I got an early start on computer audio. At the end of the last century I was using WinAmp with first a CardDeluxe PCI soundcard, then a similar card from RME, to play files on a Windows PC. After I became a MacPerson, I used FireWire audio interfaces from pro-audio company Metric Halo and an inexpensive USB-connected ADC/DAC from M-Audio. But it was with the USB version of Benchmark's DAC 1 that the computer began taking over from physical discs for my music listening. At first I used iTunes au naturel, but as I acquired more high-resolution files, I began using Pure Music to handle all the tedious audio housekeeping, assigning as a dedicated music server a G4 Mac mini I'd bought in 2006.

When Apple stopped supporting G4 machines in 2012, I bought, on the recommendation of Pure Music's Rob Robinson, a new Mac mini fitted with a 2.7GHz Intel i7 processor and 8GB of RAM. This I now run "headless"-ie, with no keyboard, mouse, or display-under the control, via our home network, of my MacBook Pro. I've used the mini as the primary source for all of my equipment reviews since 2012; it remained continuously powered up for almost three years without any problems until, one day this past June, it died.

According to the Apple Genius Bar near the Stereophile office, not only had the mini's internal power supply failed, its logic board was toast: It might take a week to repair it, depending on the availability of parts. Assuring the genius that I'd backed up my data with Apple's Time Machine utility, I left it with him and returned home to a week of playing CDs on my Ayre Acoustics C-5xeMP universal player.

Enter the JitterBug

This is when things got weird. Every disc I played sounded more solid, more corporeal than I remembered from the many times I'd played the same tracks from the Mac mini. The low frequencies sounded more authoritative, much as they had when I bought a Mark Levinson No.31 CD transport (now long since retired with mechanical problems). In hindsight, the system's sound with the computer-audio source now seemed consistently more ethereal. With one exception.

Before the Mac failed, I'd been experimenting with a sample of AudioQuest's JitterBug that I'd been given at the 2015 Consumer Electronics Show. The JitterBug, a small device costing $49, resembles a USB memory stick in size and appearance, except that it has a female Type-A USB port on the other end from the usual Type-A male plug. It is intended to be plugged into the computer's USB port, with the USB DAC plugged into its other end. According to AudioQuest, there is also a sonic benefit if a JitterBug is plugged into any unused USB port in any device that's connected to your system-even ports on devices that aren't in the direct signal path, such as a NAS or a router. However, AudioQuest warns against using more than two JitterBugs on one USB bus or in series.

Under the hood

The JitterBug is a purely passive device. All of its components-small numbers of surface-mount resistors, capacitors, inductors, and what appear to be common-mode chokes-are carried on a small, multilayer printed circuit board connecting the input and output ports. This board is housed in a plastic case just wide enough to prevent two JitterBugs from being plugged into the adjacent horizontal USB ports on my 2011 MacBook Pro. The case is also slightly too thick for two JitterBugs to be plugged, side by side, into the vertical USB ports on my Mac mini. (For my listening, I used a short USB extender cable to plug the second JitterBug into one of the ports of both computers.)

A USB connection comprises just four electrical contacts. The outer pair carries 5V DC (VBUS) and the supply ground; the inner pair carries the data signal in an antiphase, packet-based format, one element of this pair labeled Data Positive, the other Data Negative. When the positive data are at 3V, the negative data are at 0V, and vice versa, so that any noise contamination in the line will be common-mode and thus be rejected by the data receiver. The signal level doesn't change for a logic 1, but the signal level is inverted for each change to a logic 0. Each packet consists of a sync byte, followed by a packet-identification byte, then the data, and then error-checking bytes calculated from the data's content. If the error checking detects a problem in the packet's transmission, the receiver asks the source to resend that packet; otherwise, the source is asked to send the next packet. The USB receiver stores the incoming data in a temporary buffer before forwarding it to its destination.

AudioQuest says that the JitterBug conditions both the 5V voltage connection and the biphase data connections, restricting the connection's radio-frequency bandwidth to that appropriate for USB 2.0's maximum specified data rate of 480Mbits/s. As a result, there are said to be improvements in signal/noise ratio and reductions in jitter and parasitic resonances.

Listening

Back in October 2011, Stereophile spun off AudioStream.com, edited by Michael Lavorgna, to cover computer audio; earlier this year, Michael enthused over the AudioQuest JitterBug: "49 bucks buys you some of the soundest improvements I've heard for anywhere near the price." He later wrote, "I've now heard the JitterBug do its thing . . . a number of times in different systems and I easily heard sonic improvements each time." I have learned to trust Michael's ears. Even so, I wasn't expecting the degree of improvement the JitterBug wrought with the first recording I played, comparing the sound with and without the 'Bugs, via my newly repaired Mac mini. (The following comments are an amalgam of my experience with all the DACs I used for the comparisons.)

The low frequencies acquired more of the authority I experience with CDs. A song that has stuck with me for more than three decades is "Wallflower," from Peter Gabriel's fourth solo album, Security (original UK CD, Charisma 800 091-2). This very early (1986) CD suffers from the ills endemic to the primitive digital transfers of those days, but I've found that every improvement I make in my system pushes those ills a little further into the background. Listening to the ALAC file ripped from the CD, the dramatic tom-tom strokes that punctuate the buildup to the song's anthemic refrain sounded more solid, better defined in space and time, with the JitterBug. Similarly, with Phish's cover of Little Feat's "Time Loves a Hero," recorded live in 2010 (16/48 FLAC, downloaded from www.livephish.com), the drums and bass guitar both had more impact, and were better differentiated in the mix.

But it was in the midrange that the JitterBug most strenuously announced its presence. In my review of GamuT Audio's RS7 speaker elsewhere in this issue, I mention Leonard Shure's performance of Schubert's Piano Sonata in B-flat, D.960 (24-bit/96kHz WAV files). Antony Tommasini of the New York Times had written, of the CD of this performance (Audiofon CD 72010), that Shure's piano sounded clangorous in some of the most impassioned fortissimos. Without the JitterBug, yes, it does; with the JitterBug, the piano sounded more natural, more smooth in these passages.

The same was true with Martha Argerich's live performance of Rachmaninoff's Suite 2 for Two Pianos, with Gabriela Montero (ALAC files ripped from CD, Warner Classics 623594). Not only did the two pianos sound more liquid in the midrange with the JitterBug, there was a better sense of those pianos as physical objects/instruments in space. Similarly, the applause sounded more like hands clapping than generic noise.

These specific comments were all gleaned from my listening to the JitterBug in my big rig, with the PS Audio PerfectWave DirectStream and Bel Canto Black DACs. To check that I wasn't imagining things, I repeated much of my listening using a first-generation AudioQuest DragonFly and a Meridian Explorer, primarily with Sennheiser HD-650 headphones.

Overall, I would say that the JitterBug eliminated what I described as the "ethereal" character of computer audio, bringing it very close to what I experience from spinning CDs.

Conclusions

You can see from the "Measurements" sidebar that I could find no significant effect that the JitterBug had on the analog signals output by three of the DACs I had to hand. Yet with those DACs and others, I heard an improvement in sound quality that I can attribute only to the JitterBug. I hate when that happens!

Last June, Michael Lavorgna wrote that "measurements obviously have no direct correlation to enjoyment." I have no hesitation in declaring, loudly and longly, that I can think of no way to spend $49 that would make me more enjoy my computer-based audio than the AudioQuest JitterBug. Try one-or two-for yourself.