Almost every headphone listing quotes a frequency range, and almost none of them quote it in a way that means anything. Meanwhile the measurement that would tell you how a headphone sounds, a frequency response graph, appears on very few product pages at all. Learning to read both is the difference between shopping on numbers and shopping on marketing.
What a frequency range claim leaves out
A published range says the headphone produces some output between two frequencies. It does not say how much output, or how evenly. A proper specification includes a tolerance, written as something like plus or minus 3 dB, because without one the claim is unfalsifiable. A driver producing a whisper at 40 kHz can honestly be quoted as reaching 40 kHz.
Human hearing tops out around 20 kHz in childhood and falls with age, so most of these numbers describe territory nobody can hear. CCA quotes 7 Hz to 40 kHz for the C12, and FiiO quotes 7 Hz to 40 kHz for the FT3. Neither figure is a lie and neither tells you anything about how the product sounds in the range you can actually hear.
The rows are often mangled as well. The HIFIMAN SUSVARA listing shows a frequency response of “20000 Hz” in one field and a range of 20 Hz to 20 kHz in another. The first is not a response, it is the top end of the second with the rest of the sentence lost. The Moondrop CHU II DSP shows the same field as 3.8E+4 Hz, which is scientific notation for 38,000 that survived a spreadsheet.
The ranges that do tell you something
A narrow published range is more informative than a wide one, because a manufacturer only narrows a claim when the product genuinely rolls off. Koss quotes 80 Hz to 18 kHz for the KPH7. That 80 Hz floor is a real statement: this is a lightweight unsealed on-ear design and it is not going to deliver deep bass. Compare the same maker’s Porta Pro at 15 Hz to 25 kHz and you can see which of the two is aiming lower.
Business headsets narrow the range on purpose. A voice-tuned headset stating 100 Hz to 10 kHz is putting its output where speech intelligibility lives rather than chasing a marketing figure. In a spec sheet full of 20 Hz to 40 kHz claims, a number like that is a design decision showing through.
Sensitivity and impedance, the two rows that matter
Sensitivity says how much sound pressure a headphone produces for a given input, usually quoted in dB per milliwatt or per volt. It is the single best predictor of whether your source can drive the headphone.
| Headphone | Stated sensitivity | What it implies |
|---|---|---|
| HIFIMAN SUSVARA | 83 dB | Needs serious amplification |
| TOZO HA1 | 88 dB | Powered internally, figure is academic |
| Sony WH-CH710N | 94 dB | Powered internally |
| Moondrop Joker | 106 dB | Comfortable from most sources |
| FiiO FT3 | 110 dB at 32 ohms | Easy to drive for a full-size design |
The TOZO HA1 and Sony WH-CH710N rows illustrate a common confusion: sensitivity is close to irrelevant on an active wireless headphone, because the internal amplifier is fixed and matched at the factory. It matters on passive wired designs, where you supply the power.
Impedance is the electrical load. Low impedance draws more current, high impedance needs more voltage. Most portable gear is happy between 16 and 64 ohms, and the FT3’s stated 32 ohms sits in the middle of that. A high impedance headphone on a phone will simply play quietly. Pair a demanding load with something from the amplifiers category rather than turning the volume up and hoping.
Reading an actual response graph
A frequency response graph plots level in dB against frequency on a logarithmic scale. Three habits make it readable.
- Check the vertical scale first. A graph drawn over 60 dB looks flat; the same data over 10 dB looks dramatic. Compressed scales are how a mediocre response is made to look tidy.
- Ignore everything above about 8 kHz. Measurements up there are dominated by the coupler and by where the headphone sat on the fixture, so the wiggles are not reliably what you would hear.
- Read the shape, not the exact values. Where the bass shelf ends, whether there is a dip in the presence region, whether the curve rises or falls overall.
Headphone graphs are also compensated. A headphone measured raw does not look flat even when it sounds neutral, because the ear itself adds resonances. Published curves are corrected against a target, and different targets produce different looking graphs from the same headphone. Always check which target a graph uses before comparing it with another one.
Why two graphs of the same headphone disagree
If you compare measurements from two sources, they will not match, and the reasons are mostly mechanical rather than anyone being careless.
Coupler design is the biggest one. A measurement rig models an ear canal, and different standards model it with different volumes and different simulated eardrum impedances. Change the coupler and the whole upper midrange moves.
Placement is the second. On an over-ear headphone, a few millimeters of position change alters the seal and shifts the bass by several decibels. On an in-ear, insertion depth does the same thing to the treble. Careful measurers average several repositionings for exactly this reason, and a single-seating measurement should be read as approximate.
Unit variation is the third and the least discussed. Two samples off the same production line are not identical, and channel matching within one pair is a real specification that few consumer makers publish. If a headphone sounds off-center to you and the source is fine, a channel imbalance is a plausible explanation rather than an imagined one.
One more spec deserves a mention because it is almost never quoted: total harmonic distortion. A headphone that measures well in frequency response can still distort at high levels, particularly a small in-ear driver asked to produce deep bass. When distortion figures are published at all, note the level they were measured at, because a figure at a quiet level says nothing about behavior at a loud one.
Translating shapes into sound signatures
The vocabulary used in reviews maps onto graph shapes fairly consistently. A raised region below 200 Hz reads as warm or bass-heavy. A dip between 1 kHz and 3 kHz reads as recessed or scooped, and is the usual meaning of a V-shaped signature. A rise between 5 kHz and 8 kHz reads as bright or detailed, and if it goes too far, as harsh or fatiguing. An even response with none of these reads as neutral or, less kindly, as boring.
Manufacturers signal the intended shape in their marketing when they do not publish a graph. Studio monitor designs claim neutrality because their job is to reveal a mix rather than flatter it, which is the pitch behind the closed-back Moondrop Joker and much of the over-ear headphones category. Consumer products describing deep or punchy bass are telling you where their bass shelf sits without drawing it.
The practical order of checks: sensitivity and impedance to see whether your source can drive it, published range only to spot a deliberate roll-off, then a response graph from an independent measurement if one exists. Treat a 20 Hz to 40 kHz claim with no tolerance as packaging copy, because that is what it is.
