M7player$1249550 of 5671 Ω1.6 µV
FiiO M23player$699548 of 5671 Ω1.9 µV
JDS Labs Element IVdesktop$549546 of 5671 Ω2 µV
FiiO M21player$599546 of 5671 Ω2 µV
FiiO K9 Prodesktop$799546 of 5671 Ω2 µV
JadeAudio JA11dongle$20545 of 5670.5 Ω1.7 µV
FiiO BTR15Bluetooth$109545 of 5671 Ω2 µV
FiiO KA17dongle$149545 of 5671 Ω2 µV
FiiO BTR5 2021Bluetooth$109544 of 5671 Ω2 µV
FiiO K19desktop$1430540 of 5670.6 Ω3.5 µV
FiiO BTR17Bluetooth$199536 of 5671 Ω3 µV
FiiO K7desktop$199534 of 5671 Ω3.1 µV
FiiO Q15portable$440534 of 5671 Ω3.1 µV
HiBy R8 IIplayer$1999531 of 5671 Ω3.2 µV
Schiit Lyr 5desktop$799530 of 5670.4 Ω3.99 µV
FiiO M27player$1799530 of 5670.5 Ω3.9 µV
FiiO M17player$1799528 of 5671 Ω3.5 µV
FiiO K11desktop$143516 of 5671.2 Ω2.8 µV
Singxer SA-1desktop$699516 of 5671.4 Ω0.3 µV
FiiO Q7portable$750516 of 5671.2 Ω2.8 µV
FiiO BTR7Bluetooth$199515 of 5671.2 Ω3 µV
FiiO M11Splayer$499514 of 5671.5 Ω1.8 µV
Grace Design m900desktop$660506 of 5670.08 Ω5.01 µV
FiiO M15Splayer$799501 of 5671.4 Ω3.2 µV
FiiO K17desktop$990495 of 5671 Ω5 µV
SPL Phonitor xedesktop$3000492 of 5670.18 Ω5.48 µV
Violectric V550 Prodesktop$3600492 of 5670.15 Ω5.48 µV
SPL Phonitor sedesktop$1600491 of 5670.18 Ω5.48 µV
SPL Phonitor xdesktop$3200491 of 5670.18 Ω5.48 µV
FiiO KA13dongle$79488 of 5671.7 Ω2.5 µV
FiiO M33 R2Rplayer$730482 of 5670.8 Ω5.5 µV
Violectric V222desktop$1245469 of 5670.25 Ω6.15 µV
Violectric V226desktop$1600469 of 5670.25 Ω6.15 µV
Lake People G103-S Mk IIdesktop$350468 of 5670.1 Ω6.15 µV
Lake People G111 Mk IIdesktop$629451 of 5670.125 Ω6.9 µV
FiiO K11 R2Rdesktop$176427 of 5671 Ω7.2 µV
Rupert Neve RNHPdesktop$699417 of 5670.08 Ω8.03 µV
FiiO BTR11Bluetooth$36400 of 5670.5 Ω8 µV
Schiit Mjolnir 3desktop$1199181 of 5670.13 Ω25.18 µV
Schiit Valhalla 3desktop$49932 of 56725 Ω6.32 µV

A low number in the fourth column is not a bad product. A valve amplifier with a high output impedance is doing on purpose what the one-eighth rule penalises, and it was never meant for a 16 Ω in-ear; a desktop amplifier with an audible noise floor in in-ears may be flawless with the 300 Ω headphones it was designed around. What the column measures is breadth — how much of a catalogue of 567 very different headphones one product can cover — and breadth is not quality. It is also measured entirely on figures the makers chose to print, which is the only ground this site will stand on.

Why three axes and not one

Every other tool in this hobby matches an amplifier to a headphone on power. Run that over this database and the answer is almost always yes, which is what makes it a useless question. The weakest source in this catalogue — the Steinberg UR22C — still reaches 100 dB on 512 of the 567 models here. The median source reaches 565 of them. The Apple USB-C dongle (US), which costs $9, clears 547 of them.

85 of the 280 sources here have enough power for every non-electrostatic model in the database, and the cheapest of them costs $99 (Schiit Magni Heresy). If power were the question, this page could be one sentence long and the answer would be “buy the hundred-dollar one”.

Output impedance is what actually disqualifies

169 sources publish an output impedance. Apply the one-eighth rule to the headphones each of them can already power, and only 72 lose nothing at all. The rest lose some of the catalogue, and a few lose nearly all of it:

SourceOutput impedanceModels it powers but should not driveModels in the database
Bottlehead Crack (OTL kit)120Ω565567
Audient iD4850Ω560567
Audient ORIA50Ω560567
Audient iD44 MkII50Ω559567
Audient iD2450Ω555567

These are not bad products, and the numbers above are not a verdict on how they sound. The one-eighth rule is a threshold, not a cliff: a source with fifty ohms on the jack only satisfies it above 400Ω, which almost nothing in this database reaches, so the count runs high by construction. What the column does say is that these are not the jacks to plug a low-impedance multi-driver in-ear into, and that no amount of power changes that — which is the whole reason this axis exists separately from the power one.

And noise decides the in-ears

83 sources publish an output noise figure. Of those, 5 are quiet enough to stay under 20 dB SPL in all 222 in-ears here. The quietest is the Singxer SA-1 at 0.3 µV; the highest published figure is Schiit Mjolnir 3 at 25.18 µV, which crosses the line in 204 of them. The hiss page works this out for any pair.

4 products clear all three — and that is a fact about publishing

Ask for a source with power for the whole database, an output impedance low enough for every model in it, and a noise floor quiet enough for every in-ear, and this catalogue returns 4 products: the Topping L30 II at $149, the Topping A30 Pro at $249, the Topping A90 at $499, the Topping A90 Discrete at $599.

Read that carefully, because the obvious reading is wrong. It does not mean those 4 products are good and the other 276 are not. It means 4 products publish enough to prove it. The requirement is not excellence, it is three numbers on a page, and most makers do not print all three. That is the real finding here, and it is the same one the rest of this site keeps arriving at.

One page per source

Every source in this matcher also has its own page, with power into four loads, noise, output impedance and a full headphone list. Browse all 280.

↑