in ,

A stand-off down under: PureMusic vs. JRiver vs. Audirvana+

  • “Bits are bits”. It’s the number one mantra turned meme to infect mainstream thinking on digital audio. Number 2? “It’s just ones and zeroes, mate”. The corollary: if all binary data gets from point A to point B – and error checking mechanisms ensure it does – then data transmission is all that matters. Case closed.

    Not so fast, bucko.

    If points A and B are computers seeing a file being relocated from one to the other, the ‘bits are bits’ logic holds fast. So too when sending a print job out over USB. It comes unstuck when digital audio devices like D/A converters play receiver (point B) where the arrival timing of each bit matters; bit-to-bit mis…timings are referred to as jitter. Jitter causes D/A converters to perform sub-optimally, lending the resulting sound an emotionally distant quality. Look into its eyes: lights on, nobody home.

    Also influencing a DAC’s audible performance, perhaps to a degree greater than jitter, is the electrical noise (EMI/RFI) ingested by said DAC when umbilically connected to its source, often a consumer grade PC, to DAC. The latter’s internal oscillators, implemented to re-clock the incoming data stream, are extremely sensitive to electrical noise, which can easily disturb their timing accuracy.

    None too dissimilar to the mechanics of the same-sex marriage debate, there are detractors: those who not only wish to hold tight to the comfort of their world view that bits are bits and that’s all that matters, but would also wish metaphorical fire and damnation on anyone who sees things differently.

    For the sake of argument, let’s assume that electrical noise cross-contamination isn’t a factor in the quality of digital audio, that bits are bits and that’s all that matters.

    The implications of returning to this standpoint are not insignificant: 1) manufacturers of deluxe digital cables and noise filters are perpetrators of a scam large enough to make county fair-bought potions, that promise to cure arthritis and the common cold, look like they have a sound basis in science ; 2) that review sites and their staff are at best prone to the powers of suggestion or, worse, flat out liars; 3) that anyone buying such products have been totally hoodwinked.

    An exceedingly long bow needs drawing in order to assimilate the ‘bits are bits’ position.


    Compounding the issue is the pack mentality endemic to internet-based discussions. A mentality that kicks really kicks into being when said pack is confronted with something new that requires potential reassessment of current thinking.

    It usually unfolds thusly: A thread is started with a link to a deluxe cable or doohickey seeking opinions on the same. The first responder bowls in with a “that looks suss to me”-type comment. The ensuing pile-on is as ugly as it is predictable. The product is either rubbished in theory or the (perceived) absence of proof from the manufacturer pointed to as proof of its snake-oil status. “Well, this ‘ere world globe suggests the world is not flat after all…but it’s nothing more than a toy – BURN HIM!”

    Our mob’s next step is to ensure ongoing physical distance between it and the product that “couldn’t possibly make a difference”. Actual listening would bring it too close for comfort. As such, solicitations to try it for oneself are shot down. “Listening? Pah! That’s all part and parcel of the scam!” comes the snorted retort.

    Such is the combined power of mutual cynicism that it is invariably self-reinforcing. New products that don’t align with our cynics’ a) existing world view or b) wallet depth must be squashed underfoot lest sales figures gain momentum and another industry scam is perpetrated.

    Case closed – ironically – by closed minds. Ironic, huh?

    Let and live though, right? Sadly not. There are even digital audio extremists who would seek to correct anyone not sharing their particular point of view. Michael Lavorgna discusses this over at the always on-the-money AudioStream here.

    In 2015, the number one target for this kind of bully-boy treatment is the audiophile-grade Ethernet cable and to a lesser extent USB filters. Four or five years ago it was the humble USB cable and audiophile playback software applications, both of which are now more widely accepted as altering the sound of one’s digital playback, hopefully for the better.

    However, some skepticism still lingers.


    Proving that it counts open-minded skeptics amongst its members is Sydney Audio Club who last weekend undertook a whole lotta blind testing in the digital audio realm. Advance notice being Sydney specific, a post went up on DAR Australia last week. The results however are not bound by such geographical limitations.

    “When we announced the meeting format, there were people writing and calling to say we were wasting our time – the contention was that all music players sound the same. So why were we wasting our time?”, says SAC President Tom Waters.

    The skepticism didn’t end there. On the day itself, Waters’ opening remarks about there existing an audible difference (to his ears/mind) between software players brought smiles, puzzled looks and some throat clearing to the audience.

    Time to find out.

    (Before that a disclaimer of sorts. SAC isn’t a science lab. This event saw a bunch of chaps listening to different software players and voting on their preferences. Nothing more, nothing less. What follows should be dutifully ignored by readers insisting on the strictest of listening conditions. Moreover, ‘preferred’ does not immediately imply ‘best’.)

    With a mere 75 minutes at their disposal, forty or so Aussie audiophiles had convened in Epping to compare PureMusic, Audirvana+ and JRiver on a Mac. Keeping the field down to three popular favourites, even at the expense of Amarra, keeps focus tight. No need to include every option coming to market if the main focus is to prove/disprove a difference.

    The test system breaks down as follows: Kudos Audio C20 loudspeakers, Plinius SA100 power amplifier, Audible Illusions L3A pre, Bricasti M1 DAC, OPPO Digital BDP-105 disc player and a late 2012 MacMini. Loudspeaker cabling came from Nordost’s Red Dawn and interconnects were from Wireworld.

    All listening was conducted ‘blind’. No one other than Waters knew which application was being used for playback at again given time. The audience could only see the rear of the Mac’s display monitor.

    Two tracks – CS&N “Helplessly Hoping” (24bit/96kHz FLAC) and Ray Brown’s “Exactly Like You” (24bit/192kHz FLAC) – were played via each software app in turn.

    Asked simply if the audience heard a difference between options 1, 2 and 3, the audience responded resolutely in the affirmative. On matters of preference, a vote was cast by each member.

    The results? Player 1 scored 4 votes, player 2 scored 14 votes and player 3 scored 14 votes. A tie for first place between Audirvana+ and PureMusic with JRiver coming a distant third.

    [Side note: Waters himself abstained from voting – as he should – but he also believes JRiver on the Mac to be inferior to the Windows version]

    To alleviate any remaining doubt, Waters then suggested another round, this time with each player’s identity announced ahead of playback. Being served up this time ‘round, Mozart’s Requiem (24bit/192kHz FLAC), a comparatively short piece of full-on choral and orchestra. A casual vote saw preferences split between PureMusic and Audirvana+ with JRiver bringing up the rear again. Poor JRiver.


    Next up onto the blind A/B testing block, a comparison that I too have conducted at home: Steely Dan’s Gaucho in DSD vs. its 24bit/96kHz PCM equivalent. As Waters points out, provenance details for these downloads is in short supply. One can only assume they are derived from the same master.

    In my own listening tests with a Resonessence Labs Herus DAC, I heard the PCM version as having more bite on the air whilst DSD comes on softer, more relaxed. I preferred not one over t’other.

    How would SAC preferences tumble down?

    In voting, the DSD64 version notched up 18 votes compared to the high-res PCM version’s 17. These results underscore how an individual’s preference for hi-res PCM or DSD falls to a matter of taste rather than there being a out-and-out ‘best’.

    One more comparison before the curtain falls: Dave Brubeck’s Take Five is as well worn in audiophile circles as Jazz At The Pawnshop or Diana Krall. One advantage of such ubiquity is an ability to lay side-by-side a CD (from Austria’s Analogue Productions), a lossless rip of the same CD and a 24bit/192kHz PCM download of the same master. The latter two could be compared blind but the dropping a silver disc could not be properly concealed from the audience.

    Some fairly conclusive results eventuated: a single vote for the CD, 9 for the ripped CD and 13 for the hi-res download. That’s interesting. I would’ve expected the OPPO disc spinner to best the MacMini on Redbook playback alone. With the bits being read from the CD being the same bits as those ripped to the MacMini’s hard-drive, clearly there’s more going on than just binary transfer from transport to Bricasti DAC.

    The differences are most likely attributable to differences in electrical noise (as well as jitter). Noise inside a CD player’s less complex and audio-centric internals is generally thought to be less than that emanating from a consumer-grade computer like Apple’s MacMini.

    Noise is also pivotal to each software player’s SQ: 1) Caching audio into RAM before playback reduces the number of hard-drive reads during playback and so reduces electrical noise. 2) Sleeker software code can limit the need for the player to call on CPU clock cycles and again reduce associated electrical noise. iTunes and Windows Media Player are not designed this way. Neither nor Apple nor Microsoft streamline their code for optimal sound quality through noise minimisation.

    And who am I to judge if that matters not a jot to you?

    Further information: Sydney Audio Club

    John H. Darko

    Written by John H. Darko

    John currently lives in Berlin where creates videos and podcasts and pens written pieces for Darko.Audio. He has also contributed to 6moons, TONEAudio, AudioStream and Stereophile.

    Darko.Audio is a member of EISA.

    Follow John on YouTube or Instagram


    1. I don’t use Apple! So lets try the comparison on Windows and see which one wins?

      Seriously, If Jriver sounds better on Windows then maybe use the Windows version to compare? Or even with Audiophile Optimizer too.

    2. Hi John,

      Wow, the debate on bits sure becomnig si plorarized (damn it, I need to change this USB cable)

      A couple of months ago I followed an interesting discussion on a Devialet forum. Of course the Devialet is fitted with Ethernet natively, and for many owners it represents the way to go. Mention was made of various devices and tweaks to improve the sound of Ethernet, and some heard a difference and others not (link at bottom). Amongst them was mention of a fluid designed to improve the contact between plugs and cables, specifically ethernet. The fluid is called Stabilant 22 (find it on the web and at Amzaon). It reportedly improved the sound of untreated Ethernet. I mention this out of interest in its own right, but more so because it represents a further rationalisation (or ‘irrationalisation’, depending on your view) of the lengths people will go to to seek an improvement in sound quality.

      My point is that everybody has a ‘stepping off point’ for sound quality, i.e. the point at which they say “that’s it, I see no need to go any further”. I’m not sure you would ever buy Stabilant 22, so maybe your stepping off point hovers around the ‘swap cables’ levcl. Obviously, many others who share our hobby step off at the level before that, i.e. they don’t buy in to the digital cable thing. Most people with no interest in HiFi are very happy with their Beats and their Galaxy or iPhone, and probably think anyone who spends $3000 on a DAC is way past redemption. They step off way before anyone reading this post ever will.

      Me, I can’t, and never have heard a difference with USB cables. I’m now trying out an Audioquest Ethernet cable. I suspect I’ve reached my stepping off point, but it doesn’t stop me massively enjoying the set up I already have. I add that I understand those who report they do hear an improvement in digital cables and things. I can understand why they say that, and have no issue with it. If it rocks your boat, then great. But I offer my condolences to those people. You see, I’m really glad that I don’t have to worry about whether this digital cable sounds better than that one. I’ve spent more hours than I should have A-B ing this stuff, and stepping off here allows me more time to relax and listen to music, not my HiFi. That’s where we all want to be.

      Pity the person who ends up deciding the merits of different contact fluids.

      Respect, and thanks.


      • “Pity the person who ends up deciding the merits of different contact fluids.”

        Isn’t the person on the ‘stepping off’ level below you saying, “Pity the person who ends up deciding the merits of network cables.” ?

        • Hi John,

          Exactly, some of my mates think I’m nuts for paying $1500 for pair of headphones (I ain’t told them I’ve got two pairs at that price point). It’s another take on the ‘bits is bits’ argument. Paraphrasing that argument “anyone who thinks a USB or Ethernet cable makes a difference is deluding themselves”. If it works for you, and it is affordable, go for it.

          I’m also siding with Michael Lavorgna who has consistently stated that the enjoyment of music should be the ultimate consideration – there’s no right and wrong.

          Have a great day and be safe.


          • If 1) it enhances your enjoyment of music and 2) you can afford it, then have at it. Unfortunately, this kind of thinking seems to trouble some folk.

    3. I admire the enthusiasm of people like you or Mr. Lavorgna in trying to convince people what’s good for them. But, like I commented under Mr. Lavorgna’s article, you will never be able to convince people who do not want to be convinced. Even if you try to convince them of something as obvious as ‘black is black and not pink’, for example. That’s the power of the orthodoxy.

      You, guys, are wasting your time…

      • The thing is RobD, I don’t think myself and ML are trying to convince anyone of anything, just that we’d like to be able to report our findings – what we HEAR – without others trying to ‘correct’ those findings. However, I hear ya on the immobility and inflexibility of dogma.

    4. Well, I’m sure your “blind” test will be attacked for not being “double blind”. …
      John Swenson has written about measurable PC noise differences produced even when changing only the settings in the same playback software, not to mention playback from different software.
      But of course the same skeptics reject those measurements by saying it’s “impossible” they could change the resulting SQ….

      • If others don’t hear it, or hear it differently, all well and good. My distaste is for those who would proffer an opinion on the basis of a closed mind alone. To ride side-saddle with Michael Lavorgna’s pizza analogy: imagine you see a burger joint review on the web and then note that the chef is using only turkey mince. Not in a month of Sundays would your claim that those burgers “couldn’t possibly taste any good” because “beef is beef” hold up. One thing’s for sure, those turkey burgers will taste different.

    5. Michael Lavorgna, eh?
      I registered with ‘AudioStream’ in order to respond to a particularly patronising comment made by Lavorgna only to receive an email telling me I’d been banned from the site and my membership was effectively terminated.
      I had not used bad language nor made any insulting or disparaging remark but merely called him out.
      Is it any wonder some ‘audiophiles’ are ridiculed and rubbished?
      I only hope someone was around to pick up his toys and put them back in his pram. Sheesh…
      (Ooops…I hope I don’t suffer the same fate here…)

      • “Merely called him out” could mean a multitude of things and I’m sure there’s more to the story than per your precis. HOWEVER, picking over the bones of a disagreement between you and ML isn’t for here/now. As long as you don’t try to bring said disagreement to these pages all will be well.

        • Darko, in the past I have expressed myself as strongly on these pages without a dummy spit from yourself and for that I am grateful.
          ‘Said disagreement’ must have been trivial for I cannot even recall what it was about let alone drag it up now. However, it is curious that one would wish to run a site open to public comment, as a personal fiefdom where offense is so easily taken; what’s that expression about hot kitchens…?

          • Hmmmm. It has nothing to do with personal fiefdoms and (from where I’m sitting) everything to do with people wanting to parade a) how much they know and b) how much the author does not, but expressed in a way where the commentator attempts to ‘correct’ the author, often with dogma. It’s all in the delivery: “You’re wrong and here’s why!” is a mile away from “Have you thought about this issue?”.

            This is a generalised observation and most emphatically not directed at you.

            However, in my previous comment, I asked you politely not to bring it here. And yet you persist. I’ll ask again, also politely: *please desist*.

    6. Hi John,

      With the Brubeck disc playing, did the Oppo directly feed the AI preamp with its analog outs from its internal ESS dac? Or did The Oppo feed the Bricasti dac? If it did feed the Bricasti, was it using coax or optical? Thanks for the report….

    7. “Not so fast, bucko. If points A and B are computers seeing a file being relocated from one to the other, the ‘bits are bits’ logic holds fast. So too when sending a print job out over USB. It comes unstuck when digital audio devices like D/A converters play receiver (point B) where the arrival timing of each bit matters; bit-to-bit mis…timings are referred to as jitter. Jitter causes D/A converters to perform sub-optimally, lending the resulting sound an emotionally distant quality. Look into its eyes: lights on, nobody home.”

      Not so fast yourbuckoself, JD. Asynchronous reclocking is so good these days that the jitter issue is effectively put to bed. Who are you taking your lead from on technical matters? You give me cause for concern. Especially as you took on an educational role here.

      • I’m taking no issue with async reclocking. None, zero, zip. What I’ve said in the post is that jitter is one issue BUT electrical noise is perhaps an even bigger issue.

        If the jitter issue has been put to bed (as you say), why do re-clockers bring audible results? Is it because they 1) they reduce jitter or 2) provide some degree of noise insulation to the DAC…or perhaps do BOTH?

    8. It is disheartening but not at all surprising that the ‘bits are bits’ adherents not only reject the observations of others but also resist any and all attempts for them to try something new for themselves. Guys like Galileo and Copernicus were able to present reams of detailed measurements and observations but still got rejected and punished for challenging the world view. What I find exasperating is that when bit-fundamentalists are presented with observations that are inconsistent with their model those findings are airily dismissed as unscientific. A more appropriate response from anyone claiming to hold rational thought and scientific procedure dear is to seek out what is producing these results. One of my favourite quotes about scientific discovery is;
      “The most exciting phrase to hear in science, the one that heralds new discoveries, is not ‘Eureka!’ but ‘That’s funny…'”
      Isaac Asimov

      I definitely think it is ‘funny’ that better made ethernet cables are producing better results for many people so I do plan to buy some Audioquest ethernet cables to give it a go myself. I get it that a lot o people think that spending US250 on some cables is dumb or frivolous (or both) but then again nothing that is fun can’t be criticised in exactly the same way. Wish me luck as I head off on my personal journey of discovery!.

      A final thought before I depart;

      “They are ill discoverers that think there is no land when they can see nothing but sea.”
      ― Francis Bacon

    9. John, your inferences about how these bits are bits become b..i…ts , noise and jitter, sound like they are measurable. Do you know of any measurements that compare these bit sources and transmission methods? I’ve looked at different jitter measurements between transports and servers on stereophile without illumination.

      • Hey Brian. I don’t know of any measurements for the software mentioned here. As you might know, we at DAR have neither the money nor expertise to handle measurements. However, TNT host a good jitter primer here:

        Like you though, I only understand bits and pieces and assembling measurement results for an all-coherent bigger picture from which to infer audible performance is beyond my technical reach. But even if I/we could, per Srajan’s piece this week, what do those measurements of jitter and noise really tell us?

        Looking at Sooloos server measurements conducted by the world most technically proficient analyser of audio gear, John Atkinson:

        JA touches on the absence of compatibility problems with respect to the analogue output’s impedance. That’s absolutely useful intel.

        In interpreting the Jitter plot JA says, “…of more subjective significance are the lower-frequency sidebands that can be seen to either side of the central peak. These are spaced at ±86Hz and its multiples, the same frequency as the spuriae seen in fig.2. Again, this behavior may have resulted from the fact that the analog output jacks of Kal’s review sample were grounded to the chassis, which, Sooloos’s Rob Darling subsequently informed me, was due to a manufacturing error and is not typical of overall production. Even so, the calculated jitter level was still very low, at 254 picoseconds peak–peak, which makes it unlikely that this behavior underlay KR’s preference for the Source:One’s digital output feeding his Meridian digital preamp over the sound of the Sooloos’s analog outputs.”

        What does the reader take away from this? That the measurements do not align with the (listening) results of the subjective commentator? Of course, that’s not always the case – sometimes they align just so – but should we trust therefore not put quite so much stock in their importance?

        I believe John Atkinson is measuring the Antipodes DX server presently. It’s one of the best sounding transports I’ve heard to date so it’ll be interesting to see what tumbles out of that.

        On a broader note, I think what people are asking for when they request measurements on a piece of hardware is some form of visual proof. For some readers that means proof (or not) of a certain quality so that they need (not) listen for themselves.

        Now I’ll borrow from Michael Lavorgna’s pizza analogy: relying on measurements as an apriori decider on whether or not to order pizza is akin to asking for someone else to run a pizza through a blender and analyse its chemical make-up. That’s fair and reasonable when assessing the the pizza’s nutritional value. On how it tastes though, ya still gotta get off yr butt and go eat it.

        Are restaurant critics and food writers required to substantiate their claims of ‘salty’, ‘bitter’ or ‘sweet’ with chemical analysis? Not that I’m aware of. For me, the only bona fide useful measurements in this context are for E.Coli or similar. Is the food gonna make me ill?

    10. Thanks for the considered reply John. I’ll check those references. I was looking for something generalised between CD (real time) transports and media servers, to go with the listening test you mention that routed the transport. I’d like to rely on something visual/measurement but like that boy in the sixth sense, “I see dead people”.

    11. Stabilant 22 previously sold as ‘Tweek’ ‘contact enhancer’ a POLYOXYPROPYLENE—POLYOXYETHYLENE POLYMER. Now you know. My guess as a chemist (pharmaceutical not industrial) is a non-conducting material that seals out water-borne pollutants with oxidising potential.

    12. Hey John, that was a great bit of testing. I have always found Audirvarna+ to sound better on my two Mac systems. Finally deleted JRiver today, which also has a horrible interface and never comes out of beta on Mac. Cheers Robert

    13. Did the test also involve asking people if they heard a difference between tracks which had been played by the same software? (i.e. where there should not have been any difference within the bounds of the test). This would have been a very interesting control group.

      Also, presumably everyone taking part in the test could see how everyone else was voting?

      • Hey Tim. The testing methodology was as described, nothing more, nothing less. I understand people’s desire for double blind, control groups etc. but SAC’s aim here was to strike the middle ground between the (joylessness of) a strict scientific listening test and a casual playback session which is eminently more fun.

        • Surprised I was, as a happy jRiver PC user. But it’s undeniable, that in the MAC world (at least) it trails. By how much is a question. And realising thatJR /Mac is still a work in progress.
          And whether its UI superiority pulls it up.
          I have noticed a substantial improvement in JR PC with Fidelizer.
          The real test will be to run JR Mac against JR PC and I could do that . Eventually. But why, I’m having plenty enough goosebumps already. ‘Happy enough, not interested in possible improved SQ? Hand in your Audiophile Card.’

    Dieter Burmester, 1946 – 2015.

    KIH #25 – What do measurements really tell us?