I am using Amarra's 2.0 software to run my Mac Mini Server into a dCS stack. I went this route for two reasons: 1) i thought the sound (codec? Process) of ITunes was not very good; and 2) I had invested so much time into loading all my music into ITunes that i wanted a software that works with ITunes.

Sometime in 2011 I began thinking about the possibility of doing a review on high quality audio players for the Mac OS X operating system. Even though Windows users still represent a big part of Headfonia readers, I choose to focus on OS X for now because it’s easier to do as I use Mac myself, but also because for most developers, OS X seems to be the platform of choice when it comes to high quality audio playback.The last few months, I’ve tried and tested the majority of available audio players for OS X. In total, I’ve tested a total of nine players: three from Sonic Studio (Amarra, MINI, and Junior), Audirvana and Audirvana+, Pure Music, Fidelia, Decibel, and BitPerfect. Comparative reviews are always difficult to do, especially when you have more than five different products in one article. After I started working on this article, I realized that it’s quite a crazy project to take on, but I realize that if I only did a comparative of three different products then it’s only a matter of time before someone come and ask “how does this compare to that?”.

Many of these players offer such a rich level of control and customization, and it is beyond the scope of this review to try to cover every single function offered by each player. I realize that despite clocking in over 7,000 words, the article is far from perfect, but it’s been a crazy few months keep so finally I decide to publish this article and get on with working on more normal reviews. Anyway I hope you guys will find this to be helpful. There are a lot of things I missed in terms of covering each player in detail, but if you’d post it on the comments section I will try to fill them in.I’ll start with the table of contents for those of you who want to skip a section or two. Table of Contents. General Facts About Audio Players –.

The Sonic Studio Players –. Audirvana Free and Audirvana Plus –. Pure Music –. BitPerfect –. Decibel –. Fidelia –.

Common Settings Parameters –. Classification, Final Thoughts –.

Appendix: The Concept Of Neutrality –General Facts About Audio PlayersI have had the chance to use these players with some very high end equipment and indeed the improvements I get from these players are good enough to warrant the purchase. True to the “garbage-in garbage-out” principle, the app you choose for music playback represents an important part of the “source”.

Any distortion created at this level is going to be passed on to the subsequent components on the system, and so it’s very important to get things right.Of course the whole “Is it worth the money?” question is going to be the first thing people ask. One thing I need to state before going to the reviews is that the sonic improvements that you’ll get from these audio players are very subtle. Undoubtedly the hardware quality is more important than the software, so a $300 DAC is still going to be better than a typical $100 DAC despite the player used on the computer. This is why the pricing of the players is going to be a big part of the decision. Take a guess? Take 6 guys and ask them to draw a simple mountain.

How’ll the drawings look?What I’m saying is, I don’t know how each of the designers optimized the data transfer path. I know that they’re all trying to minimize jitter. But clearly they all wrote their own line of software, and the different approaches would be one possible explanation for the different sound.Again, I never thought that an audio player could have a sound signature. I’ve always been an iTunes user and I thought that was fine.

Until I start comparing these different players. I think what he is trying to say is that he doesn’t see how a player that is bit-perfect can deviate from neutral, as timing only becomes an issue once the signal is clocked, the data whooshing around the PC’s innards being constantly buffered until it reaches the audio interface.The arguments I’ve seen to explain this inconsistency mainly make reference to effects on the power supply, which don’t really make much sense with reference to the design of PC power distribution and the ATX12V spec.I’m not trying to stamp on any parades, but I really don’t see how these can work. Surely when the mechanisms involved are so exotic some of the companies making large sums out of this (look at the price of Amarra!) can actually produce some measurements to show the jitter reduction capabilities of their software with a modern PC?. From my point of view, I think we’re all looking at this from a simplistic point of view.

For instance, when you say “timing only becomes an issue once the signal is clocked, the data whooshing around the PC’s innards being constantly buffered until it reaches the audio interface.”I mean, we don’t make a living designing Operating Systems, and we have no idea what’s happening behind the pretty GUI.I find that from my point of view, the easiest thing is to do is to just listen to them, and that’s what I shared on this article. Except that the reason for “who cares?” is that we may not know the actual reason for many years. In the past, there have been plenty of things that had noticeable effects, but the reason was not yet known.But it is senseless to suggest that we should all wait years to find out “why” before making our next purchase decision.BTW, your jitter discussion avoids the inconvenient fact that those buffers get full. Aside from that, this is similar to the 1980’s “bits are bits” discussion, and in both cases, mistakenly assuming that consumer equipment perfectly achieve their specs, rather than merely coming “close enough to work most of the time”.

My jitter discussion does not “ignore” that: buffers getting full has absolutely no bearing whatsoever on jitter.I’m not “assuming” anything either: bit perfection is not something that is difficult to achieve. My point is that bit-perfect players (and if you’re suggesting something like Foobar is not bit-perfect the developers would like a chat) must, by neccessity, deliver the same data to the soundcard’s hardware buffer. If this was not the case computers would be unable to function in even a vaguely reliable and predictable way: they wouldn’t really function at all.This leaves the only mechanism whereby differences can be caused as one of the power supply, which seems very unlikely.I think it is reasonable that companies selling things utilising unlikely mechanisms provide some modicum of proof that their stuff actually does what it says on the tin.I also take issue with the idea that audiophiles heard issues in the past before science caught up. I can’t think of any recent examplesfrom jitter to feedback, the idea that audiophiles heard it before it was understood is more a myth than reality.

I tried the demo of Amarra 2.4 and found it a little buggy indeed. Adding a song in the playlist make the listening song to lag half a second, the control of the windows is odd, Amarra doesn’t respond for few seconds when clicking on it to move the main window, exiting the program clear the playlist, etcTried Audirvana + also and instantly understood what you said about Amarra black background. The sound of Audirvana is different, more aerial and still a bit more crisp, but a little less detailed. Yet the program works like a charm.So it’s Audirvana for me.Well still looking for one with playlist management and I would be in paradise.Tested with Audinst HUD-mx1 and (bad) Koss UR-40 (I really want to change those). Hmmm, I’ll get the only free one (Audirvana free) and give it a shot.Now just being a curious soul, I do wonder how audio players can possibly have different sound sigs. I’m not saying they don’t as obviously Mike’s able to tell them apart. But does having different sigs imply that they all essentially apply different EQs to the music?

In a perfect world, (yeah, I know) all the different players do should be opening up the same audio file and sending the same bit-coded music digitally to the same DAC/amp/headphones. Honestly though, I tend to believe that a good player should be truthful (neutral) to the music file that it opens and only offer the options of EQing your music as you like, instead of imposing some sort of EQ on you without asking whether you’d want it or not.

If one player is grainy and the other is smooth, that is not part of an EQ since an EQ alters the frequency response and grain is not part of a frequency response.Or if one player has a deeper soundstage, or a blacker background, that is also not a part of an EQ process.I think we all can imagine “neutrality” as a relatively easy concept in the realm of ideas. However, when we move to the real world, it’s very hard to determine what true neutrality is.But the confusion here, as with most people who’ve asked the same question over and over again, is simply the fact that having a character implies (wrongly) that they have applied some sort of an EQ.I hope that makes sense. You are missing the point. A “rebuttal” is not the same word as “proof”.The point is that it is a waste of time to start over on a 10,000 post discussion that has already occurred (with very dedicated people on both sides of the argument). It’s also a waste of time – on any Forum – to start a NEW “creation vs evolution” thread, or a NEW “warming vs denier” thread or a NEW “conservative vs liberal” thread.As someone has already said, Headfonia is entirely based on the method of listening and then changing only one thing, and listening again, and then reporting on what you heard.A “sound quality denier” would have no interest in any such report, so it is pointless to read this site. You may have noticed that I have been reasonably careful in not attacking everything as wrong, merely stating that I am skeptical.As I said earlier, I believe the method of simply listening for differences can co-exist with some degree of skepticism when it comes to the causes of these differences. Accepting everything you hear “as-is” with no real thought as to what you are hearing cannot be a healthy attitude, IMHO.If you are saying that.that.

is the avowed attitude of the site, then perhaps I am wasting my time. First, I cannot speak for the “avowed attitude of the site”, just for my perception of what Mike and L have said. The site does have a statement on this issue at:From my own perspective, I am a professional software engineer and my diploma is in audio engineering.

Having used computers from the HP-65 through the iPad and most everything in between, I do understand bit perfect, and the ability of modern computers to copy terabytes of data without any resulting error in the target files is very gratifying. But of course, errors happen behind the scenes and the inner algorithms reread automatically to correct the errors.

Now in the case of playing CD’s -vs- digital tracks on the computer, I’ll take the computer any day because the CD errors that happen in real time can be heard, whereas the ripped tracks have been error corrected, and real-time errors don’t matter unless they can’t be corrected (rarely a problem). But while data read errors probably don’t happen often enough to worry about with software players, I find it very difficult to believe that modern PC CPU’s and their I/O processors are unaffected in playing a music track perfectly with perfect timing.

I’m not talking about jitter (I don’t think so anyway) – just normal playback. I can watch a cursor having irregular movements on the screen when I’m not touching anything and when the computer is idle. I am certain the things that affect that cursor are affecting music playback, despite the best efforts to buffer out the interruptions. One thing you could do to improve playback on a PC is shut off all network connections, then get into the O/S and shut off all administrative processes etc.

And eliminate any other background processes. Or use DOS on a standalone PC without Windows or any Windows stub running, and no network connected. Thanks for the great article!I am using K550s with the E17 and BitPerfect with my iTunes library (the best setup under 500$, IMHO), but I have one small problem. The E17 doesn’t appear to have 88.2 support (at least according to BitPerfect’s analysis), so it forces me to upsample my redbook files to 96 instead of the more natural power of two upsampling. Is this a problem with BitPerfect, or is it true that the E17 is incapable of 88.2 playback? Is it really that big of deal with my modest equipment to simply use the 96 upsampling setting?

It sounds fine to my ears, still a marked improvement over leaving the files at their native 44.1.Would it be worth my money to upgrade to Sonic Studio’s 49$ option? Would that be the best option if I want full iTunes integration?There are a lot of inconsistencies in this article though. For example, you say BitPerfect is more spacious than Decibel in one section, then say the complete opposite in the following section.Thanks for you help!.

Hi @headfonia:disqusPhones: Senn HD650DAC/Amp: Audinst HUD-MX1Source: MP3 @ 320kpbsGreat article as always. I’ve been playing with Fidelia on a Trial, along with Audivarna Plus (also the trial version).

Overall I think Fidelia sounds “better” with the FHX Crossfader enabled. Without this though there really isn’t much difference to my ears (perhaps not surprising given they both use 64-bit iZotope.What I would say is that I slightly disagree with your view on the usability of the Fidelia iTunes integration. I found this to be really clunky when trying to browse my over-bloated library (20,000+ tracks). It feels like going back to iTunes 10 years ago.

I found the Audivarna integration MUCH better as you basically carry on and use iTunes to pick the music (even going as far as switching the Audivarna display off which gives you a well known library function but with the grunt of Audivarna to handle playback)I do have a question for you re Fidelia’s Crossfeed function. I can definitely hear a change in the music as the degrees and intensity are notched up. I’m finding it difficult to describe this change. Sort of like moving the sound forward (physically). It does seem to come at the sacrifice of volume (actually I’d imagine gain). Is there a recommended setting for these controllable values? I know the easy answer is to play until I find something that sounds “right” but I’m quite naive when it comes to the technicalities of what Crossfeed is doing.

If you have any simple explanations of this function to help me better make the right judgement call (that being what my ears say is right plus what technically is a good starting point) that’d be greatOne final note. I have BitPerfect too. To be honest I really didn’t hear an improvement vs standard iTunes and certainly no where near Fidelia and Audivarna’s improvements. I was disappointed by the available music players for Mac. On Linux, I have used Amarok 1.4 earlier. When I switched to Mac, I have mostly used iTunes and Songbird but I had my problems with both.What I wanted was basically a player which is simple, can handle an infinitely large music directory and has a PartyShuffle/DJ-mode function.

It should also support all most important sound formats (flac, ogg, mp3, m4a, wma, ) and maybe some other things.Because I didn’t found that, I started my own Open Source project:It is simple and is all centered around a main queue (looks a bit like the old Winamp, XMMS or other simple players). The main queue is always in PartyShuffle-mode, though. It shows some of the recently played songs, the current songs and the upcoming songs.

It plays always the songs from the top of the queue and then removes it from there. Once the queue becomes too empty, it intelligently adds new songs to it (based on context and ratings).It is also powerful, e.g. It has its own volume loudness normalization algorithm. And is has Last.fm scrobbling support. And some other basic things.It supports basically all existing sound formats.Because it is Open Source, everyone can contribute and make it better.

The code is simple and mostly Python, so it is easy to work on it. HiA very useful comparative review that I am finding a valuable guide to auditioning alternative players for streaming on my Mac Laptop, however I think you, and (as far as I can tell) all other reviewers, have missed one important point:If one of these players is installed to provide better local playback via USB or Firewire audio interfaces does it inhibit the serving capabilities of the computer doing the streaming?At the moment my music serving via ethernet is confined to using Airplay with iTunes to play remotely 16/48 (i.e.

16/44.1 converted by iTunes to Quicktime) but I intend to move on from this to hi-res in the near future. Testing Decibel I discovered that if I use this for streaming it prevents Airplay output so I cannot listen remotely.Could you please summarise the capabilities in this area of the players you tested?ThanksDavid. HiA very useful comparative review that I am finding a valuable guide to auditioning alternative players for streaming on my Mac Laptop, however I think you, and (as far as I can tell) all other reviewers, have missed one important point:If one of these players is installed to provide better local playback via USB or Firewire audio interfaces does it inhibit the serving capabilities of the computer doing the streaming?At the moment my music serving via ethernet is confined to using Airplay with iTunes to play remotely 16/48 (i.e. 16/44.1 converted by iTunes to Quicktime) but I intend to move on from this to hi-res in the near future.

Testing Decibel I discovered that if I use this for streaming it prevents Airplay output so I cannot listen remotely.Could you please summarise the capabilities in this area of the players you tested?ThanksDavid. Mike, I wasn’t clear on the configuration you used in your sound quality (SQ) testing. Were you using the Mac merely as a music server, feeding a digital signal to an outboard DAC, or were you comparing SQ based on the Mac’s analog output? Assuming it’s the former, how could these programs control the D/A conversion parameters of an outboard DAC? Also, you mentioned these programs have the capability to handle very high sampling rates (96 kHz). But the Mac’s optical output (via the 3.5 mm mini-jack) is limited to 96 kHz, so how were you able to output higher sampling rates — did you use the Firewire or Thunderbolt ports?

Mike, I a bit confused (because I don’t know this stuff).You talk about better than Core Audio up sampling but don’t most USB DAC’s require the output to go via Core Audio?Also all those advanced features confuses me, I thought it was really a matter of the decoder of the audio file (FLAC, MP3, ) decoding the format correctly and simply sending this information bit perfectly/correct to the USB and let the DAC do the magic? So my question here, should any player that will decode the files correctly not be as good as any player when it comes to sound quality as it it is the DAC that does the stuff?Thank you.