From Munich to Denver with AudioQuest Ethernet cables


munich_2014If you’d have told me two years ago that some USB cables sounded better than others I might not have laughed in your face but I’d have certainly afforded you a quizzical look – not sceptical to the point of refusing an audition but sceptical to the point that it would have been too low on my list of priorities.

Eventually, I did take the time to audition a USB cable. And then another. They each sound better than a stock, printer-type USB cable and they both sound different to each other. And yet broader scepticism abounds. There are folk who “refuse to believe” (note the speculative language), the implication being that they’ve yet to sit down and, y’know, listen.

If you’d have told me at the beginning of this year that some Ethernet cables sound better than others I might have given you a similar response: that a) I have bigger fish to fry and that b) many readers really don’t like it when you try to tell them that digital cables make a difference. I’d learnt that from my excursions into USB.

On matters of Ethernet, even broader, more rigid scepticism abounds. There are many folk who “refuse to believe”, the unintended implication being that they’ve yet to sit down and, y’know, listen.

Steve Silberman in the AudioQuest hospitality suite in Munich 2014

Midway through my conversation with Steve Silberman in the AudioQuest hospitality room at this year’s Munich High End he brought up the topic of Ethernet cables sounding different. Was the collective audiophile consciousness’ refusal to believe why I tried to brush the subject away? Quite possibly.

“AudioQuest’s main business IS cables and of course Silberman would say as much – it affect’s his employer’s bottom line”, I reasoned internally. I concentrated hard on keeping such thoughts away from my mouth. Alas, my face must’ve given me away. Silberman was adamant that I attend their Ethernet cable demo upstairs. “Just listen – what do you have to lose?” asked Silberman. He had me there and he knew it – check.

Joining the throng was Stephen Mejias, fresh from the Stereophile boat. Here was a journalist I respected. Surely he too wouldn’t steer me wrong?

Time now for a point of order. Show coverage — as Srajan Ebaen of 6moons recently described it to me, it’s all sizzle and no steak. He’s right – it’s mostly impossible to draw conclusions on an individual component’s contribution to a system. Heck, even the room makes its own, sometimes unpleasant, contribution to final presentation. It’s why I rarely venture an opinion on specific bits of kit in show reports unless something is glaringly obvious. It’s also why my own reports are filed under ‘showcase‘.

How some commentators call out differences between DACs, sometimes with a comparison to the previous year’s event, I’ll never know. I do know that they are better/braver men than me. For me, audio showrooms provide a faint whiff of what each component in the rack might be doing but it’s only the collective performance that can be reliably assessed.

However, if the room is hosting a game of swapsies, it’s a different story. In such scenarios it’s easier to tease out performance deltas. Play a song or two, remove component A, insert component B and then replay that same song or two.

The AudioQuest system in Munich centred on a NAD M2 integrated amplifier, Elac 407 loudspeakers and Furman power conditioning. A NAD M50’s network interface received digital audio streams from a Synology NAS via a router. The M50 then fed a NAD M51 DAC.

All weekend AudioQuest swapped out Ethernet cables for anyone with sufficient curiosity willing to sit down and listen. That’d be my hypotheses and me. Null: there are no sonic differences between Ethernet cables. Alternative: there are some sonic differences between Ethernet cables.

AudioQuest demo space, Munich High End Show 2014

First up, a generic Ethernet cable – I don’t recall its make or model. In the context of the above hypotheses, it didn’t matter. I was only interested in there being audible differences between ANY two Ethernet cables.

After thirty seconds of music, a pause whilst the Ethernet cable was switched over to AudioQuest’s RJ/E Vodka cable. Price? US$255 and up. Play was hit on the same song again…

…chin, meet floor. There was a difference: more micro-dynamic attack, more intensity. Wanting not to believe what I was hearing I requested that the original cable be returned to the mix. Silberman and co. obliged willingly and sure enough, vividness took a step backwards. Null hypothesis rejected, alternative hypothesis accepted: sonic differences between Ethernet cables were real.

Before you conspiracy theorists jump in, allow me to play that role on your behalf. I’m not suggesting that the chaps from California had deliberately hobbled the control cable but let’s assume, just for a moment, that they had done exactly that. How would they have hindered its (audible) performance? With error-correction built into TCP/IP and UDP, any given cable would either work or it wouldn’t, no?

Furthermore, I realize that a single listening session isn’t wholly conclusive. However, my experience with Ethernet cables at the AudioQuest room in Munich was enough to see me further my investigation rather than close the door on it for good. With curiosity energetically stirred, I requested cable review samples from AudioQuest on the spot.

AudioQuest demo space, Munich High End Show 2014

Time passed, AudioQuest Ethernet cables arrived (and were scheduled for November listening), more time passed, I arrived at RMAF.

Attending the AudioQuest ‘Computer Audio Demystified’ seminar on Sunday morning the room quickly filled to capacity. People crouched on the floor and huddled behind the door. Our host again was Steve Silberman whose favourite phrase is “Do not confuse a clear path for a short distance”.

Whilst it’s plain that Silberman is a digital audio evangelist he’s also someone who doesn’t leave the obvious to chance. In his exposition of iTunes configuration he leaves nothing and no one behind.

The first forty-five minutes were dedicated to foundation level computer audio:

1) how to rip with iTunes – Silberman recommends AIFF
2) how to make iTunes sound better with Audirvana+ or Amarra
3) seeing your computer as an audio component
4) keeping your audio library on a separate drive
5) keeping that drive on a separate bus to the DAC (e.g. Thunderbolt for storage access, USB for DAC connection)
6) Solid state drives are a winner!
7) more RAM = better sound
8) quality of RAM matters – Silberman recommends for OS X and Samsung for Windows users. “Avoid Kingston”.
9) higher processing power computers tend to sound better
10) Gordon Rankin (whose custom code AudioQuest use in their Dragonfly DAC) says laptops tend to have lower RMI/EFI than desktop PCs and therefore tend to sound better

That’s a solid amount of information that even the most experienced computer audiophile could sink their teeth into.

Steve Silberman presenting ‘Computer Audio Demystified’ at RMAF 2014

A quick demo of the (very) audible benefits of Audirvana+ 1.X over vanilla iTunes playback set the scene for what I’d come to hear: an Ethernet cable swap. This was to be Round 2! A network-connected dCS digital front end, Simaudio by Moon amplification and Vandersteen loudspeakers formed the basis of the system.

“Streaming is where digital audio will land”, reckoned Silberman. A quick A/B revealed to the audience that a MacBook direct-wired to the router with a bog-standard Belkin cable sounded ballsier and grippier than hearing the two connected over WiFi. “See? ANY cable is better than wireless!” proclaims our host.

The Belkin Ethernet cable was then swapped out for an AudioQuest Pearl variant (US$25 and up). This change afforded Talking Heads’ “Sax and Violins” more on-the-nose dynamic punch and revealed more surface textures. A sense of “WTF!” spread through the room. Mind-bending stuff.

On paper, Ethernet cables shouldn’t matter. In reality to anyone in attendance that day in Denver, they do. For this commentator, here were results from two different shows, each with different rooms and different hardware configurations. On both occasions I’d witnessed a difference between Ethernet cables. My findings match those of Audiostream’s Michael Lavorgna. It would seem that EVERYTHING in the digital audio chain matters. Even the stuff we think couldn’t possibly matter, matters.

I understand why levels of scepticism run high and will continue to run high. Unless you’ve heard it with your own ears you probably won’t believe it either. A shade over a third of the 300ish people that voted in last week’s poll reckon that audiophile-grade Ethernet cables are pure snake oil. Conversely, two thirds are either a) open to the possibility that Ethernet cables can sound different or b) are already convinced that they do.


With this audio show coverage I’ve described the ‘what’ – my initial experiences in hearing a difference between Ethernet cables. For some readers, that will be a confusingly bitter pill to swallow. (Some would rather knock the pill from the offering hand).

The ‘why’ is beyond my ken (for now at least). That’s for engineers to investigate. Feel free to discuss possible reasons why in the comments section below but to keep this conversation on point, refuseniks and ‘snake-oilers’ are directed to the comments section in the first instalment.

Stephen Mejias and recently corresponded on the matter:

“Here’s what I’ve learned since joining AudioQuest: Wired Ethernet offers uncompromised data rates at distances up to 328ft for Cat7 and 180ft for Cat6, avoiding the precipitous drop-off in data rates common to wireless transfers. At a distance of only 20 feet, some wireless routers lose as much as 85% of their data throughput. These losses in throughput cause dropouts in your media, higher-variable compression rates, and an overall loss in audio fidelity. As you increase the number of devices on a wireless network, you decrease the potential bandwidth available to each device.”

“Even generic Ethernet cable offers audio performance that’s generally clearly superior to the fastest, most robust wireless networked-audio system. Moving beyond the limitations of wireless systems, the specific cable matters, too, of course. Cat6 cables support network speeds up to 1Gbps; Cat7 cables support network speeds up to 10Gbps at 100-meter lengths. All of the cables we sent you are built on the Cat7 foundation.”

At home with the three cables sent to me by AudioQuest is where we’ll next pick up this conversation. Stay tuned.

Further information: AudioQuest ‘Computer Audio Demystified’ white paper | AudioQuest retail price book

Written by John H. Darko

John lives in the NOW + HERE = NOWHERE. He derives an income from the ad revenues of DAR. John is also an occasional staff writer for Stereophile, 6moons and TONEAudio.

Twitter: DarkoAudio
Instagram: DarkoAudio
Facebook: DAR


Leave a Reply
  1. Should try local, Brisbane specialist cable manufacturer . They blow Audioquest out of the water.

    I use VUE – VU2 USB cable with very modest system Rotel RA 1070 integ amp, MacBookPro17″. Calyx 192 DAC, Pure Music software and a pair Quad book shelf speakers, very modest indeed, previously had Pass Lab amps with B&W 802 Diamond speakers, so I am familiar with good sound. The difference the Vue cables made was quiet unbelievable, aided with assistance of the Calyx and Pure Music software, although system is modest at least I get clean sound, and very listenable when combined with HD Tracks HD music, however, not all HD music is the same. Anything is better than crappy CD recordings. Have also tested the Calyx Femto DAC, in a word fantastic, and my first / introduction DAC was as HRT Music Streamer +11, that sold me on DACs.

  2. Hi John. I AM a believer since I do believe and trust my ears as well as you do. But zi read an interesting comment from a disbeliever yesterday who stated: does the internet work with audioquest ( or other brands ) audiocables? This paradigm keeps me as a believer awake and allergic to snake oil audio business… I respect Audioquest products and their solutions for the current problems we still encounter in digiyal audio playback. Our human ears seem to be much more sensitive to jitter or other very difficult to measure effects. The magazine the absolute sound recently tried (and to my opinion failed) to measure these audible effects properly. But audio and music is my passion and I have encounteted other unexplaibable, but audible phenomena in the past as well, like demagnitising CD’s and applyi g green marker on the side of a cd etc… so somewhere in the transfer from digital to analog minute errors are made, which seem to be audible. Nowadays, our digital data are tapped online from the internet using a streamer or offline using computers and harddrive / nas solutions. What fascinated me is that ‘we’ audiophiles do not seem to care or understand what might happen with these digital data before it flows into our nas or hdd. Even more fascinating is that the internet and all streaming audioservices apply a ‘cut and paste / torrent’ method of and suddenly our music sounds correct..
    So what is the real problem we are dealing with in our digital audio playback systems..? I cannot understand that all these errors are solved by improving the interface / cables only. There must be a more fundamenl way to deal with and solve these errors completetly. To my opinion, the solution is there, but it is not in the interest of many of these respectable companies to solve it. Ovrr the last 20 years, the costs of processing a gigabite of data has dropped enourmously. We now see HD tv and 4K tv using videoprocessing technology at much higher and more complex rates than with high-end audio gear. It all happens within a few microchips usi g buffering technology. First in firtst out ( FIFO ) algorythms. So I assume a true solution for a 100% puristic processing of digital audio via internet lies within the core of a well designed microchip. One complete 16/44 cd contai s not mire than 800 mb of data. The speed of streami g and processing such data is much less than a 4K video stream.. So we just need a company who will understand this and solve our little frustrations once and for all. Sorry for all cd player and cable manufacturers, but I believe the solution to stream all audio not only bit perfect but also harmonically petfect either wireless or via the existing (non audiophile) cable networks in or soil is near.

      • Hi John:
        Important question. Do you know WHAT the control cable was? Cat5, Cat6, Cat7?

        If if it was Cat5 or 6, ALL bets are off, and it was a bit of a sham.

        Audioquest ethernet cable is Cat7 and Cat7 that is mission critical approved can be had for 1/10 what audioquest is charging.

        • Hey Andre – In Denver it was a Belkin but I’m unsure if it was Cat5/6 or not. Can you link us to cheaper Cat7, please?

          • Hi John:
            I can tell without a doubt it was not Cat7. With all due respect to audioquest, and I think they make some nice products, their cable demos are a bit of a…well..I won’t say it.

            See link here for various brands of Cat7 cable on amazon. There are dozens of other sources as well. I I can tell you that some of these cables have been tested for mission critical applications by the US military.

            Also, I really don’t think Mr. Mejias’s quote about data loss is based on actual data, unless he can provide that.


          • I also just “features” that quite honestly just seem ridiculous for an Ethernet cable. For instance on the AQ website, they claim:

            DIRECTIONALITY: All audio cables are directional. The correct direction is determined by listening to every batch of metal conductors used in every AudioQuest audio cable. Arrows are clearly marked on the connectors to ensure superior sound quality. For best results have the arrow pointing in the direction of the flow of music. For example, NAS to Router, Router to Network Player.


            SOLID 10% SILVER CONDUCTORS: Solid conductors eliminate strand-interaction distortion and reduce jitter. Solid silver-plated conductors are excellent for very high-frequency applications, like Ethernet audio. These signals, being such a high frequency, travel almost exclusively on the surface of the conductor. As the surface is made of high-purity silver, the performance is very close to that of a solid silver cable, but priced much closer to solid copper cable. This is an incredibly cost effective way of manufacturing very high-quality Ethernet cable

            What kind of bugs me is they are marketing these cables like analog cables, and I think, I don’t know for sure, that they feel this is the only way they can entice audiophiles.

            Now, I will tell you for a fact if these cable improve the sound over my industrial grade Cat7 from Rosewill, and other trusted brands, i will eat crow. 🙂

            But I am not willing to fork over several hundred dollars to find out at this point.

  3. From what I have been reading so far, like analog RCA and speaker cables, Ethernet cables are a “necessary evil” (discarding wifi) and to very lazily copy/paste quote Roger Skoff of “THE VERY BEST CABLE POSSIBLE IS NO CABLE AT ALL”. His caps btw.
    Some streaming devices do give us the “no cable at all” option in the form in internal memory, SD card etc.
    Data does not need to traverse a network – its right there in the box and that *should* mean zero signal degradation (im open to suggestions why this may not be the case). If this IS the case, audio played via the streamers internal memory should be better on all accounts than the best Ethernet cable money can buy (again, correct me if im barking up the wrong chicken).

    Before I upgrade I would like to be convinced that there is degradation occurring with what I currently have with me here.
    I would like your/the greater internets thoughts on whether an A/B test of two identical files, one played through an ethernet network and the other loaded into the streamers internal memory will suffice to demonstrate any audible shortcomings.

    Cheers! B

    • Hi Bruce, intuitively I agree with Roger Skoff’s statement . It seems obvious that stored data streaming from a large RAM buffer, which is closely interconnected to the DAC IC, will sound better than stored data on a NAS streaming through relatively long distance , even via the best ethernet cable available . If this approach is not feasible for technical reasons, I will be very interested to understand why.

  4. There is now no doubt in my mind after rising to your challenge last week.

    I was very sceptical but open minded and only one A to B comparison as per your experience with Audioquest was enough to convince me of the differences. I suggest a big portion, if not all, of the snake oil contingent hasn’t even bothered to try it.

    I don’t understand why this is the case but I’m a qualified engineer with both electrical and mechanical background and to my knowledge I’m not mad or delusional and can tell if one cable sounds ‘different’ to the other simply by listening.

    Think of it another way, how smug are we to actually try it, be surprised and confused at the result but enjoy the benefits.

  5. I currently use generic Cat cable and the results have been dramatic. I have switched from Cat5 to Cat6 to Cat7 and have never looked back. The difference from Cat6 to Cat7 was amazing. More depth, more dynamics, more EVERYTHING! I hope to be able to eventually afford to replace ALL my streaming devices (PS3, Laptop and two Blu-Ray players) with Cat7. Then move on to upgrading the quality of the cable itself. All the cabling in my system is AudioQuest and soon to be Ethernet.

  6. I attended an Audioquest demo at a HiFi show in Windsor, UK, in 2013. The Ethernet cable from a Linn streamer to a hub was swapped from a standard Cat5 or Cat6 cable to an Audioquest low range Ethernet cable and back and forth a few times to make sure the audience could hear a difference.

    I was unprepared for the surprise. Ethernet is a packet switched network, how could there be a difference? But, to my ears and to that of everyone in the audience (there were no naysayers in my demo) the system sounded more detailed, spacious and natural with the Audioquest cable in place. I am sure the same effect could be felt with other brands, but kudos to Audioquest for daring to host a demo to illustrate that there is, indeed, a difference.

    Quite why this should happen is beyond my ken. But it worked. So, if I one day have my music on NAS and need Ethernet, I’ll be trying aftermarket cables myself. Anyone who says ‘it can’t work therefore it isn’t working’ should patiently review the evidence and revise their opinions!

  7. I stream hires music from a NAS to a Macbook laptop, and depending where I am in the house, I do this through hispeed wifi or ethernet cable. I use Audirvana+ as the playback software, which allows me to set a memory buffer for playback. My music is then not streaming in real time from my NAS directly through to my USB cable & filter, reclocker and DAC, it has been buffered already in the computer memory. I don’t notice any difference when going between wifi or ethernet cable connection in this case, and I check every aspect of my system to the point of being anal. I will need PLENTY of listening tests to convince me of the merits of an audiophile ethernet cable for my setup. Fortunately, a friend is an Audioquest dealer… 🙂

  8. I didn’t think this was worth trying because I only have basic Ethernet cables, but I eventually got around to seeing if I can hear a difference.

    I recently acquired a secondhand Linn Sneaky DS which comes with a 3m grey Cat5 cable which I’ve been using for the last few weeks.

    I swapped the Linn Cat5 cable for a 1m Dick Smith Cat6 cable that I had connected to the Apple TV. The Cat6 cable was 1-2mm thicker and less flexible than the Cat5 cable.

    The difference was pretty astounding. There had been a midrange mush previously with the Sneaky which had now mostly disappeared. It also sounded like the treble was rolled off previously, now it was the opposite, it took on an almost tinny sound in the top end. The difference wasn’t subtle. Where I previously found the Sneaky to be smooth and non-fatiguing it now bordered on edgy, harsh, and fatiguing. The sound could probably be described as “better”.

    After listening for a while I decided to plug the Apple TV back in using the Linn Cat5 cable. All of sudden the harshness disappeared and the sound was somewhere between the original sound and the harsh sound.

    Ultimately I prefer the sound with the Apple TV plugged in because it takes some of the harshness off the sound!

    I suspect the chipsets used for the Ethernet and Wireless receivers in network streamers are not designed to the same standards as say a DAC chipset. There probably isn’t a large enough market for Ethernet chipsets specifically designed for audio applications, but it appears it could be of some benefit considering even plugging another device into a router can affect the sound.

    • Just a small update.

      I decided to remove everything else from the router (I had a TV and network receiver plugged in), and the harshness that I noted earlier was gone, so it wasn’t in the Dick Smith cable per se.

      However the sound became dull and uninteresting. I thought to reconnect the Apple TV using the Linn Cat5 cable and the sound took on a richness that reminded me a bit of a valve amp I used to use. I’m not sure whether it was introducing some subtle 2nd order harmonics or whether it was acting like a filter, but I much preferred the sound with it plugged in. Note that the Apple TV is not in the playback chain, just hanging off the router.

      I’m intrigued how affected the sonics are by connecting various cables into the network. Should network streamers be more isolated than this?

      Btw – thanks for bringing up the topic. The Sneaky is sounding a lot better.

  9. This is false information!
    “Cat6 cables support network speeds up to 1Gbps; Cat7 cables support network speeds up to 10Gbps at 100-meter lengths. All of the cables we sent you are built on the Cat7 foundation.”
    Per International standard ISO/IEC 11801 Information technology, Cat5e supports 1Gbps (100mhz), cat6(250mhz) and cat6A(500mhz) support 10Gbps. Proposed cat7 is for 10gig over 100 meters and rated for 600mhz. Category 7 is not recognized by the TIA/EIA.
    Most home routers and switches are 100mbps so you will need to upgrade them too.

  10. I definitely have to try the Audioquest diamond ethernet. Its actually reasonably priced as far as top of the line audiophile cables go.

  11. So where are people inserting these better cat6 and/or cat7 cables?

    My hifi is on a different floor from my cable company router. My cat5 cable comes from the router to a Linksys SE2500 5-Port Gigabit Ethernet Switch in the room w the HiFi. From the switch, one cat5 to my Music streamer, one cat5 to mac PC, and one cat5 to NAS.

    I can guess to change out the cat5 in the hifi room, but I doubt I’d pull high end Audioquest Ethernet cable through my house.

    Would changing just the cable in the hifi room make that much difference? I’m seeing parallels between home AC wiring and just changing that last six feet w a high end AC cable.


    • In reply to Dave

      I have a budget Chord Cat 7 Ethernet cable between my streamer and gigabit switch with a standard Cat 5 between the switch and the NAS and this was enough to generate differences.
      Luckily the ‘difference’ was very favourable which is not always the case. I’m not sure if changing the Cat5 between the switch and the NAS will make any difference but will try this in the new year.

      So for now your parallel with the mains wiring would seem to hold true.

  12. John,
    My first experience that USB cables matter was with a Microsoft Digital Sound 80 in 2000. One of the guys on the USB design team gave me replacement cables for it and said “try these”. The cables he gave me made the largest improvement of any cable change I’ve ever made. So I’ve known about USB cables can make a difference for a long time. I found Digital Audio Review in early 2013 when I was looking for a DAC for my office. Even the Modi benefits from having better quality copper wire in its USB connection. I didn’t go overboard but WireWorld Ultraviolet was a satisfying upgrade. When I purchased a DAC for my east coast system I started with the Ultraviolet then moved up to Wireworld Starlight. I experimented with a silver plated copper USB cable but found it too bright. I can safely conclude that for my sonic tastes improving the quality of the copper and shielding in USB cables can be worthwhile.
    As for Ethernet cables, the results are mixed. There are a lot of cables out there that barely work for audio but still may function for data transfer if not as well. There are good reasons why Fluke DTX-1800 Cable Analyzer costs $9,000 US, you can make sure your network wire is functioning properly. There are a lot of problems with copper cabling. It is not uncommon to find slightly flawed Ethernet cables that are not good enough to carry the packets properly. I visited a client last month in another state and he turned on The Trishas. They didn’t sound right and after our meeting we traced the problem to an Ethernet cable just slightly flawed. He swapped it out and the improvement was obvious.
    When you first reported you were blown away by an Ethernet cable my first thought was the comparison was not level matched. Blown away is often explained by a .3 decibel difference in the volume. I didn’t think any more about your experience in Munich until I moved west last month and needed to think about my own Ethernet cabling. One of the first things I examined in my new environment was the possible sources of noise in Ethernet cabling. These are: florescent lightning, heating ventilation and air conditioning systems, radio, TV and electronic devices. I borrowed a Fluke DTX-1800 from a client and had him make me two CAT 7 cables, bought a pair of CAT 7 cables from a local company and ordered two AudioQuest Pearl Ethernet cables online. I verified the cables with the Fluke and tested each pair carefully level matching each pair. I could detect no audible difference between these cable sets.
    Next step was to monitor the quality of service in the network with a PRTG Network Monitor. The jitter with the CAT 7 cables above is a little less than my when network was wired with CAT 6 cables so quality does matter and it can be measured. Whether audiophile grade cable is necessary to keep jitter inaudible seems to be an open question.

  13. John, Steve Silberman, nice chap that he is, is misinformed on many aspects of computer audio, and sadly spreads this misinformation to people who don’t know any better. I attended his RMAF “Demistifying Computer Audio” seminar at RMAF I recall. I was unimpressed with some of his statements, particularly in relation to different brands of RAM and Hard Drives sounding different. He relates this as fact, and fails to offer any kind explanation or reasoning. NOT done when dealing with computer science! He claims there are many things “we don’t know”. I don’t believe Steve has any formal computer science qualifications. A very poor reflection on the audiophile community at large.