Measurements

To measure the t103z's performance, I used a G.R.A.S. Type RA0045 ear simulator, a Clio FW audio analyzer, a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface, and a Musical Fidelity V-Can headphone amplifier. Measurements are calibrated to ear entrance point (EEP), the point where your ear canal intersects with your earlobe. I used the standard, single-flange silicon tips in the medium size because these fit the ear simulator best. I inserted and reinserted each earpiece several times, and settled on a position for each that gave the most representative result.

I did most of the measurements of the t103z with the red filters, then tried measuring with the other filters for comparison. As you can see in the first frequency response plot, the response of the t103z is unlike any other we have measured. There's a sizeable upper bass/lower midrange bump centered at 300 Hz, a slight lower treble peak at 2 kHz, and a very narrow peak at 10 kHz. With a typical IEM, we'd see a broader bass bump between, say, 40 and 300 Hz, a strong peak at about 3 kHz, and perhaps another peak at 6 or 8 kHz. The response with the red filter is so far afield from what I've seen in the past that I wouldn't venture a prediction as to its psychoacoustic effect-I'll just have to let the panelists' subjective judgments stand on their own.

As you can see in the second frequency response plot, the effects of the filters are considerable. The response curve of the black (high-pass) filter is fairly similar to the red (flat) filter, with a little less bass response and +7 to +8 dB more treble response above 3 kHz. The response of the yellow (low-pass) filter is radically different, and much more similar to the measured response of a typical IEM; it's almost the same as the response of the red filter except that it has a much larger and lower bass boost that peaks at 60 Hz.

(Post-script: After this review was published, I had a chance to discuss these measurement results with a Torque Audio representative, who said that the results were essentially the same the company had gotten. He postulated that the G.R.A.S. coupler commonly used to measure in-ear monitors may be incapable of accurately measuring IEMs with ports. The red and black filters are both ported while the yellow one is not. Thus the disparity between the measured response of these filters-especially the red one-and our panelists' subjective impressions. I can't say without further investigation whether or not this is correct, but it's certainly a plausible explanation.)

Adding 70 ohms output impedance to the V-Can's 5-ohm output impedance to simulate the effects of using a typical low-quality headphone amp has no significant measureable effect.

Total harmonic distortion (THD) at 100 dBA is just a bit higher than we usually measure with IEMs, hitting 4% at 100 Hz and maxing out at 5.5% at 20 Hz. Isolation is typical for an IEM. With the single-flange medium tip, it's -10 to -29 dB above 1 kHz. Switching to the double-flange tip improved isolation by -5 to -10 dB above 3 kHz, but that result reflects the way the tips fit the G.R.A.S. coupler; results with your ears will likely vary.

Measured impedance is just about dead flat at 15 ohms. Average sensitivity with a 1 mW signal at 16 ohms rated impedance is 102.4 dB from 300 Hz to 6 kHz.

Bottom line

No matter which of the t103v's valves you prefer, there's no question that this is the most user-friendly application of the interchangeable filter concept to date. The t103v is a well-engineered, thoughtfully designed headphone in its own right, and the addition of the filters makes it vastly more versatile than almost any other IEM on the market today. We're eager to hear what the upcoming filter sets sound like, too.

Meanwhile, I think we can all agree with Will's summation of the t103z: "We music lovers deserve to have stuff like this."