This review page is supported in part by the sponsors whose ad banners are displayed below |
 |
 |
In our pages Nick Bedworth's Audiophilleo 1 & 2 review contains useful jitter background. The only meaningful way to assess specific audible rather than measurable jitter effects—whilst determining where things turn specmanship and stop mattering—would be to deliberately compromise just the jitter performance of a given device. Audiophilleo's designers provide such an educational tool. They include in their converter a second inferior clock. Curious owners can switch it in to educate themselves on the effects in what otherwise remains an identical circuit. In normal hifi reviews such jitter specificity never factors of course. Statements on one DAC being superior because of lower jitter are flawed unless the jitter aspect was first properly isolated from D/A conversion, upsampling/oversampling, I/V conversion, the output stage, output impedance and other factors. Without such protocol one simply cannot compare two completely different DACs, then categorically state that one sounded better because it had the superior jitter specs.
|
|
PureMusic's audio setup interface recognized the KingRex as 'SPDIF OUT'
and supporting all common sampling rates up to 192kHz |
Back to the KingRex. My comparison context had the usual iMac running OSX 10.6.5 and PureMusic 1.73 in memory play with pre-allocated RAM. This front end streamed AIFF files via USB directly into the Burson Audio's asynchronous USB input. Alternately it streamed into the KingRex first, thence via Black Cat Cable Veloce S/PDIF cable into the Burson. Round II replaced the Burson with the Weiss. Here I also compared the Veloce coaxial cable to the AES/EBU link KingRex had included. Because the Burson has a very high-quality stepped attenuator, it connected directly to my FirstWatt F5 amplifier. The Weiss lacks analog attenuation. It thus leashed to the StereoKnight Silverstone Balanced transformer volume control. This eliminated more active circuitry and digital attenuation. In my digs passive or direct drive into the F5/Tango R amp/speaker combo are the two most transparent and highly resolved setups I can muster. Given the hairsplitting I anticipated for this assignment, drafting my most powerful microscope seemed à propos
|
 |
|
|
|
The verdict of Round I was rather easier than predicted. The detour path through the KingRex altered textures. Instead of for example having a singer's voice appear robed in her own recorded ambiance, she appeared nude in the familiar ambiance of my room. The air injection which acoustic reverb creates behind and around a voice or instrument dried out. These very fine spiderwebby decay trails, reflections and harmonic sprays that add the recorded 3rd dimension were less visible. Naturally the stereo image itself didn't shift. It didn't collapse or flatten out. The performers didn't move forward, backward or sideward. Everything stayed put. The only thing subtracted was some evidence of a foreign acoustic getting mildly superimposed or inserted into my own. Why this would also influence tone textures is easily understood. Remember singing in your empty bathroom before you hung up towels and curtains, added furnishings and laid out mats to damp that nice echo effect which so enhanced your voice? Now minimize the severity of that gross example to something very subtle and fine. That was the difference between the asynchronous direct feed and the isochronous indirect feed: less recorded space, drier textures.
|
 |
In the grander scheme of hifi things, changing cables is likely to effect bigger changes.
The losses encountered with the KingRex were exceedingly minor in scope if—from the high-end perspective where fractional percentage points are chased—not immaterial or irrelevant. For just $189 however and in my context of utter redundancy where the UC192 was 5th wheel rather than necessity, this seemed a good showing. Incidentally there's no cheating. While PureMusic can be told to downsample files to a source's best capacity, it couldn't in this instance because what it saw was the 24/192 KingRex, not the 24/96 Burson. It thus logically refused to play back 24/192 files at 24/96. As far as it was concerned, the DAC (which in reality was a DDC only) could handle full data density. The result? No sound out of the Burson until I selected tracks at or below its native resolution. Thus don't make the mistake of believing that the preceding device's resolution can 'override' the specific component where actual D/A conversion occurs.
|
|
|
Round II was a rerun of round I yet fainter in magnitude. Switching the Weiss DAC2 between the AES/EBU and coaxial feeds incoming from the UC192 showed a difference where I'd not expected any. Coax proved to be more focused and firmer. Whether this involved formats or cabling of dissimilar quality—Chris Sommovigo's Veloce specialty cable was likely superior to KingRex's own—I could no more determine than claim that the Weiss held a small edge because Firewire trumps USB.
Conclusion: As though to spite the bits-is-bits brigade, whether internal to a DAC/player or external, D-to-D conversion (here USB to coaxial, AES-EBU or I²S) is no matter of pure immutability. Presuming that here the greatest variable is jitter, its effect in these leagues appears to be the degree by which ambient micro detail is recovered. Unlike changes in frequency response and tonal balance we expect from that most nonlinear hifi device the loudspeaker; unlike effects from harmonic distortion which discussions over tubes and transistors tend to focus on; these low-jitter differences are far finer in makeup. As such they are far less blatant and disturbing. This places them far lower on the totem pole of hifi essentials to worry about.
Given the $189 price tag on the KingRex micro component and how closely it trailed the internal D-to-D conversion of far more expensive machines which could eliminate its second cable... the UC192 delivered exactly what a computer audio transport ought to. For legacy digital it builds an affordable bridge to that shiny USB port at the business end of a laptop or PC whilst passing up to 24/192 files. Fence sitters should find it an ideal way to test the waters. If that's who KingRex had in mind primarily, the UC192 hits the inner circle of the bull's eye. If only the reluctant would give it a go. As reader emails portray, plenty of audiophiles continue to resist uncompressed streaming audio. Occasionally I manage to entice one to jump. The feedback then is predictable:
|
• "I had no idea."
• "I'm listening to much more music than before."
•"Why did I wait so long?"
Indeed. On all three counts. Now you're out. And this c.a.t.'s outa the bag too... |
|
|
|
|
|
 |
 |
|
 |
