AQ Jitterbug Impressions

Over the past few days I have had the opportunity to try a jitterbug in two of my systems. I thought I would share my subjective impressions. Please understand minor difference were heard and described but for many these may not be so apparent or appear as large, you know we tend to exaggerate scale since we have no way to reference.

My headphone setup is first and it is modest at best, nowhere near as revealing or open as my speaker setup. And the cans are just decent midfi by standards here, but I have them equalized and like them for my listening purposes. System consists of Win laptop with Foobar playing aiff files from a fire wire connected drive into a Halide USB Bridge which goes into a non DS MHDT Havana dac (borrowed from Effusion) then into the Asgard 2 amp and then the Mad Dogs.

First test was to drop the Jitterbug into an adjacent usb port next to the bridge. And I tried it on two other USB ports. Not sure if I could hear a difference between the different ports, but the conclusion was the same, Bad! It really made the music sound lifeless, dull and kinda like a poor mp3 file. Enough of that.

On to the Jitterbug in front of the Bridge. Using the Live Alison Krauss album beginning with “Faraway Land” and progressing into Jerry Douglas doing “A Tribute to Peador O'Donnell / Monkey Let the Hogs Out” was a good test. Great recording, good hall ambiance, clean vocals, nice reverb, well defined instruments and signature Dobro sound are the hallmarks. My headphone system does not even get close to my speaker system and much better cans would help here I am sure, but it does okay on these tracks. The Jitterbug reduced a fair amount of the Dobro overtones as well as hall reverb and made it sound like it was a bit bassier than without. It also presented the sensation of a blacker background. The focus of the guitar in the soundstage was locked but you lost a good amount of hall ambiance and things did not breathe as well as they did without. Some folks have mentioned that in their systems it sounded like bass got more pronounced and vocals came bit forward. I noticed a bit of that as well, but think it may be more a reduction of highs that makes hearing the bass and vocals easier. For me on this system, the Jitterbug just sucked too much life, air and upper treble out of the music to be an addition to my system

I am really at a loss to explain why a USB device that is supposed to reduce USB noise ended up affecting audible upper frequencies in the converted analog out, but hey I will let the measurement folks see if hey can explain it. I do know that it is not dissimilar to what the Wyrd did with my HRT dac. That combo made highs compressed sounding and removing it allowed them to come back. And the Wyrd with the Modi 2 uber actually sounded pretty nice with the Wyrd with a slightly recessed midrange being the most noticeable change along with some tightened or bigger bass.


Another thread here points to a “what's best” forum post that shows some measurements of the Jitterbug. It shows very little noise as compared to a direct usb connection under 24k but shows lots of new added measured distortion all the way up to 130khz. While all of these spikes are well below -100db maybe they can affect bits that end up at the dac chip and how that affects conversion. I can only surmise that maybe what I am hearing is similar to what I heard when I had a number of switching power supplies in my main audio systems power circuit. After removing a printer, non-audio computer, router and more from a wall outlets on the same circuit as the audio system and placing them on separate one, I noticed that soundstage width and highs opened up. Maybe this noise is like that RFI and is actually impacting something that eventually affects, sound. Well, enough of my trying to explain it, I hear what it does.

On to the main speaker system. Much more resolving, and unfortunately for the Jitterbug the Matrix X-Sabre is being fed by a power legless usb cable so most of the gains from supposed tweaks to USB are not as easy to come by. The Wyrd did absolutely nothing, pro or con, no difference. Placing the Jitterbug into a port adjacent to the usb cable had ta similar effect to the main system. Things were not quite so lifeless but it seemed to be a negative here as well. Hall ambiance shrinked a lot, overtones were damped, yet one might come away saying they liked it. Violins became less steely and more wooden for example, but in other cases vocals became more smooth and rounded. The sound is definitely different, quieter guitar notes just got lost, and Alison's signature breathy vocals became more simple and rounded. And some midrange body was lost in the guitars as well, almost like the volume wa turned down, so instruments gort a little leaner. Yes the result of this test was not good in my opinion.

Running the Jitterbug in line with the powerless USB cable presented a slightly different picture. Almost the same result as with the Jitterbug in a separate port with a bit more highs being evident and less compressed sounding overall. End result, no thanks!

How this noise reduction device affected highs and spatial cues, got me thinking about another possible test. Since I was hearing an impact on the music here, what would happen if I sent ones and zeros from a CD to a hard drive but first through the jitterbug. Yes, external DVD drive connected via usb cable into the jitterbug then into the computer. I ripped a new Agnes Obel CD with and without the jitterbug. How can this be? results were similar to what I heard with the dac, only on a much smaller scale.

I can see how some folks might like the effect the jitterbug has on their system. To conclude then in my systems, the Jitterbug made audible differences in presentation. It smoothed vocals a bit, made bass appear slightly rounder, reduced treble overtones, air and spatial cues and made the background appear more black or quiet. It also sucked a fair amount of life and vitality out of the music. If you like what it does, great, you may wish to also try it on your external drive and CD rips

 

Picked up Wolvebain's AQ Jitterbug recebtly for some comparative listening (thank you kindly sir!).

Using the ultra revealing SPL Phonitor / Sennheiser HD800 combo (with parametric EQ -3dB @ 5500 hZ to somewhat offset the peak in the HD800's frequency response) these are my impressions:

- NuPrime uDSD (bus powered): an improved sense of space, better separation and focus, blacker background, superior attack and decay, slight edge taken off upper mids, making the music seem to "flow" better. If this was my primary listening DAC I would be able to justify the expense on the Jitterbug quite easily.

- NuPrime uDSD (8000 mAh battery pack powered): This is where it gets tricky. The improvements heard with the Jitterbug are virtually indistinguishable from the benefits of the Y-cable data/power split. Jitterbug in/out ... inconclusive. It did not detract from the music. But also I am not so sure it improved anything significantly. Many more hours of comparative listening on a wide variety of music will be required to make any meaningful conclusions.

- Wadia 121: Virtually indistinguishable. Certainly not a clear cut affair. Again much more listening required.

There you have it after about an hour of back and forth comparative listening ... my subjective observations, to my ears, in my system, using a small selection of music, and maybe reaffirming what I wanted to hear.

What I am prepared to put money on is that:

1. In some applications the Jitterbug has clearly audible benefits
2. This is more pronounced in bus-powered applications
3. And it is probably made irrelevant by more sophisticated DACs

Would love to hear more opinions.

 

As we say in South Africa ... Ja ja.

I did not say most resolving - did say ultra revealing. The HD800 to me remains a reference for transparency, imaging, transient response and detail retrieval. Far more than any other pair in my collection or what else I have heard. That to me is revealing. And to my ears no other amp I've heard shows off these technicalities better than the Phonitor. Another tick for revealing. So I stand by my comment that that the combination is ultra revealing (not most resolving ... don't know where you got that).

And now that we've turned this back form yet another Phonitor bashing thread, do you have any constructive comments regarding the Audioquest Jitterbug to share?

 

My experience so far runs more or less along the lines of the other posts above. The other day I stopped by the local shop to listen to the Audioquest NightHawk and Meze 99, and while i was doing that, the sales dude feverishly tried to sell me the AQ Jitterbug, promising i'd enjoy sheer audio nirvana. I gave in and picked a unit, as it was only 50 euros. I looked at the little thing with contempt and plugged it in with a smirk, expecting it to make everything worse. My on-the-go setup is MacBook 12 > Asus Xonar U7, powered over USB > modded ATH-m50x . And, what do you know... It did sound better... in some respects. There was initially some extra sense of clarity, bite and overall impact. However, there was something funny going on. If I can use TV analogies, the overall contract of the picture was improved but it was a little less nuanced. You know, like those LED TVs with dynamic contrast set to "showroom" mode. The contrast is wow, but when you look closer you realize the details are masked, because shades blend in.

When I went to the office, I had to test with my better setup. That's the modded HD650 with the cobalt magnets, Etheraudio PCM58 based dac and my 6n1p-ev single ended tube amp. With this setup the "contrast" difference was gone, I was left only with the lack of nuances between shades. There's less micro-dynamics, less detail, less bite. Overall, I'd describe the sound as rather grey and boring. Not a significant difference, but you don't have to look for differences. They're immediately apparent.

I don't think this gadget works as a jitter solver. Quite the opposite, I honest believe it introduces more jitter, as it simply smears up the picture just a little bit. I even tried it on my speakers system and the issue was even more apparent. What it does seem to do well, is to stabilize the voltage (to some extent) and filter out power line noise (to some extent). So, if you're using a USB powered dac/amp, there are some benefits.

I just don't understand all the rave about the Jitterbug in all these reviews I read today. My only logical explanation is that my PCM58 dac's USB receiver has some weird circuitry that supposedly is dealing (somehow) with jitter. Whatever it's doing, I guess, it's just doing it better than the Jitterbug. Perhaps other people are using equipment where the bug does have some positive effect on the sound.

Bottom line, I got what I paid for. A 50 bucks gadget won't notably improve the sound quality. However, if you're running a USB powered dac/amp combo off a laptop, it may sound better to your ears... or not. YMMV.

 

It shouldn't reduce jitter, directly - at least not in the DAC - since you're talking USB and there's no clock on the USB connection that is relevant to the DAC's output or jitter numbers. Data either arrives fast enough or it doesn't - and it's received into a buffer and then clocked-out of that, by a clock in the DAC, and fed to the converter IC. So, if it does affect the DACs jitter(I believe AudioQuest posted some measurements at some point that suggested it had a minor effect on jitter), it'd be down to noise on the wire causing issues with the clock in the DAC.

It's essentially just a passive filter (caps, common mode chokes). Oddly it is applied to the data lines as well as the power lines ... which, in theory, would smooth out the differential levels of the signaling voltage ... and that's not something, as an engineer, I'd really want to be doing.

Here's an interesting experiment (one I've performed myself, but it'll be more useful/enlightening to try it yourself). Using a fast external USB storage device, do a large file copy and time it. Unplug the drive. Reboot the PC. Add a Jitterbug to the connection to the storage device, and repeat. Compare the results. You can draw some interesting conclusions from them.

Once you understand what's going on there* ... it's worth bearing in mind that USB Audio 2.0 does not have any kind of packet re-transmit capability. Data either arrives correctly or it doesn't. There is a CRC value that lets the receiver know if the data did, in fact, arrive as sent but there's no ECC data there, so all the receiver can do is decide if it wants to treat that data specially or just take it as-is. It can't deterministically correct it though, as it has no way to know what the data should have been.

Now, this is not to say that Jitterbug cannot have a positive effect ... it does filter noise. But whether the cure is worse than the symptoms is going to be highly dependent on a number of variables.

...

For the hell of it, and because I was getting a bit tired of just bitching about USB Audio (and figured I should, at least, take a quick stab at seeing how simply it could be made better) I put together some cables that a) separate the data and power lines into separately shielded, discrete, optimized and to-spec wire runs b) allow the power line to be severed from the source and/or supplied from an external low-noise supply and c) include both an EMI/RF filter and a passive power-line filter** ... neither of which touch/affect the DATA lines.

If you're so inclined, these are not difficult to put together (maybe a bit fiddly) and the parts aren't expensive; and they provide an interesting contrast to the Jitterbug (and interesting effects when combined with something like the Wyrd). For portable use (which is the only time I personally use USB audio, unless specifically testing it), they don't need an additional dongle, which is nice.

There are, of course, multiple products, cables in particular, that do the separate power/data line thing. Some are sanely priced, some are rather silly. Building your own, basic, version should be meaningfully cheaper than even just the Jitterbug. All should be more useful with something like a Wyrd or a Lightspeed Revive in the chain ... assuming you have a DAC that's worth fussing with the inputs on in the first place.

*In interpreting the results - if performance drops, you're getting bit-level data errors - which, contrary to popular belief, USB Audio 2.0 cannot correct.
**A bit harder to do yourself in a very small form-factor unless a) you're proficient with electronics and b) able to hand-solder SMT parts reliably.

 

Great post, thank you. I wondered what's inside the Jitterbug and even fancied some thoughts there could be some dedicated oscilator inside. So, there's no special tech, accurate crystal and low noise on the signal and power lines remain the only way to reduce jitter. Interestingly, I have made an experiment in the past by running the USB wires a few turns into a ferrite ring. The more turns, the worse the sound was getting. I'm not an engineer, but a keen hobbiest, and never sought out the answer, but this has always baffled me, as the USB signal is differential. You said "in theory, would smooth out the differential levels of the signaling voltage ... and that's not something, as an engineer, I'd really want to be doing." Can you elaborate on "smoothig out" a little bit? I think you have the answer to what baffled me.

I never separated the lines on a USB wire, but your comment is encouraging and I should try it. I'll test the jitterbug with my external SSD. I've actually experienced lower transfer rate with poor chinese USB in the past.

 

With a theoretical perfect differential setup, you'd expect there to be perfect cancellation of the field from the positive leg by the opposite offset from the negative and, thus, there'd be no net magnetic field induced and the core would have no effect. The problem is that it's NOT a perfect differential setup. Spacing between conductors and the core will vary and that will change the relative strengths of the induced effects. Multiple windings will exacerbate the effect in two ways ... first, it'll likely upset the cable geometry a bit (especially with very flexible USB cables) and second, there will be multiple interactions between signal and the core. The effects are small, but cumulative.

If you want to understand more about the limitations of differential signaling with a simple two-wire twisted pair, you can read up on Star-Quad cable geometry - which exists to make practice closer to theory!

That's in relation to the noise filter circuit, rather than the ferrite (though in practice it'll be apparent in both cases). In short, the structure of a noise filter of the nature in the Jitterbug is to smooth out variations in the voltage on the wire (noise and/or variation in source voltage stability). They do this, in hugely simplified terms, by trying to attenuate higher voltages than desired and by "topping up" any lower-than-spec voltages. That's great for a power feed, but on an alternating high/low signal, i.e. the analog voltages that represent the digital data, you don't want to be trying to "even out" the values to a single, consistent, level. You'll either "round off" the edges of what should be a perfect square wave or alter the actual voltage differential, which can cause issues with reading the signals correctly ... and higher speeds make it worse.

 

This is very interesting. I am aware that due to imperfections you can't have an ideal choke. However, I always assumed the impact of these unavoidable imperfections would be rather small. Yet, every next turn into the ferrite deteriorated the sound audibly, so I always thought there must be something else at play. However, I can take what you claim easily, for a good reason. People often laugh at me for wasting money on 0.75mm 0.999 silver (which is real hard to find) wires for my USB cables. However, it does sound better, even though the properties differences between silver and copper are small, and rationally it shouldn't make an audiable difference.

Okay, I get this. So, even though the filter is supposed to "cure" external noise, it does actually damage the signal a little bit. I'm curious what this filter is. In my experience placing a passive filter in the signal path always led to deterioration of sound quality. Is there some voltage regulator on the power lines?

Thank you for taking the time and effort to expand on the topic. I don't have the time to become more engineer-savvy, so to speak, due to kids, work and so on, but I am really enjoying any interesting information.