Testing Noise-Cancelling Headphones

Posted by: Chris Martens at 2:02 pm, December 2nd, 2008

One of the most enjoyable and illuminating projects I’ve taken on of late has been testing a group of noise-cancelling headphones for Playback. The first batch of reviews (covering models from Audio-Technica, Creative Labs and Monster Cable) appears in the current issue of Playback (playback.avguide.com), while the second batch of reviews (covering models from Bose, Sennheiser and Sony) will appear in Playback 15, which is slated for release in early December. It occurred to me that you might enjoy learning how I tested the headphones and then, if you like, trying some listening experiments of your own.

At first blush, you might think it would be easy to test noise-cancelling phones. Just find a noisy environment, plug them in, and have at it, right? But upon reflection, I’d say “wrong,” because that approach—appealing though it is in terms of simplicity—isn’t readily repeatable and thus does not create a level playing field for purposes of comparing the headphones. What I needed, I reasoned, was a “test bed” that would meet several requirements. The test set-up needed to:

•    Be capable of producing high levels of noise (comparable, say, to the volumes you’d experience in the interior of an airliner in flight).
•    Be repeatable from test to test.
•    Be capable of showing not only how much noise reduction various noise cancellation circuits provided, but also of show which frequencies were affected by the circuits (some circuits cancel out a lot of noise at some frequencies, but not at others).

To meet these goals, I decided to set up two systems: a reference noise generator, and a reference headphone playback system.

Reference Noise Generator: my reference noise generation system consisted of the following:

•    Source: Musical Fidelity kW SACD player set on “Repeat” and playing the “Pink Noise” track from the Ayre Acoustics Irrational But Efficacious system conditioning CD.
•    Amplification: NuForce P9 preamplifier and a pair of NuForce Reference 9 Special Edition v.2 monoblock amplifiers.
•    Speakers: A pair of Mirage OM-28 omni-directional loudspeakers.

Here are a couple of points worth nothing about the noise generation system. First, the “Pink Noise” track produces even amounts of noise from octave to octave from the bass region though the midrange on up to the highs. When you listen to noise-cancelling headphones, then, you get a more accurate picture of how the various noise cancellation circuits work and how effective they are. Second, the Mirage speakers are omni-directional, meaning that they have very even power response throughout the room. As a result, noise more or less comes at you from every direction, just as would be the case in a plane, train, or bus. Third, the system can, when called upon to do so, play pretty loudly so that simulating very noisy environments isn’t a problem (well, within reason). Fourth, the NuForce preamp has a very finely graduated and clearly marked volume control that allows exactly repeatable volume settings during noise susceptibility tests.

Reference Headphone System: my reference headphone playback consisted of the following:

•    Source: iPod Classic loaded with a group of about 25 well-recorded and losslessly-encoded (both WAV and ALE format) music tracks representing a mix of pop, rock, jazz, and classical material.
•    Source, part 2: Wadia Model 170 iTransport (a device that can pull digital audio data directly from the iPod, thus bypassing the iPod’s good-but-not-great DACs and analog audio circuits.
•    Source, part 3: PS Audio Digital Link III DAC, connect to the digital outputs of the Wadia iTransport.  This DAC offers substantially better sound quality than the iPod’s built-in DACs and, as such, constitutes a “be all that you can be” output device for the iPod/Wadia.
•    Amplification: PS Audio GCHA headphone amplifier—a versatile and very high quality amp that let me hear the noise cancelling headphones at their very best.

Some comments: Is this a way better headphone playback rig than most people own? Yes, but I figured that in a review context readers would want to know “What can these ‘phones do under ideal circumstances?” I didn’t want the playback chain to limit the headphones’ performance in any way. However, just to keep everybody honest, I also drove each of the test ‘phones directly from the iPod’s headphone jacks to make sure the ‘phones worked well in that context, too.