Music in the Round #90: Merging+Player PL8 Page 2

Still, I'm not deaf to the siren songs of cable companies, or the many audiophiles whose praises pepper their ads. Yeah, I want more, but I'm skeptical that there's any to be had, and I sense that very-high-priced cables amount to audio jewelry. Nonetheless, recently I succumbed to temptation and took a flyer on some wire.

Revelation Audio Labs Prophecy CryoSilver Reference DualConduit USB 2.0 Digital Link ($649/1.25m): Revelation Audio Labs' Prophecy DC cables came with my Baetis Prodigy-X. server, and I was impressed with their construction. I then discovered that RAL had a USB cable with their signature CryoSilver conductors and with dual Type A input connectors to converge on a single Type B output connector, with one input for signal and another for USB power."


The separation of signal and power leads, with the option of providing that 5V power from a quieter, cleaner source than the main board or PCIe card of a computer, can work with a specialized USB card such as SOtM's tX USBhubIN USB board, which can derive its 5V output from an external DC source. In my server, power comes from a dedicated 9VDC output on a HDPlex 400W DCATX power supply. What if RAL's DualConduit cable could do the same thing via direct connection to the user's DAC?


The DualConduit's substantial construction makes it relatively stiff, but by slowly, firmly applying a steady torque I was able to manipulate the two Type A ends into a pair of USB mainboard ports on my server, and the other end into the USB input of my exaSound e38 DAC. It performed just fine, but when I compared it with the output from the SOtM via an AudioQuest Coffee USB link ($279), I heard some slight imaging instability of voices in the 3D soundstage of medium to large choruses. The sound was quite good, but it suffered in direct comparison to the SOtM-AQ route.

Of course, because both USB ports feeding the DualConduits are from the same motherboard, one might wonder: what value could there be in sending the same 5V signal over a separate cable, only to have it recombined with the signal lines at the other end? On the other hand, the Prophecy CryoSilver Reference DualConduit does provide an opportunity to supplant the motherboard's 5V with a better external supply, which I drew from yet another output on the HDPlex. Now those voices in the chorus were unwavering and distinct. The result of swapping the DualConduit cable over to the SOtM outputs in place of the single-line AQ Coffee was a toss-up.

Which approach I use will depend on the situation. I like having the SOtM for its accurate isolated clock, its external DC supply, and its outstanding performance with only a single high-quality USB cable. On the other hand, if your server can't support the installation of a dedicated audiophile USB board, RAL's Prophecy DualConduit with a good 5VDC source is another way to optimize your server-DAC link.

Signal HiFi SignalCable MagicPower power cord ($69/3'): Of all possible system cables, the one that I believe has the least potential to influence a system's sound is the AC power cord. In fact, I recall reviewing a power amplifier late in the last century and, at the insistence of the manufacturer, buying a specific power cord for the review. It bothered me to think why, if the amp performed at its best only with this cable, it wasn't included with the amp. As it turned out, that cord, despite its premium price and the amp maker's recommendation, made absolutely no improvement in the amp's sound. I keep that cord still, as a memento.


In my opinion, a power cord must be as passive as possible, and carry adequate current for all conditions. That is all. So, when AudioQuest installed a passel of AC conditioners in my system, they also brought along several of their most expensive power cables—which, they insist, are required for the optimal performance of their Niagara-series power conditioners. I have no way to dispute AQ, as the overall effect of this brute-force installation was thoroughly successful. In the end, I decided to buy the Niagara 5000, but soon realized that the impressively massive AudioQuest Hurricane power cord wasn't flexible enough to fit behind it in my rack.

What I wanted was a sturdy, 1m-long, 20A cord with a NEMA 6-20P three-pin plug and an IEC C19 connector to fit the Niagara 5000, and it had to be flexible enough to fit in the narrow space between the Niagara and the wall. I could get a Chinese-made generic for under $10 from several vendors but I imagined the Niagara spitting it out in distaste. I also tried any number of familiar audiophile brands, but few were willing to assure me that their product was flexible enough.

Then I stumbled on Signal HiFi. Their SignalCable 20A power cord is based on 10 AWG stranded, high-purity copper with hospital-grade Marinco terminations. The 20A connectors are optional at no extra cost. We exchanged a few e-mails to confirm that a 3' MagicPower cord ($69) would fit my needs. When it arrived, it was perfect, and for me it still is. I may buy some more.

Aqvox AQ-Switch-SE audiophile network switch
Power cords, RCA and XLR interconnects, speaker cables, S/PDIF, TosLink, USB—and now there's even more to bother with as we add networking, with its own cables, connectors, and switches. But the old controversies remain. On one side are experts who argue that network protocols ensure bit-perfect transmission and that timing and jitter problems are unimportant until after the audio datastream has been extracted. On the other side are those concerned about the audible effects of timing and data correction, noise, and galvanic isolation. I don't know enough to take a side. My experience has been that the networked part of my system works or it doesn't, and that the faster it is, the rarer the problems or failures.


As part of a recent renovation I upgraded my home network, which these days runs at 0.95GB from end to end, greatly exceeding the requirements of even discrete multichannel music. Nonetheless, I was fascinated by Aqvox's AQ-Switch-8 audiophile network switch (€398). Actually, I was as fascinated by the Internet buzz as by the descriptions of the product on Aqvox's website, which are rich in audiophile lingo but poor in useful technical information. Nonetheless, there are assertions of Aqvox's protection of its proprietary designs and access to the innards is prevented: "The enclosures are holo sealed and all modifications are covered with resin compound." I guess it piqued my curiosity—or my perversity.

I asked Aqvox to send me the AQ-Switch-8; instead, they sent their upgraded AQ-Switch-SE with matched power supply (€798), plus one Excel High-End LAN-/Network cable, 1m (€249) and six LAN-Detoxer RJ45 Caps (€99.90/set of six).

The AQ-Switch-SE looks like and is in fact labeled a D-Link DGS-108 8-port gigabit network switch, but it bears an Aqvox logo on the top and a foil-seal tamper warning on the side, and is distinctly heavier than the stock D-Link DGS-108 switch I bought to compare it with. Aqvox's power wall wart was also different, and larger. The instructions insisted that only the Aqvox supply be used, regardless of the source or specifications of any other supply. Aqvox also asked me to use the impressive Excel Network cable between their switch and the targeted output device, and that the LAN-Detoxer Caps be inserted in any unused RJ45 jacks. I chose to use the Excel cable between the switch and my Baetis Prodigy server; all other cables were Belden CAT6a. Installation and use were unremarkable, and there were no problems in operation.

I came to this task confidently doubting that modifying a network switch could make an audible difference, assuming it already met Ethernet standards—and indeed, my response to the first hearing was "So what?" It sounded fine, but different in no obvious way from the stock D-Link and ASUS switches I'd been using.


So: Experimenter bias confirmed. But to be sure that I was being fair to the Aqvox, I wanted swift A/B comparisons. That wasn't possible—the AQ-Switch-SE requires at least three reconnections and, by then my ability to hear subtle differences between the two would be lost. I had to settle for nonblind auditioning with the same set of reference recordings over the next three or four days.

I picked three files—recordings of unaccompanied chorus, solo acoustic piano, and orchestra—listened to each several times over (between other selections), and took notes. Then I switched the switch. I ran through this switching cycle twice; once with switches 1-2-3, the other with switches 3-2-1, but all cables remained the same, with the Aqvox cable connected to the server and Belden CAT6a links to the LAN port and the Merging Technologies NADAC Merging+Player. In the end, I came to no clear conclusion—with each change of switch, I usually heard, with delight, details and perceptions that I perceived as new. My notes seemed to reveal slightly less enthusiasm when I switched to the ASUS from the D-Link or the Aqvox, and slightly more when I switched to the Aqvox from either of the other two.

I'm not sure what that means in the context of my own system. I'll probably replace the ASUS switch with the D-Link, since the cost of the latter—approximately $30—is trivial. It wasn't clear to me that the AQ-Switch-SE didn't provide an improvement, and I'd consider getting one if the cost weren't so far from trivial. Aqvox's website is filled with praise from happy users, but for now I remain insufficiently motivated to join them.

supamark's picture

I agree they should include at least a 1 year (and seriously, for this kind of money, lifetime subscription - it's only $500 retail) since the box won't work without Roon. I'm sure they could get it cheaper than retail and fold it in to the price if their margins are so slim.

The other problem with this approach (which would disqualify it were I in the market for this sort of thing) is that you're buying a $14k (w/ tax) piece of equipment that requires software from a third party to operate. When buying high end audio you're always gambling that the company will stay in business and support the thing, but now you've got to hope *two* separate companies stay in business and support their things (just because a company is still around doesn't mean they support all of their products).

Kal Rubinson's picture

There's no assurance that any company will outlive you but the NADAC Player does not become a doorstop if Roon goes belly-up.

First, Roon has stated that, in the event that they close-down, they have a policy of maintaining a download version that will allow the user to continue using Roon forever without needing access to them.

Second, the NADAC Player can, without Roon, still operate as a network endpoint/DAC from any server.

Rlotzkar's picture

Nice series! I've been researching Network Renderers and almost went for NADAC style to avoid USB cables. I chose a different route (Bricasti M5 > Legacy Wavelet & HDPlex 200). So now I need to reconsider USB cables because only 1 of these devices uses a "normal" USB plug (Wavelet uses MicroB & HDPlex uses Type C). So the question arises, do USB cables make a difference...

I've used split cables (have 2 LHL 2G split cables) and can hear the difference vs decent quality straight cables. Now that I'm looking for a custom cable I'm curious about your experience with USB Cables. It seems to be a recipe for flame wars on various audio forums. Bits are bits and don't have sound...

So how did the AQ Coffee stack up against that RAL fire hose?

Thanks for the power cord referral, another one I'm looking for!

Kal Rubinson's picture

I am not particularly interested in comparing cables and only do so if there is a particular feature that seems to warrant attention. I tried the RAL because of it's split config. As I said, it was appreciated with the plain-Jane motherboard USB outputs but offered no advantage over the AQ when used with the externally-powered SOtM tX-USBhubIn.

adrianwu's picture

I have been experimenting with the Microrendu, which allows music data to be streamed over an ethernet based network to be converted to feed an USB-based DAC. The beauty of this gadget is that it allows a direct connection without the need of a USB cable. We have tried different Ethernet cables going from the switch to the Microrendu, and the difference is quite marked. We used a long cable (10 M) so the difference could have been more exaggerated than say a 1M cable. A shielded cable (Cat6a or Cat7) is certain preferable, with noticeable tightening of the bass impact and better imaging. The PS for the Microrendu also matters, and a DIY regulated linear supply gave a darker background and more relaxed, more musical presentation than the ifi wall wart.
We also had the 2 channel NADAC on loan for audition during several sessions over a week. Not impressed at all. Compared to the Lampizator Golden Atlantic/Microrendu, the NADAC sounded clinical, unmusical and uninvolved. This could just be our taste in music reproduction, but the Lampizator was better at expressing the microdynamics and the sound was fuller and had more nuance. Impossible to say which was more "accurate", but for enjoyment, we much preferred the Lampizator.

otaku's picture

Hi, I just bought a Signal MagicPower cable, basing my decision at least partly on this review. To my ears, my Bel Canto C7R sounds a lot better with this cord, which is surprising since the Bel Canto contains circuitry similar to their Virtual Battery Supply. What I am thinking is that rather than improving the sound of the receiver, the MagicPower cuts down on 60 cycle hum radiated from the power cord into the nearby speaker cables and interconnects, particularly the phono interconnects.