For those who were wondering, here's all the additional information I've gleaned about these headphones over the years.
------------------------------------------------------------------------------------------------
My previous comment about this headphone line's quality control is only partly correct. The per-unit matching is usually good, but the unit-to-unit
variance of these drivers is obscenely high. To date, I have owned 10 of these headphones -- 4 DR-Z5s, 3 DR-Z6s and 3 DR-Z7s -- and measured 11 of them. Not a single pair of them had anywhere close to the same response at the driver's primary acoustic resonance, which is approximately 2kHz. Here's an overlay graph of one channel of each unit, with the two fabric-earpad DR-Z7s omitted, matched at 500Hz:
As much as
+/-6dB of variance within the audible frequency band. With the exact same earpad. In the exact same position. On the exact same amplifier and measurement coupler. Here's the two fabric-pad DR-Z7 units overlaid with each other for good measure:
The enitre response above 500Hz is completely different! One pair ski jumps into 3.5kHz like an AIR series AT while the other one gives us a big fat hump of upper midrange and tanks the treble. Big oof.
The best result I've seen overall is from my friend's "unicorn edition" DR-Z5, with about +3dB of reactance from 500Hz to 2kHz and the most bass
and treble of any of the units I've measured. Sadly, this headphone's left channel perished shortly after he bought it, and it wound back up in my hands awaiting a donor unit with a similar-sounding driver or a miraculous repair job.
Meanwhile, the
worst result comes from this particularly crap pair of DR-Z6 I bought about a year ago, with one blown driver (of course) and a mighty
8dB rise from the same middle midrange frequencies into a sharp 2kHz peak. Ouch. This thing is unlistenable. What the heck is going on here??
---------------------------------------------------------------------------------
From left to right: DR-Z5 (worn pad), DR-Z6 (new pad), and DR-Z7 (fabric pad):
The problem with these drivers lies in two aspects of their construction. The design itself is quite good, but variance in how it is assembled results in a couple of glaring issues:
- The thickness of the film used varies wildly, likely due to this material being rather new at the time, and the units with thicker film have a lower mechanical breakup frequency than the thinner drivers resulting in more breakup at the driver's primary resonance frequency.
- The "unicorn" DR-Z5 has a significantly thinner diaphragm than the other units, and produces higher frequency -- and lower SPL -- "crinkling" noises when deformed by hand. Typical units almost sound like a plastic water bottle and release much more energy when deformed this way due to the additional material present. In other words, the thinner film has less mass and releases less mechanical energy as it deforms, resulting in both less total energy and less resonance/persistence.
- The exact same thing happens when the suspension flexes for the driver's excursion, although much less catastrophically, and when reflected waves from the interior of the driver housing and from the ear canal collide with the film and transfer energy to it.
-
The DR-Z6 and DR-Z7 also have an electroplated palladium driver coating -- which does reduce the resonance of the material slightly -- but simultaneously slows its transients down and increases its overall mass, resulting in more acoustic breakup and thus, more 2kHz glare.
- The dimpling pattern on the center dome is completely random. Every single driver has a different pattern than the next. I forget where I read this bit of info and unfortunately do not remember the specifics, but the way Sony formed this pattern on the center dome was not via the now-obvious "just build the dimpling into the mould" option, like how modern companies such as Beyerdynamic do it. Judging by the end result, the film was likely heated in some way to bring it to its semisolid glass-transition state, and then had air blown at it from the rear to form bubbles. (This is only a guess; if someone remembers where that bit of info was disclosed that would be very helpful.)
- The dimples are important because they help break up the wave returning from the head and send it in random directions. Comparatively, a hard flat surface will reflect sound exactly where it came from, which will promote resonance.
-
The problem is that, sometimes, the entire dome surface is not dimpled. I have circled a portion of this driver in red, where the membrane is still flat and reflective towards the edges. This isn't the best example I have but the rest of my drivers are put away inside their housings, so please excuse my lack of desire to go digging for it. The worst example I've seen has about 50% of the center dome's surface left un-dimpled. Ridiculous. This should never have made it into a headphone.
----------------------------------------------------------------------------
I do not recommend bothering with these headphones until someone (probably me) works out a way to produce aftermarket drivers to a tighter spec than Sony managed in the 1970s. Even ignoring all the production issues the cable is extremely fragile and soldering to the drivers is almost guaranteed to destroy them due to the haphazard solder tab design. I own 10 of these headphones, only 3 of them still work, and only
one is something I would actually use regularly -- of course, my friend's broken pair. Everything else sounds somewhere between "dreadful" and "mediocre".