A Sampling of Headphones with Audiophile Aspiration (Vol.1, No.5: Winter 1977/78)

Reviewd below:

Fontek Minifon A-4 ; Koss 'Auditor' Dynamic/10 ; Koss 'Auditor' ESP/10 ; Stax SR-X/Mark 3 ; Stax SR-5 ; Yamaha HP-1

---------

We examine four electrostatics and two electrodynamics (the latter, as it turns out, mainly for comic relief).

Headphones are exactly like loudspeakers and, at the same time, totally different. The similarity lies in the electroacoustic transducer format: in each case, the output waveform of an amplifier is translated into the excursions of a diaphragm, and the same principles apply.

The difference is in the coupling to the ear: loudspeakers actually try to imitate life by producing a sound pressure gradient in the listening room, whereas headphones bypass the natural listening environment and "mainline" the pressure field directly into the ear canal.

As a result, loudspeakers tend to sound more natural, since real-life interfaces are maintained; on the other hand, headphones have a better chance to preserve waveform accuracy all the way to the eardrum, since the signal goes through fewer transformations and, be sides, the smaller diaphragm is more precisely controllable.

The impression critical listeners generally form about headphone listening is that more sonic information is revealed ("I hear things I didn't know were on the record") but that the experience is ultimately fatiguing and unsatisfactory. We feel that a pair of first rate electrostatic headphones, such as the Stax SR-X/Mark 3, is an invaluable "audio loupe''-a magnifier for examining what goes on in a particular piece of equipment when inserted into a known chain of components.

We also find the uncanny clarity of such head phones exhilarating, but in the long run we can't derive much musical pleasure from their use. We prefer our music by total immersion rather than intravenously, not only because it feels more normal that way but also because low bass is perceived more through the body than through the ear canal. Nevertheless, no serious audio enthusiast should be without a pair of good headphones; just to be able to play at any loudness level without intruding on others is a good enough reason.

Incidentally, none of the above remarks apply to binaural sound reproduction; that's a very different ball game from stereo, to be explored in depth in a future issue.

How we measured them.

The frequency-domain and time-domain measurements we made to evaluate speaker systems (see our Number 4 issue, page 20) are equally applicable to headphones. The problem, of course, is coupling. How do we create the same interface between the head phone diaphragm and the measuring micro phone as exists in actual use between the headphone diaphragm and the ear? The standard solution is an "artificial ear," which is an acoustic coupler made of metal and de signed with a somewhat arbitrarily chosen "official" internal volume of 6 cm? Since your coupler and mine are made of flesh, blood and cartilage, and since the cubic capacity of yours isn't necessarily the same as of mine, we aren't too happy with this solution. It doesn't appear to simulate the on-the-head acoustic transfer conditions closely enough to yield results we'd be willing to accept as "official." As a matter of fact, pressing the headphones a little tighter to your ears or moving them a few millimeters will sufficiently alter the existing acoustic impedance matches to create a whole new set of frequency response characteristics.

We therefore decided to work around the problem, taking a whole series of iterative measurements, some with no coupling at all, others with loose coupling by hand, tight coupling by hand, probing with the microphone between the ear and the diaphragm, etc., etc.

The composite picture that emerged, while perhaps not quite accurate quantitatively, pro vided reliable information about the basic frequency-domain behavior of the unit under test, such as smoothness or roughness, major dips or peaks, upward or downward slopes, and so forth. We're satisfied that our measurements reflected the realities of on-the-head performance, since the results correlated pretty well with what we heard. All of this applies, of course, only to frequency-response testing; the time domain can be investigated with nearfield readings of pulses and tone bursts, just as if the headphones were small loudspeakers.

How we listened to them.

The program material we used for our comparative listening tests consisted almost exclusively of direct 15-IPS copies of 30-IPS original master tapes, mostly recorded with only a single pair of microphones for the main pickup. The tape deck used was a highly modified Stellavox 'Stellamaster'; the amplifier in each case was the Bryston 4B (except that the two Stax models were also checked on an experimental tube amplifier especially designed for them by Julius Futterman, with the electro static elements driven directly off the output tube plates).

Headphone listening is a solitary experience, but we made sure that at each listening session there were at least two pairs of educated ears available to swap phones and com pare notes; on occasion there were three.

The findings reported below represent the unanimous conclusions of staff members and consultants; there were no major disagreements at any point during the tests, and minor disagreements were quickly resolved upon repeated listening to the same passage.

The most surprising discovery was not that electrostatic headphones are superior even to the best electrodynamics but that the latter aren't even close. Listening in quick succession to the Koss dynamic, for example, and either one of the Stax electrostatics elicited giggles from each auditioner. The difference was funny. Since the price differences aren't so staggering (especially in the case of the Stax SR-5), we don't see how the audio purist can be expected even to consider dynamic headphones for critical listening.

Fontek Minifon A-4 Specs Corp., 1169 E. Chess Drive, Foster City, CA 94404.

Fontek Research Minifon A-4 electrostatic unit with C-4 coupler, $300. Tested sample on loan from distributor.

Fontek headphones are designed, we're told, by a Mr. Niwa who was formerly with Stax and involved in the development of the SR-X. The Minifon A-4 electrostatic is being sold as a Stax beater, which it isn't-not quite, anyway. But it's good enough to have given us the impression on first hearing that it just might be.

The A-4 has a bright, forward sound, with lots of presence-not the crude, zingy sort, just very detailed with a good, clean edge.

That's probably why some people rate it number one. Compared to the Stax SR-X/Mark 3, however, it comes off as a wee bit edgy, aggressive and, at the same time, slightly woofy. It's not an altogether neutral reproducer. This is especially apparent on complex material with a rich harmonic structure, such as massed brasses.

The frequency response is quite flat, though under certain coupling conditions it appears that there's a loss of upper highs with respect to the level of lower highs. The SR-X is definitely smoother overall. In the time domain, the Fontek exhibited a moderate amount of ringing; the rise time was not quite as fast as that of the SR-X; furthermore, our suspicion of some frequency response irregularies received time-domain confirmation on square pulses. On the whole, good but not brilliant test bench performances.

Physically, the Minifon A-4 is mini in deed: very small and light, designed to be worn with very little pressure on the head, but not quite as beautifully finished as the SR-X/Mark 3. The coupler box that goes between the amplifier and the phones looks well made, although we found an unforgivable piece of loose wire rattling around inside our sample, apparently a paring from the transformer leads before they were soldered into place. The coupler will drive two pairs of headphones but does not incorporate a headphones/speakers switch.

We must add that we've heard rumors of unit-to-unit variations in Fontek products, al though we certainly don't believe that the A-4 we tested was in any way defective. If there are better-sounding samples out there, they must be awfully good.

Koss 'Auditor' Dynamic/10 Koss Corporation, 4129 N. Port Washington Avenue, Milwaukee, WI 53212. Auditor Series Dynamic/10 stereo headphone, $85. One-year warranty. Tested #0495, on loan from manufacturer.

This is the top-of-the-line electrodynamic stereo model of the world's leading headphone manufacturer. As such, it comes off as a distinct disappointment. Compared to even the least expensive electrostatic we tested, it sounds like a real mess-jumbled, unfocused, not at all transparent, rough, and quite fatiguing. Also rather uncomfortable on the head. In all fairness, though, it must be pointed out that the competition in this group of six headphones was quite formidable; as one of our most experienced auditioners remarked about the Dynamic/10, after agreeing with its last-place ranking, "If you think this is bad, you ought to hear what really bad sounds like." True enough.

Our measurements confirmed the listening tests: the Dynamic/10 is totally incoherent in the time domain. Pulses of any duration are reproduced (if one can speak of reproduction in such a case) without the slightest resemblance between input and output. That almost certainly explains the lack of transparency and focus; the rough, irritating quality is probably explained by the ringing we observed on tone bursts. The frequency response also shows considerable roughness; between 2 kHz and 23 kHz there are jagged peaks and dips all over the place, and the bass rolls off at approximately 12 dB per octave below a 4 dB peak in the lower midrange (in the neighborhood of 300 Hz-measured with tight coupling). Not a nice response at all.

We don't quite understand why Koss is marketing the Dynamic/10 as an audiophile product. The Yamaha HP-1, for example, is a considerably better dynamic headphone and costs $20 less.

Koss 'Auditor' ESP/10 Koss Corporation, 4129 N. Port Washington Avenue, Milwaukee, WI 53212. Auditor Series ESP/10 electro static stereo headphone with E/10 energizer, $300. One year warranty. Tested #0249, on loan from manufacturer.

As all good audio freaks know, the Koss ESP/9 was hailed as a reference standard by a number of reviewers and assorted audio professionals when it came out a few years ago.

We have no opinion on the validity of that judgment, since we haven't tested the ESP/9; what we have here is the new model that Koss claims is a substantial improvement over all their previous electrostatics. The ESP/10 is both the most expensive and the most elaborate Koss headphone to date.

The top-of-the-line image is apparent from the massive E/10 energizer with its two VU meters, double outlets for two pairs of phones, selector button for phones/speakers, LED over load indicator, etc. Very impressive. More impressive, in fact, than the sound that results when these goodies are activated.

Not that the ESP/10 sounds bad; Koss couldn't get away with that for $300. But it unquestionably sounds the least accurate among the four electrostatics we tested.

Specifically, transparency and definition of inner detail are well below the standard set by the other three; in addition, the aggressiveness we noted in the Fontek is also evident, perhaps even more of it. On the whole, a somewhat veiled, unfocused and not particularly pleasing sound, at least by comparison.

The frequency response of the ESP/10 is very flat at its "sweet spot," maybe the flattest of all the headphones tested in terms of maximum observable amplitude deviations, but the Stax SR-X/Mark 3 has a smoother curve.

Under other conditions of measurement the Koss appears to have a very depressed mid range; which condition corresponds most closely to on-the-head use is hard to tell.

In any event, a nasty standing wave exists at 11 kHz and another, not as obnoxious, at 20 kHz: we suspect the cavity behind the diaphragm to be the probable cause. Pulses were reproduced by the ESP/10 with slower rise time, more time smear and more ringing than by the SR-X; compared to the Fontek, however, the Koss wasn't significantly inferior.

Thus the audible superiority of the Stax is well supported by the lab data; why the Fontek also sounds better isn't quite as clear, although certainly more explicable than the reverse would be.

One more thing. The ESP/10 is rather uncomfortable to wear, its earcup and head band design being identical to that of the Dynamic/10. At 3%: times the price, Koss could have done a little better. Overall, this ain't no reference standard.

Stax SR-X/Mark 3 American Audioport, Inc., 1407 N. Providence Road, Columbia, MO 65201. Stax SR-X/Mark 3 electrostatic "earspeaker" with SRD-7 adaptor, $230. Tested #02115 (adaptor #4062), owned by The Audio Critic.

This is without doubt the finest headphone we have heard and measured to date, but you must keep in mind that there are a number of contenders (such as, for example, the Infinity electrostatic) that we haven't been able to test so far. Anything better than the SR-X/Mark 3 would have to be pretty sensational, though.

Phenomenal clarity, startling definition of detail, and freedom from colorations characterize the sound of these phones; the rendition of the input signal appears to be essentially complete, except of course for the low bass. In fact, there isn't much to say about the sound of the SR-X/Mark 3, since the usual anomalies that occasion instant comment just aren't there. Some people (not the ones we associate with) claim to detect some hardness or over brightness at the higher frequencies, but we're virtually certain that what they're hearing is a merciless resolution of the intermodulation sidebands generated by their inadequately aligned phono cartridges.

The frequency response of the SR-X/ Mark 3 is very flat and exceptionally smooth (the smoothest of all the headphones tested) out to 20 kHz and even beyond. Under certain conditions of acoustic loading there develops a large peak at 20 kHz; we could see it on the spectrum analyzer but couldn't hear it. In the time domain, this was the unit that rang the least, had the fastest time rise and created the least time smear among the six headphones we measured. The correspondence here between audible and measurable performance is as good as a reviewer who doesn't believe in magic could possibly ask of Mother Nature for corroboration.

The SRD-7 adaptor, which supplies the bias voltage for the electrostatic elements and provides the impedance match between the power amp and the phones, features out lets for two pairs of Stax electrostatics (any model) and convenient switching between loud speakers and headphones (a red light comes on when the latter are energized). One thing that disturbed us somewhat was the introduction of 0.13 ohm additional speaker lead resistance per channel when the main speakers are switched through the SRD-7. We find that excessive, especially in cases where there are long speaker leads to begin with. For $230 Stax could have put in thicker wire to help keep damping factors high.

When we started The Audio Critic, the Stax SR-X/Mark 3 was already our reference headphone. After five issues' worth of listening tests, it still is.

Stax SR-5 American Audioport, Inc., 1407 N. Providence Road, Columbia, MO 65201. Stax SR-5 electrostatic ''ear speaker" with SRD-6 adaptor, 3130. Tested #01228, on loan from owner.

For 56; cents on the dollar, Stax will put you within a hairsbreadth of SR-X/Mark 3 ownership. We can't think of another 'son of" product in the field of audio that approaches the performance of the high-priced "daddy" product as closely as the SR-5.

What do you give up when you buy the SR-5 instead of the SR-X/Mark 3? A tiny bit of smoothness and that ultimate refinement of inner textures-that's about all. The two models sound remarkably alike. In the laboratory, the only definite difference we could nail down was very slightly better pulse reproduction by the SR-X. On tone bursts, for example, both were equally excellent. We gave up after a while trying to find a major difference.

The SR-5 is slightly larger than the SR-X and, unlike the latter, completely surrounds the ear with its bulkier earcup. It's still very light, however, and comfortable to wear. Both models are open-backed and therefore not completely sealed against sounds from the outside.

Bargains are few and far between in equipment for the audio purist, but the Stax SR-5 is definitely one of them.

Yamaha HP-1 Yamaha International Corp., Audio Division, PO Box 6600, Buena Park, CA 90622. HP-1 Orthodynamic headphones, $65. Tested sample owned by The Audio Critic.

All we can say about this one is that it's a hell of a lot better than the Koss Dynamic/10, not to mention assorted nondescript electro dynamics, but it just hasn't got the focus and definition that seems to be the exclusive province of the electrostatics. It has excellent amplitude response, though, and you can hear it in the solidity, smoothness and balance of the sound: +2 dB from 40 Hz to 1 kHz, +3 dB from there on up to 10 kHz. The bass is particularly impressive; there's some roughness, however, in the top octave, with a big peak at 19 kHz preceded by a characteristic suck-out at 11 kHz. On the whole, flatter and smoother than most dynamics. Very listenable, too, with out anything disturbingly unmusical to be singled out. Dynamic range is excellent.

It's in the time domain that things fall apart; the HP-1 is totally incapable of reproducing pulses of any duration. Furthermore, on tone bursts there's unmistakable evidence of ringing and time smear. Hence the lack of ultimate clarity and resolution of inner detail.

It would be unfair to dismiss the HP-1 without mentioning that it's extremely light, rugged, comfortable to wear and pleasant to use. Still, we can't wholeheartedly recommend it to the purist.

Recommendations

Even though six models constitute a very small sampling of what the headphone market offers today, there's a good chance that the recommendations below would have been the same if we had conducted a much broader survey. They represent the best of current thinking and execution.

Best headphone so far, regardless of price: Stax SR-X/Mark 3.

Close to the best at a much lower price: Stax SR-5. SE

---------

[adapted from TAC]

---------

Also see:

Sophisticated Speaker Systems, Large and Small: Our Unending Survey

Various audio and high-fidelity magazines