Why your DAC's output impedance matters more for headphones than its DAC chip

I had a listener email me last month convinced his new DAC/amp combo was "damaged" because his Grados sounded thin and his Audeze sounded fine on the same box. Nothing was broken. What he'd stumbled into, without knowing it, was output impedance mismatch — a spec that sits three lines down in the manual, usually in ohms, that most buyers never read and most reviews never mention. It's not glamorous. It doesn't photograph well for a product shot. But it changes what you hear more reliably than the DAC chip printed in bold on the box.

I spend most of my listening time with headphones on, not speakers in a room, so this is a spec I've learned to check before I check anything else. Chip generation, sample rate support, MQA decoding — none of that tells you whether a given DAC will actually sound right with a given pair of cans. Output impedance does.

What output impedance actually is

Every headphone output has some internal resistance sitting between the amplifying circuit and the jack. Call it the source impedance. Your headphones have their own impedance too, printed on the spec sheet — anywhere from around 16 ohms for some IEMs up past 300 ohms for old-school studio cans like the Beyerdynamic DT 1990 PRO. The relationship between those two numbers, not either one alone, decides what actually reaches your ears.

The rule of thumb, inherited from decades of professional audio practice, is that a source's output impedance should be no more than roughly one-eighth of the headphone's rated impedance — sometimes stated as a damping factor of 8:1. Fall well short of that and two things happen. Frequency response ripples, because headphone impedance isn't flat across the band, it swings with the driver's own resonances. And bass control loosens, because a high-impedance source doesn't grip a low-impedance driver the way a properly low-impedance one does.

Why this bites low-impedance headphones the hardest

Multi-driver IEMs are the worst-case scenario here, and it's not close. Many have impedance curves that swing wildly across the frequency range — dropping to single-digit ohms at some frequencies, climbing well above that at others — because of how the crossover networks are built between drivers. Pair one of those with a dongle DAC or headphone output that has, say, 5 or 10 ohms of output impedance, and you get audible frequency response shifts. Not subtle ones. I've measured differences of a couple of dB in places, which is enough to make a neutral IEM sound bright or bass-shy depending purely on what it's plugged into.

Full-size dynamic headphones with a flatter impedance curve are more forgiving. That's part of why a pair of high-impedance planars can sound remarkably consistent across wildly different sources, while a sensitive IEM sounds like a different product depending on the dongle. This is also why "it sounded amazing with the reviewer's rig" is one of the least reliable sentences in this hobby — the reviewer's source and your source may have nothing in common electrically, even if both are labelled "hi-res".

Something I mention in our guide to the best audiophile headphones and I'll say again here: impedance interaction is precisely why headphone auditions at a shop, using the shop's amp, tell you less than you'd think about how that headphone will behave on your own gear at home.

Where the industry actually gets this right

Desktop DAC/amps built with headphones as the primary job — rather than as an afterthought on a network streamer — tend to get output impedance right because the designers had no choice but to think about it. The RME ADI-2 DAC FS, for instance, publishes its output impedance figures openly and it's a genuinely low number across its various outputs, which is exactly why it copes so well with a wide spread of headphone loads from sensitive IEMs to high-impedance planars. Chord's Mojo 2 is another one that was clearly engineered with headphone loads front of mind rather than bolted on. Compare that to some all-in-one streaming amps and network players, where the headphone jack is very much a secondary feature and the output impedance spec often isn't published at all — which, in my experience, is usually a sign it isn't especially low.

I'd go further and say this is one of the more mildly contrarian positions I hold in this job: I think headphone output as a checkbox feature on streamers and integrated amps is mostly there to tick a spec sheet, not because the engineering effort went into making it good. If your headphone listening actually matters to you, a dedicated desktop DAC/amp built for the job — see our rundown of the best desktop DAC/amps — will almost always outperform the headphone jack on a lifestyle streaming amplifier, even one that measures beautifully on the speaker side.

How to actually check this before you buy

Manufacturer spec sheets vary wildly in how honestly they disclose output impedance. Some publish it clearly in ohms. Others bury it, round it favourably, or simply don't mention it at all — in which case, treat that omission as information in itself. If you can't find the figure anywhere, assume it's not a strength of the product.

A rough field test, if you own a genuinely low-impedance IEM and a higher-impedance full-size headphone: plug both into the same source at matched loudness and listen for tonal shifts, especially in the upper bass and lower midrange, where impedance curve dips are most common on multi-driver IEMs. If the IEM sounds noticeably thinner or brighter than it does on a source you know has low output impedance — a good dedicated headphone amp, for instance — that's your answer.

The Beyerdynamic DT 1990 PRO MK II and the Sennheiser HD 660S2 are both good reference points because their impedance curves are relatively well-behaved and documented; if something sounds off with either of those on a given source, the problem is very likely upstream, not in the headphone.

What this has nothing to do with

Output impedance is a completely separate conversation from bit depth, sample rate, or DAC chip generation, and I think a lot of buyers conflate the two because both live inside the same box. A DAC chip decodes digital audio into an analogue voltage. The output stage — which includes that impedance figure, along with output voltage and current delivery — is what actually drives your headphones from that voltage. You can have a technically excellent, measurement-topping DAC chip sitting behind an output stage that's mediocre for headphone loads. It happens more often than the marketing suggests, particularly on products where the headphone output is clearly a secondary feature bolted onto a streamer or preamp built primarily for speaker duty.

This is also distinct from the hi-res audio conversation I get asked about constantly — whether 24-bit/192kHz files actually sound better than 16-bit/44.1kHz, whether MQA did anything meaningful. Those are real questions worth having, but they're upstream of the issue I'm describing here. You can feed a headphone amp the most pristine hi-res file in existence and still get compromised sound if the output impedance is wrong for your headphones. The file format doesn't fix an electrical mismatch.

Balanced outputs don't automatically solve this

I get asked a lot whether going balanced fixes impedance problems. Sometimes, yes — a well-engineered balanced output often does have lower output impedance than its single-ended counterpart on the same device, partly because balanced circuits tend to get more design attention as a premium feature. But it's not automatic. A balanced output built cheaply, or one where the manufacturer focused engineering effort on the single-ended jack instead, can still have output impedance that's too high for a fussy IEM. Check the actual published figure for each output rather than assuming the connector type tells you the answer. I wrote more on the practical differences in our headphones buying guide, but the short version is: balanced is a connector and wiring scheme, not a guarantee of a better-engineered output stage.

What I'd actually do about it

If you own or are shopping for sensitive multi-driver IEMs, I'd treat low output impedance as close to a non-negotiable spec, ahead of sample rate support or DAC chip pedigree. Products like the FiiO K11 and the iFi Audio ZEN DAC 3 are worth a look partly because both were designed with a wide range of headphone loads in mind rather than as an afterthought bolted onto something else.

If your main headphones are higher-impedance dynamics or planars with well-behaved impedance curves — many Sennheiser and Audeze designs fall into this camp — you have more headroom to prioritise other things: output power for planars like the Audeze LCD-X, which need real current to sound their best, or simply which amp topology and tonal character you prefer. I've covered the power and impedance side of that pairing question in more detail in our desktop DAC/amp guide, and it's worth reading alongside this if you're building a rig from scratch.

Either way, the fix is cheap once you know what you're looking for. It doesn't cost more to buy a source with a genuinely low output impedance; it just requires reading past the sample-rate bullet points on the box, which — I'll be honest — is not what most of us are trained to do when we're standing in a shop with a credit card in hand.

A quick reality check on numbers you'll see quoted

You'll sometimes see output impedance figures thrown around casually in forum threads without context — "under 1 ohm" treated as some universal gold standard. It's a reasonable target, but the number only means anything relative to the headphone load you're actually running. A source with 2 ohms of output impedance will behave essentially identically to one with 0.5 ohms when driving a 300-ohm dynamic headphone; the 8:1 ratio is comfortably met either way. That same 2-ohm difference matters enormously more on an IEM that dips to 8 ohms somewhere in its impedance curve. Context is everything here, and I'd be sceptical of any spec sheet or forum post that quotes an impedance number as though it's a badge of quality in isolation, rather than a figure that only makes sense next to your actual headphones.

None of this makes the DAC chip irrelevant — conversion quality and output stage design both matter, and the best products get both right. But if you're forced to choose which spec to interrogate first when you're headphone-first like I am, skip past the chip name and go straight to the ohms. It's told me more about how a source will actually sound with my own collection than any decoder generation ever has.

— Eleanor Shaw, Headphones & Personal Audio Editor