Headphones you can compare on an axis
Every model ships with its measured frequency response, taken on a rig we describe in full — so "warm" and "detailed" are things you can check rather than things we say.
We believe in measuring what we sell
Two decisions, and what each one costs.
A published response curve
Measured on a 711 coupler, five reseats averaged, raw and compensated curves both included. It is the only way two headphones can be compared without hearing them side by side.
Everything replaceable
Pads, cable, headband and driver, sold separately and fitted with a screwdriver. It makes the housing slightly thicker than a bonded one, which is a trade we make on purpose.
The range
Four models. The differences are stated as measurements, not as tiers.
OB-2
Open back
- ✓Open back, 32Ω, 104dB/mW
- ✓Wide stage, minimal isolation
- ✓Velour and hybrid pads included
- ✓1.5m and 3m cables
- ✓Not for offices or trains
CL-1
Closed back
- ✓Closed back, 38Ω, 101dB/mW
- ✓−18dB isolation at 1kHz
- ✓Leather and hybrid pads included
- ✓1.5m and 3m cables
- ✓Narrower stage than the OB-2
ST-4
Studio reference
- ✓Closed back, 60Ω, 99dB/mW
- ✓Flattest response we make, ±2.5dB
- ✓Individual measurement card in the box
- ✓Detachable balanced and single-ended cables
- ✓Benefits from an amplifier
IE-1
In ear
- ✓Single dynamic driver, 16Ω, 108dB/mW
- ✓−22dB isolation at 1kHz
- ✓Four silicone and two foam tip sizes
- ✓1.2m cable, replaceable at the connector
- ✓No microphone; this is not a headset
The measurement bench
Three things about the rig, because a curve is only worth what the method behind it is worth.
A 711 coupler, five reseats
Every curve is the average of five reseats on an IEC 60318-4 coupler. A single seating can move the treble by 4dB, which is more than the difference between two models.
Raw and compensated, both
We publish the uncompensated trace alongside the Harman-compensated one. Only publishing the compensated curve hides the choice of target, which is where most of the argument is.
Unit-to-unit tolerance stated
±2.5dB from 20Hz to 10kHz between any two units off the line. A tolerance nobody quotes is a tolerance nobody measured.
Hear the difference, feel the build
Three things you notice in the first week and keep noticing.
Simplicity of fit
One headband adjustment, no app, no pairing dance. The clamp is set at the factory to 4.2N and stated on the box.
Sound you can place
A wide stage on the open backs and a narrow one on the closed, said plainly rather than sold as better.
Built to be opened
Four screws and a Phillips head. Every part inside has its own order code printed beside it.
What owning one is actually like
Five steps, none of which needs an app.
Plug in
A 3.5mm cable in the box and a 6.35mm adapter. No pairing, no firmware, nothing to update at the wrong moment.
Check it against the card
The measurement card in the box is that unit, not the model average. Compare it to the published curve if you want to.
Swap the pads when they flatten
Around eighteen months of daily use. ₹400 a pair, two minutes, and the tonality shift from worn pads is the thing most people mistake for the driver ageing.
Replace a cable, not a headphone
The connector is a standard 2.5mm mono jack at each cup. Any third-party cable fits, and we will tell you which ones we have tested.
Sell it with the warranty attached
Two years, transferable, because these get resold and a warranty that dies on resale quietly lowers what yours is worth.