“Nobody in the history of audio (pro or consumer) has ever advocated for the addition of noise!”
These words, from AudioQuest’s Steve Silberman, reveal a major motivating factor in the development of the company’s JitterBug, a USB noise filter released in June of 2015 ($49 USD). Silberman’s point is that if no one believes that introducing more EMI/RFI noise into a digital or analog playback system is a sonic benefit, shouldn’t the corollary hold true? All else being equal, shouldn’t the reduction of such noise result in better sound?
Noise is the enemy of good sound, and the JitterBug’s raison d’être is to reduce EMI/RFI from both the data and power sides of USB transmission -- on the power side, both the 5V DC USB voltage bus (aka VBUS) and its associated ground. If the JitterBug actually does reduce noise without introducing new problems of its own, it would seem, in theory, to be a sound (sorry) idea, especially for $49. Still, some people -- many of whom have never actually listened to the device -- have called foul: “heresy,” “snake oil,” “marketing BS,” “separating a fool from her money,” etc.
That’s right -- there are those who claim that, particularly in USB-based digital audio, all that matters is a signal that is “bit perfect.” From that point of view, the only meaningful threshold is basic competence of digital delivery. According to holders of this worldview, strongly promoted by a vociferous few on the World Wide Web, its truth is easily demonstrated: If two different USB cables can pass the competence test and deliver all those perfect bits from host to peripheral (e.g., a “competently” engineered, USB-capable digital-to-analog converter), then forget anything you might hear: There cannot be any audible difference between the USB cables. If you do hear a difference, it must be the result of confirmation bias, the placebo effect, or some other effect that can be psychologically explained away.
Were this true, the audio world would eventually recognize that any competently designed and manufactured USB cable by definition can have no sound of its own, and everyone would simply buy a Belkin Gold Series USB cable for $10 on Amazon. Extrapolating from that, if “bit perfect” is all that matters, it wouldn’t matter what brand of computer you use, which operating system (Windows, Mac, Linux), which hard drives, which playback software (Amarra, JRiver, Roon), or any choice within any other variable in the digital chain -- as long as the DAC receives a bit-perfect stream of ones and zeros. After all, a competently engineered DAC will reclock the bitstream and feed its internal converters that same bit-perfect stream. It is from this perspective that devices such as the AudioQuest JitterBug are reviled. Reduce noise? Can’t make a difference.
Not everyone agrees. Ever since my first foray into computer audio over a decade ago, it has been my observation that quite a bit of the system preceding a USB DAC can make considerable, even profound differences in the sound, despite a bitstream that has been confirmed as “bit perfect.” This is true even of USB DACs that don’t tap into the VBUS and/or galvanically isolate the USB input receiver components from the DAC’s D/A and analog output sections. As for the AudioQuest JitterBug, I was intrigued to see what a low-cost device the size of a thumb drive might do in my systems.
A thingy. A whatchamacallit. A doodad.
From Steve Silberman I received a ziplock bag containing a small handful of JitterBugs, enough to let me test in several different systems, both in the signal chain and in a parallel capacity. I also took a close look at the JitterBug’s standard slick packaging, which does a nice job of describing the device’s intended effects. A thumb-drive-sized device wrapped in a heavy-dutyish plastic cowling, the JitterBug is quite small: about 1.5” long (excluding connector), half again as wide, and a little less than half an inch thick. Still, it’s big enough to prevent two JitterBugs from sharing most immediately adjacent USB ports. There are no lights, bells, or whistles. Other than the case itself, all you see is a USB Type-A male connector on one end and a rubber-covered slit with a Type-A female connector on the other -- exactly what you’d expect to see in an in-line USB device. When I cracked open the case, I saw that these connectors are mounted on a narrow, rectangular circuit board covered with traces and tiny resistors, capacitors, inductors, and chokes.
The JitterBug is a purely passive device. It is not an active USB hub, reclocker, and/or signal regenerator along the lines of SOtM’s tX-USBhub, Schiit Audio’s Wyrd, or UpTone Audio’s USB Regen. Compared to those devices, the JitterBug’s three-pronged mission is relatively simple: 1) preserve the data integrity of the audio bitstream flowing from computer to USB DAC by eliminating parasitic resonances, USB jitter, and packet errors; 2) lower the noise on the VBUS conductor pair (power and ground); and, similarly, 3) filter out from the USB data lines all EMI and RFI.
The JitterBug meets the specification for a USB 2.0 High Speed device with a maximum data rate of 480Mbit/s. This is critical information for two reasons. First, it defines a threshold for the data-line filtering -- the JitterBug targets noise suppression only above the data-transfer frequencies, to prevent unintended violations of the integrity of the data itself. Throwing out the baby with the bathwater would do no one any good, and would undermine Prong 1 of the JitterBug’s mission. Second, because the JitterBug is an inline device to be inserted between host and peripheral in a network otherwise capable of higher transfer function (e.g., between USB 3.0 devices), the system will recognize the 480Mbit/s limit and accordingly reduce the flow of data. Not that this should matter for any USB DAC -- I’m unaware of any music-data transmissions that require more than USB 2.0 throughput -- but if a JitterBug is inserted between the computer and a USB 3.0 hard drive (why would anyone do that?), the system will recognize the AudioQuest as a gating item and lower the transfer rate to match it.
Made in the USA
In speaking with Gordon Rankin, chief designer of the JitterBug, I was surprised to learn that it’s made here in the US, primarily of North American-made parts. Only the plastic case is made in China. Rankin pushed for domestic manufacture for multiple reasons, and AudioQuest found a factory in Ohio that uses state-of-the-art assembly procedures and equipment, and offers levels of efficiency and precision that, in the end, cost about the same as would outsourcing manufacture to Asia. AQ’s DragonFly and Beetle DACs are also constructed at this facility. The circuit boards and electronic devices, too, are all locally sourced. The prices of such devices have been reduced to commodity level; apparently, they would cost essentially the same regardless of country of origin.
In designing the JitterBug, Rankin built 14 prototypes. After the first few, he set up a test board on which dozens of circuit alternatives could be quickly connected, tested, and listened to. As the circuit design approached finalization, another board was built that allowed for evaluating various inductors, chokes, resistors, and capacitors, some with pots to facilitate fine tuning. With Rankin charting the course, and using more than 30 different USB DACs for multiple-data-point testing and listening, the feedback given at key junctures by Silberman (Vice President of Development), Garth Powell (Director of Power Products), Joe Harley (Senior Vice President of Marketing and Product Development), and AQ’s founder, Bill Low, helped guide the development of the JitterBug.
Given AudioQuest’s consistent advocacy of optimal directionality in signal cables and power products, I wasn’t surprised to learn that signal directionality was designed into the JitterBug. I was surprised to learn that AQ drilled down to even the smallest parts, in one instance requiring the maker of one of the JitterBug’s inductors to run custom batches with directionality indicated. Such attention to detail in design was continued in assembly, an area in which local manufacture provides benefits. Every JitterBug is subjected to optical testing to ensure proper assembly and orientation, and each is also tested electronically before being inserted in its case. Being mere miles from the assembly line, rather than half a world away, enables real-time oversight and responsiveness, from questions about which solder to use to any of the myriad other day-to-day realities of production.
Deploy, deploy, deploy
Silberman encouraged me to experiment, and, with my bagful of JitterBugs to play with, I was ready . . . but not before being given two friendly reminders. First, while recommending that JitterBugs be used in parallel to the subject data line, AudioQuest cautions they should not be used “in series”: that is, don’t plug one JitterBug into another and expect an improvement in sound quality. This made sense to me -- I’ve long understood that independent filters, when used in series, can create unintended interactions that adversely affect the sound. Second, AQ recommends that no more than two JitterBugs be connected to any single USB bus. Fair enough.
So what did I do first? I tested the “in series” warning, of course! My music servers are dedicated CAPS Zuma computers, and include SOtM tX-USBexp PCIe-to-USB soundcards. These specialty cards include an ultra-low noise regulator for digital, clock, and USB power circuits, along with power-input noise filters, PCI express slot filters, and an active noise canceler for clock power. With the JitterBug inserted in the SOtM soundcard output, the sound got considerably worse. While not unlistenable, it had picked up a harshness that was a big step -- or five -- back. It took only the opening bars of Al Green’s “Tired of Being Alone” (Al Green Gets Next to You, 16-bit/44.1Hz FLAC, Hi Records) to hear the adverse impact the “filter-on-filter” effect had on the otherwise warm, fulsome tone of Green’s voice. JitterBug removed, and along with it any harshness. Warning validated.
What about a parallel deployment on this same server? The SOtM USB soundcards have two power-related options. For DACs that don’t require USB power, an Off switch on the card can be toggled to deactivate the VBUS and stop all power from going to the DAC. Because neither my Ayre Acoustics QB-9DSD or Wavelength Crimson USB DAC requires USB power, Off is my default setting. As a result, I expected to hear no difference when I inserted a JitterBug in the server’s other USB outputs. But on plugging a ’Bug into the unused USB port on the server’s faceplate, I heard an increase in the music’s sense of ease; everything sounded slightly, if sometimes only very slightly, more relaxed and more solid. Surprised, I spent the next day or so inserting and removing a JitterBug from that and various other parallel USB outputs on the server, as I swapped out DACs and changed USB cables. The change I heard remained consistent. The only logical explanation was a reduction in noise residing in and/or from the computer’s internal power supply. It wasn’t a marked improvement, but it was consistently there.
The fact that a slight audible benefit would result from the mere insertion of a $49 JitterBug in a server’s unused USB output made me reconsider using the second power-related option available via the SOtM soundcard. This card has an external power connector for bypassing the computer’s noisy power supply, allowing it to be served by an external supply such as SOtM’s own Intelligent Battery Power Supply, which I have now ordered (unfortunately at many multiples of $49). So much for the specialty server.
I next deployed the JitterBug with more typical partnering gear: several Apple Mac Minis (various vintages) and my MacBook Pro (2013), with either a second Ayre QB-9DSD or my Ayre DX-5DSD. The results were in every case positive. The biggest improvements always occurred when one JitterBug was connected in series between the Apple computer’s USB output and the USB cable feeding the DAC. A secondary improvement, perhaps 20% the magnitude of the first, occurred when I inserted a second ’Bug in an unused USB port on the same computer. Again, these changes comprised a greater sense of ease and solidity. For these comparisons I relied on Ola Gjeilo’s “North Country II,” from the 2L-TWBAS 2012 Sampler (24/176.4 FLAC, 2L/SoundStageRecordings.com), a perennial favorite impeccably recorded and intimately familiar. The weight and presence of Gjeilo’s acoustic piano and the accompanying flugelhorn and cello, the image stability and purity of tone -- all were enhanced whenever JitterBugs were in use. Consistently making the sound more engaging, the Mac Mini produced greater benefit than the MacBook Pro -- perhaps because, whenever I use the latter for audio, I run it on battery power.
When I consciously focused on what happened as I inserted and removed JitterBugs, what was immediately obvious was the sense of solidity. Over longer listening sessions, and days during which the configuration of JitterBugs remained unchanged, it was the sense of ease that became more evident. Those moments when I connected with the music emotionally came more quickly and frequently. The improvements wrought by the JitterBug(s) were most evident with files ripped from recently released CDs and downloads of popular music that have all the telltale signs of high compression. In one such test I alternated between a high-resolution download of “Say Goodbye,” from Beck’s Morning Phase (24/96 FLAC, Capitol), and the “EQ The Vinyl Exp” version of the song (320kbps MP3, Capitol), made available via a download card included with my LP. Like other commentators, and despite the resolution differences, I greatly prefer the Vinyl Experience version, presumably for the apparently greater dynamic range of that mix (the Vinyl Experience file, reportedly the result of a needle drop, evinces a different set of mastering decisions between vinyl and digital releases). My preference remained, though by a narrower margin, with the JitterBug in the playback chain. Both files sounded better with the JitterBug, but the more relaxed nature of the sound disproportionately benefited the more heavily compressed high-rez track, ameliorating the other differences between the versions. For me, the inability to deeply connect with music from a digital source is the great barrier that separates it from music reproduced by an all-analog playback chain. Did my digital front end, with the JitterBug(s), come close to what I experience with vinyl? No. But it took a positive step in that direction.
Finally, I deployed JitterBugs in other network locations. I plugged one into the USB jack on the front panel of my Ayre DX-5DSD (intended for the playing of files from memory sticks), and one into my network’s remote Synology NAS. I was never able to conclusively identify any positive sonic difference, but as I also heard no degradation of the sound, this is an area where further experimentation may make sense.
Nothing (much) to lose
In a hobby in which chasing down even marginal sonic gains while avoiding deleterious tradeoffs can cost hundreds, thousands, even tens of thousands of dollars, a $49 device -- or, better yet, a $98 pair of them -- that can effect varying degrees of audible improvement is a no-brainer. My trials of the JitterBug indicated that, as long as AudioQuest’s few warnings are heeded, there are no downsides, and consistently positive results of varying degrees. Even if, ultimately, the JitterBug produces no improvement in the sound of a particular system, $49 is a minimal risk, and provides an engaging opportunity to experiment with and listen to your digital playback system in a new, simple, and potentially rewarding fashion. C’mon -- what have you got to lose?
. . . Peter Roth
AudioQuest JitterBug USB Noise Filter
Price: $49 USD.
Warranty: One year parts and labor.
2621 White Road
Irvine, CA 92614
Phone: (949) 585-0111