Details

Parent Category: Products

Category: Headphone Measurements

Created: 10 March 2019

Reviewed on: SoundStage! Solo, March 2019

I measured the Momentum True Wireless earphones using laboratory-grade equipment: a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, and a Clio 10 FW audio analyzer. For isolation measurements, I used a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface. A MEE Audio Connect Bluetooth transmitter was used to send signals from the Clio 10 FW to the earphones. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed. Note that because of the latency introduced by Bluetooth, I wasn’t able to do a spectral decay measurement, and of course my usual impedance and sensitivity measurements are irrelevant for wireless earphones. If you’d like to learn more about what our measurements mean, click here.

The above chart shows the Momentum True Wirelesses’ frequency response measured with the KB5000 and KB5001 anthropomorphic simulated pinnae. This is not terribly far from normal, although many earphones will have a bigger bump in the bass, and there’s less energy in the 5kHz region than we typically see.

This chart shows the Momentum True Wirelesses’ right-channel response compared with several other earphones, including the Jabra Elite Active 65t (another true wireless model), the Sennheiser HD Free (a wireless model with a cable connecting the earpieces), and the AKG N5005 earphones, the designs that currently best conform to the “Harman curve,” shown in research by Harman International to be the preferred in-ear headphone response for most listeners. You can see that while the Momentum True Wirelesses aren’t outliers, their response is generally flatter, with less bass output and somewhat attenuated response between 3 and 6kHz.

This chart shows the operation of the unusual tone control within Sennheiser’s Smart Control app. The dark blue line is the response with the tone control centered. (This is a measurement with a real-time analyzer and pink noise, so it looks much different than my usual frequency-response measurements, which are done with logarithmic chirp tones.) Note that a wide variety of tonal ranges are possible through the app. Although precise settings are not possible with this app, it has a range of about +4/-7dB in the bass and +/-6dB in the treble.

Because of the latency of the Bluetooth connection, it was possible for me to get a stable measurement of distortion versus frequency only at the extremely loud level of 100dBA. Even at this high level, distortion is typically around 1%, and peaks out at about 2.5%.

In this chart, the red line indicates an external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. The Momentum True Wirelesses’ isolation is about average for earphones fitted with silicone tips, although not as good as can be achieved with a model using over-ear cable routing, such as the Massdrop x NuForce EDC3s.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 10 March 2019

Reviewed on: SoundStage! Solo, October 2019

I measured the AU-Flex ANCs using laboratory-grade equipment: a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, and a Clio 10 FW audio analyzer. For isolation measurements, I used a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface. A MEE Audio Connect Bluetooth transmitter was used to send signals from the Clio 10 FW to the earphones. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed. Note that because of the latency introduced by Bluetooth, I wasn’t able to do a spectral-decay measurement, and of course my usual impedance and sensitivity measurements are irrelevant for wireless headphones. If you’d like to learn more about what our measurements mean, click here.

The above chart shows the AU-Flex ANCs’ frequency response measured with the RA0402 ear simulator (I wasn’t able to get an adequate seal using the KB5000 and KB5001 simulated pinnae). Interestingly, there’s little of the usual peak around 3kHz, but a strong peak around 5kHz, and that peak maxes out at about 5.5dB stronger than the bass response, all of which corresponds well with our listening impressions.

The impulse response shows that the latency with the MEE Connect is 175ms. This isn’t bad for Bluetooth earphones, which typically measure a little over 200ms latency, and it’s way better than the true wireless earphones I’ve tested, which typically have more than 300ms latency. The upshot is, you might experience some lip-sync problems when you watch videos using the AU-Flex ANCs, but it probably won’t be horribly distracting.

This chart shows the frequency response with ANC on and off, and in monitor (pass-through) mode. This is an amazing result, as the responses match perfectly except for slight differences in level. I’ve never seen such a close match before with noise-canceling headphones.

This chart shows the AU-Flex ANCs’ right-channel response compared with the passive Campfire Comet and AKG N5005 (the earphone said to best reflect the Harman curve) earphones, and the Sennheiser HD 1 Free Bluetooth earphones. You can see there’s ample extra energy between 4 and 6kHz, and not much low bass.

Because of the latency of the Bluetooth connection, I could not use the Clio’s sine sweep function to measure total harmonic distortion (THD) versus frequency, so I did discrete THD measurements of sine tones in one-octave steps. Note that distortion is very low at all frequencies and levels.

This chart shows the passive isolation with ANC off, the active isolation with ANC on, and the isolation (or lack thereof) in monitor mode. The red line indicates an external noise level of 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds.

This chart pits the AU-Flex ANCs’ isolation with ANC on versus two noise-canceling models (Bose QC30s and Plantronics BackBeat Gos) and a non-noise-canceling model (Sennheiser HD 1 Frees). The AU-Flex ANCs’ isolation is pretty good, but not world-class; you’ll likely still notice some noise coming through when you fly on a commercial airliner.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 01 March 2019

Reviewed on: SoundStage! Solo, March 2019

I measured the Next headphones using laboratory-grade equipment: a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, and a Clio 10 FW audio analyzer. For isolation measurements, I used a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface. The headphones were amplified using a Musical Fidelity V-CAN and an Audio-gd NFB-1AMP. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the Nexts’ frequency response. Surprisingly, considering our listeners’ reactions, this response is within the range of normal for headphones, but we can see two idiosyncrasies that correspond with our subjective impressions. There’s a shallow but unusually broad bass boost -- it spans about five octaves, while dynamic headphones might typically show a stronger but much narrower boost about two-and-a-half octaves wide. Also, the peak at 5.7kHz is a little stronger than the 3kHz peak. I’d normally expect that peak to be about 3 to 6dB weaker than the 3kHz peak.

This chart shows how the Nexts’ tonal balance changes when they’re used with a high-impedance (75 ohms) source, such as a cheap laptop or some cheap professional headphone amps. Using the higher-impedance source produces a very slight increase in the bass of about 1dB at 50Hz.

This chart shows the Nexts’ right-channel response compared with a few other audiophile headphones in the same approximate price range as well as with the AKG N700NCs, the headphones that probably come closest to conforming to the so-called “Harman curve.” The Nexts are close in some ways to the AKG N700NCs, although with a weaker and much broader bump in the bass, and more energy between 5 and 6kHz.

The Nexts’ spectral decay (waterfall) chart shows near-zero resonance across the entire audioband. There’s just one very high-Q (i.e., narrow) resonance centered at 2.3kHz, which corresponds with a quarter-wavelength of about 1.5”, so it might be occurring within the earpads.

The Nexts’ measured total harmonic distortion (THD) is unusual in that it’s basically flat at 3% in the bass at both of the test-signal levels I use for measuring. Normally it would rise at the higher level. Although it’s a little strange, I’d guess it isn’t a big deal, as 10% is the commonly accepted threshold for distortion audibility in subwoofers.

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. The isolation measurement of the Nexts suggests they’re really more of a semi-open-back design than a pure open-back design like the HiFiMan Anandas -- i.e., they do offer a modest amount of isolation at frequencies above about 1.5kHz.

The impedance magnitude of the Nexts is just about dead flat at 16 ohms (same as the rating), and the impedance phase curve is similarly flat.

Sensitivity of the Nexts, measured between 300Hz and 3kHz, using a 1mW signal calculated for 16 ohms rated impedance, is 97.0dB. That’s a couple dB above the breaking point between “needs an amp” and “works OK plugged straight into a smartphone.”

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 20 February 2019

Reviewed on: SoundStage! Solo, February 2019

I measured the Stellias using laboratory-grade equipment: a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, and a Clio 10 FW audio analyzer. For isolation measurements, I used a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface. The headphones were amplified using a Musical Fidelity V-CAN or an Audio-gd NFB-1AMP. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the Stellias’ frequency response. This is a little flatter than I typically measure from audiophile headphones. There’s a very mild resonant bass bump centered at 100Hz, and less of a pronounced peak between 2 and 4kHz than most audiophile headphones exhibit.

This chart shows how the Stellias’ tonal balance changes when they’re used with a high-impedance (75 ohms) source, such as an inexpensive laptop or some cheap professional headphone amp -- or more likely in the Stellias’ case, a tube amp with transformer output (although that still wouldn’t be near 75 ohms). Using the higher-impedance source produces an increase in bass of 2 to 2.5dB below 400Hz.

This chart shows the Stellias’ right-channel response compared with the Focal Clear, HiFiMan HE1000 V2, and Sennheiser HD 800 S headphones. You can see a bit of family resemblance between the Stellias and the Clears, although they have significant differences. The HE1000 V2 and the HD 800 S headphones have a lot more treble energy and less bass energy than the Stellias.

Across almost the entire audioband, the Stellias’ spectral decay (waterfall) chart shows a little more resonance than I’m used to seeing, but it’s almost entirely damped out within 6ms (one full cycle at 167Hz), so it’s very unlikely to be audible.

The Stellias’ measured total harmonic distortion (THD) is inconsequential at the loud listening level of 90dBA (measured with pink noise). At the extremely loud level of 100dBA, it gets up to around 4% below 100Hz, but in the bass, 4% isn’t a big deal. (The generally accepted figure for audible harmonic distortion in subwoofers is 10%.)

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. Isolation of the Stellias is in the same ballpark as other closed-back models such as the Audeze LCD2 Closed-Backs and the NAD Viso HP50s -- plenty enough to shield you against jibber-jabber and light background music in the unlikely event you bring your Stellias to Starbucks.

The impedance magnitude of the Stellias averages about 40 ohms, and runs as high as 56 and as low as 31 ohms. That swing in the bass is what produces the difference in response when switching from a low-impedance source to a high-impedance source. Despite the fairly large magnitude swing, the impedance phase is reasonably flat.

Sensitivity of the Stellias, measured between 300Hz and 3kHz, using a 1mW signal calculated for 35 ohms rated impedance, is 104.0dB. For a large set of audiophile headphones, that’s really impressive -- feel free to plug them straight into your smartphone if you like.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 10 February 2019

Reviewed on: SoundStage! Solo, February 2019

I measured the Ultimate Ones using a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, a Clio 10 FW audio analyzer, a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface, and a Musical Fidelity V-CAN amp, with an Audio-gd NFB-1AMP used for distortion measurements. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the Ultimate Ones’ frequency response. This looks typical except for the pronounced dip between 300Hz and 2kHz. Note the roughly three-octave-wide bump in the bass response. Because the bump is so wide, it will tend to make the bass sound moderately elevated in level rather than boomy.

This chart shows how the Ultimate Ones’ tonal balance changes when they’re used with a high-impedance (75 ohms) source, such as a cheap laptop or some cheap professional headphone amps. There’s barely any difference, just a roughly 1dB bass boost between 30 and 60Hz with the high-impedance source.

This chart shows the Ultimate Ones’ right-channel response compared with a few other headphones in the same price range (AudioQuest NightHawk, HiFiMan HE400i, and Sennheiser Massdrop HD 6XX). Note that these are all referenced to 94dB at 500Hz. Obviously, the Ultimate Ones’ deep midrange dip will make them sound quite a bit different than the other models shown here.

The spectral decay (waterfall) chart shows a little more resonance in the bass than I’m used to seeing, but it’s at -20dB right from the start and falls below -40dB after 8ms, or about one-and-a-half cycles at 200Hz, so I would be very surprised if it’s audible.

At the loud listening level of 90dBA (measured with pink noise), the Ultimate Ones’ measured total harmonic distortion (THD) is a little high below 100Hz. The bass distortion doesn’t rise much at the extremely loud level of 100dB, but there’s a little more distortion between 1.3 and 2.2kHz.

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. I’d classify the Ultimate Ones as semi-open-back, as they offer more isolation than fully open headphones such as the HiFiMan HE400i’s, but less isolation than a typical closed-back model such as the NAD Viso HP50.

The impedance magnitude of the Ultimate Ones averages about 37 ohms and doesn’t vary much from that, and the impedance phase is also flat.

Sensitivity of the Ultimate Ones, measured between 300Hz and 3kHz, using a 1mW signal calculated for 35 ohms rated impedance, is 97.0dB. So they’ll probably play loud enough from most smartphones, although you might not be able to get them really cranking.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 20 January 2019

Reviewed on: SoundStage! Solo, January 2019

I measured the HE6se headphones using a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, a Clio 10 FW audio analyzer, a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface, and a Musical Fidelity V-CAN amp, with an Audio-gd NFB-1AMP used for distortion measurements. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the HE6ses’ frequency response. This is comparable to what I’ve seen with many HiFiMan models, and not far outside what I’ve seen from most open-back planar magnetics, except that the peak between 3 and 4kHz (which is generally considered necessary to create the illusion of hearing real speakers in a real room) is a few dB higher than normal.

This chart shows how the HE6ses’ tonal balance changes when they’re used with a high-impedance source, such as a cheap laptop or some cheap professional headphone amps. (Of course, the headphones’ low sensitivity means it’s practically impossible to use them with a low-quality amp.) There’s essentially no difference at all, which means the HE6ses’ tonal balance probably won’t change if you decide to use them with, say, a high-output-impedance tube amp.

This chart shows the HE6ses’ right-channel response compared with three other open-back planar-magnetic models (the Audeze LCD-Xes, Focal Clears, and Quad ERA-1s with the leather pads). The strength of that peak between 3 and 4kHz is evident here; no way these headphones won’t sound a little bright.

The midrange hash -- the very narrow, high-Q resonances between 3 and 5kHz -- of the HE6ses are seen in almost all the spectral decay (waterfall) measurements I’ve done of planar-magnetic headphones, but here they’re focused across a smaller band and are higher in magnitude. This corresponds with the big peak in the frequency response in this same range. Meanwhile, the bass resonance is practically non-existent, the least I can remember measuring in any headphones.

The HE6ses’ measured total harmonic distortion (THD) is pretty low, except for a weird little bump to about 3.5% centered near 1.5kHz. This is at the extremely loud listening level of 100dBA, though.

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. The HE6se headphones offer among the least isolation from outside sounds that I’ve ever measured -- but of course, open-back models aren’t supposed to isolate you from outside sounds, and the fact that the HE6ses provide so little isolation also means that the acoustical impedance of their rear grilles is very low, which may have contributed to the big sense of space I heard from these.

As usual with planar-magnetic headphones, the impedance magnitude of the HE6ses is almost perfectly flat. It’s about 65 ohms through the entire audio range, and the impedance phase is also flat.

Brace yourself, because the sensitivity of the HE6ses, measured between 300Hz and 3kHz, using a 1mW signal calculated for 50 ohms rated impedance, is 79.2dB, more than 4dB lower than the already-low rating of 83.5dB. Which means you will definitely need an exceptionally powerful amp for these.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 20 January 2019

Reviewed on: SoundStage! Solo, February 2019

I measured the ERA-1s using a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, a Clio 10 FW audio analyzer, a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface, and a Musical Fidelity V-CAN amp, with an Audio-gd NFB-1AMP used for distortion measurements. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the ERA-1s’ frequency response with the perfed leather pads. What’s unusual here is the small peak centered at 1kHz, and the fact that the peak at 3kHz is lower in magnitude than usual; typically, the peaks in a headphone’s response between 5 and 10kHz are a few dB lower than the 3kHz peak.

This chart shows the acoustical effects of the perfed leather and velour pads. The velour pads should deliver a softer treble and a fuller tonal balance, at the likely expense of perceived treble detail and spaciousness.

This chart shows how the ERA-1s’ tonal balance changes when they’re used with a high-impedance source, such as a cheap laptop or some cheap professional headphone amps. With the high-impedance source, there’s just a barely measurable (and not audible) reduction in bass of a fraction of a decibel from 20 to 30Hz.

This chart shows the ERA-1s’ right-channel response compared with two other open-back planar-magnetic models (the HiFiMan Anandas and the Audeze LCD-Xes), as well as the AKG N700NCs, a closed-back model said to deliver response very close to the “Harman curve,” shown in research by Harman International to be the preferred over-ear headphone response for most listeners. The ERA-1s are generally similar to the other planar magnetics.

The midrange hash (very narrow, high-Q resonances) shown in the ERA-1s’ spectral decay (waterfall) chart is typical of planar-magnetic headphones, although the ERA-1s’ hash covers a wider band than most.

The ERA-1s’ measured total harmonic distortion (THD) is generally low, although actually slightly higher than I typically measure from planar-magnetic headphones. It reaches an audible level only in the bottom octave of bass (20-40Hz), and at an extremely loud level, but most music has little or no content in that range, and most of the distortion harmonics are so low in pitch that they’ll be hard to hear.

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. Like almost all open-back models, the ERA-1s offer little isolation from outside sounds -- much less than that of the Audeze LCD2 Closed-Backs also shown here.

As usual with planar-magnetic headphones, the ERA-1s have an almost perfectly flat impedance curve, running about 21 ohms through the entire audio range, and impedance phase is similarly flat.

Sensitivity of the ERA-1s, measured between 300Hz and 3kHz, using a 1mW signal calculated for 20 ohms rated impedance, is 93.4dB with the perfed leather pads and 96.1dB with the velour pads. This means that while you will likely get usable volume if you plug the ERA-1s straight into a smartphone, you’ll get the best performance from the ERA-1s if you use a high-quality portable music player or amplifier.

. . . Brent Butterworth
brentb@soundstagenetwork.com

Details

Parent Category: Products

Category: Headphone Measurements

Created: 20 December 2018

Reviewed on: SoundStage! Solo, December 2018

I measured the LCD2 Closed-Backs using a G.R.A.S. Model 43AG ear/cheek simulator/RA0402 ear simulator with KB5000/KB5001 simulated pinnae, a Clio 10 FW audio analyzer, a laptop computer running TrueRTA software with an M-Audio MobilePre USB audio interface, and a Musical Fidelity V-CAN amp, with an Audio-gd NFB-1AMP used for distortion measurements. These are “flat” measurements; no diffuse-field or free-field compensation curve was employed.

The above chart shows the LCD2 Closed-Backs’ frequency response. This is a typical measurement for audiophile-oriented headphones, with maybe just a dB or two more bass than many large planar-magnetic models offer.

This chart shows how the LCD2 Closed-Backs’ tonal balance changes when they’re used with a high-impedance source, such as a cheap laptop or some cheap professional headphone amps. This is a little surprising, because most planar-magnetic headphones show no variance on this test. With the LCD2 Closed-Backs, the higher-impedance source produces a boost of 1 to 1.5dB between 1.5 and 3kHz.

This chart shows the LCD2 Closed-Backs’ right-channel response compared with two other audiophile-oriented headphones (the Fostex TH909s and Focal Elegias), as well as the AKG N700NCs, a new closed-back model said to deliver response very close to the “Harman curve,” shown in research by Harman International to be the preferred over-ear headphone response for most listeners. The LCD2 Closed-Backs are clearly in the ballpark of a “standard” headphone response, with perhaps a little more treble and a slightly fuller midrange than we see in most over-ear headphones.

The LCD2 Closed-Backs’ spectral decay (waterfall) chart shows slight comb filtering effects between 2 and 6kHz that are commonly seen in planar-magnetic headphones, but they’re far better-damped than usual, probably because of the closed back and perhaps some damping material inside.

Measured total harmonic distortion (THD) of the LCD2 Closed-Backs is near zero, barely above the noise floor of the measurement.

In this chart, the external noise level is 85dB SPL, and numbers below that indicate the degree of attenuation of outside sounds. The LCD2 Closed-Backs’ isolation is about average for headphones of this type, and notably better than a conventional, smaller set of closed-back headphones such as the NAD Viso HP50s (also shown) -- and, of course, better than Audeze’s open-back models.

Here’s an unusual result -- most planar-magnetic models have a dead-flat impedance curve, but the LCD2 Closed-Backs’ impedance response has a large peak of 108 ohms, centered at 2.4kHz; through most of its range, the impedance magnitude is flat at 72 ohms. Despite the fairly large magnitude swing, the phase response shows just a mild shift at the same frequency. This peak is why the response of the headphones changes when they’re used with a high-impedance source.

Sensitivity of the LCD2 Closed-Backs, measured between 300Hz and 3kHz, using a 1mW signal calculated for 70 ohms rated impedance, is 97.5dB (the rating is 97dB). That’s high enough that you’ll get a somewhat usable volume from typical smartphones, but low enough that you’ll want to use a high-quality portable music player or amplifier for most of your listening.

. . . Brent Butterworth
brentb@soundstagenetwork.com