January 2011 issue - Page 6 - Mac Mini

Article Index
January 2011 issue
Page 2
Page 3
Page 4
Page 5
Page 6 - Mac Mini
All Pages


Having read your replies to Leigh Penny’s questions in the past two issues, I really feel you should take a long, hard look at yourselves when it comes to computer audio. You either embrace it or you don’t. People write to you for advice because of your knowledge and expertise. And when it comes to most matters hi-fi, you display this admirably. So if you choose to give advice relating to computer audio, you owe it to your readers and to yourselves to research your answers more thoroughly or simply admit that you don’t know.

Bob Hurst had it about right (I loved his “charmingly misinformed”) but having printed his letter you then carry on regardless with your (Noel’s) second reply.

Some comments here.  You seem surprised that your Mac would only output 44.1 or 48kHz via USB to a “96kHz capable” Cambridge Audio DacMagic. Well actually it isn’t. This device (as are many others) is limited to 48k via USB. In your defence Cambridge Audio don’t exactly shout about this.

I also have a suspicion about your jitter analysis. You write that jitter from your Mini was fine at 44.1 yet not so good at 48kHz. But you don’t say what files you were playing. My suspicion is that you were playing 44.1kHz files. Played back at their native resolution – no problem. But to play back at 48k, OSX would resample them real time. This is a BAD THING. And would explain your results (had you been able to measure 96k more badness would have resulted). Which brings us neatly back to Leigh Penny’s (LP from now on) question regarding playback of files having different sample rates.

Here you write that output rate is set in ‘Audio MIDI Setup’ and that once set “you do not have to change it. All files are converted to this rate.” Whilst technically true this is appalling advice (see BAD THING above). Here in particular you should hang your head in shame.

Fortunately there is a solution. LP doesn’t offer any information as to what software he’s using for playback but with a Mac, iTunes is a fair bet. Until quite recently this would have been a delicious irony; the slickest of user interfaces coupled with the (allegedly – don’t sue me!) worst sound quality. The solution is software available from either Pure Music or Amarra (Google them). They both do similar things. They interface with iTunes, taking over playback duties with a reported increase in sound quality whilst retaining iTunes’ functionality etc. (I would like to make it clear, I use neither Macs nor iTunes so cannot confirm the sound quality issues but the weight of evidence from users is pretty overwhelming). My suggestion would be for Pure Music. It’s significantly cheaper ($129 vs $695 for Amarra) and also available as a trial download, free for 15 days. Can’t really lose.

There is also a second very worthwhile benefit; automatic sample rate switching. So your CD rips will play back at 44.1kHz and any hirez downloads or rips will be played at their native resolution up to 192kHz (hardware permitting) without having to change any settings.

Back to LP. All you need now is a DAC. And I’m afraid I’m going to have to disagree with Noel again.  How best to get your digits out of your Mac seems a fair place to start. Cat, skin, ways, many! Without getting too involved, and especially if you will only be connecting your Mac to the DAC, then asynchronous USB would be my choice (not to be confused with asynchronous upsampling). I hear what Noel says regarding isolation from the noisy computer environment. Valid but now addressed by several non-optical options and only one of many issues.

An aside: Noel alludes to Naim Audio’s website ref their own DAC. They don’t like USB connection to a computer. I have no problem with Naim – they’re a wonderful company. But they would say that wouldn’t they? Watch them change their minds once they have their own implementation. Contrast with dCS for example, who use asynchronous USB; and they know a thing or two about digital.

For £330 or so the HRT StreamerII+ would seem to represent a fine mix of sound quality with value for money (and a 5 globe review from Patrick Cleasby in the August issue). It also claims a “completely isolated path between the computer and audio system” although the details here are somewhat vague.

I would suggest that for now there would be little to gain spending more until you reach the Ayre QB-9 at about £2k. USB only (asynch of course) the Ayre isolates from the computer using opto-couplers and also provides both balanced and single-ended out. I’m surprised Noel doesn’t already have one!

Both these machines are currently limited to 96kHz but 192k will follow soon. The Ayre will almost certainly be upgradeable while HRT are supposedly bringing out a 192k version in the near future if this is important to you.

If other inputs are required then the new Arcam rDac might be worth a look (£299ish). Again it has asynchronous USB input (here licensed from dCS) as well as conventional optical and electrical S/PDIF inputs for plugging in set top boxes and the like.

Finally (to LP again) if you were to buy the HRT, with the money saved from your £1,300 budget you could further improve sound quality by replacing the Mini’s hard drive with an SSD (solid state drive). It doesn’t need massive capacity as your music files would best be stored on an external drive, preferably network attached (ethernet, not USB) and away from the system or even in another room. Isolate the computer from the mains with a power conditioner/filter, get a decent quality USB lead and buy an iPad.

Okay the iPad won’t make any difference to the sound but what a controller it would make.

Andy Holland,



The Mac Mini computer is not purpose designed for audio.


Thanks for all the info Andy – and your opinions. It is very interesting, if very much in the assertive computer audio / internet tradition based on internet folklore.

That USB outputs 48kHz maximum into a Cambridge DacMagic warns that what USB outputs is determined by what is connected, unlike S/PDIF. I should have made this clearer, but it gets very long winded if every option is catalogued and explained.


The Rohde & Schwarz generator output a 48kHz test signal to the Mac Mini direct and this was clearly stated in our reply; it was not upsampled from a 44.1kHz file as you have surmised. Your subsequent arguments rest on the assumption that the Mac will degrade the signal if asked to sample rate convert. Sample rate conversion is a mathematical process based on an algorithm. It does not necessarily cause degradation, unless the algorithm is a poor or compromised one. Internet folk lore might think otherwise but in real life these situations are never so black-and-white; it comes down to quality of implementation.


The most likely source of jitter above 44.1 sample rate lies in a clock multiplier; these can be very crude. The Mac Mini likely runs a 44.1 clock.

Noise isolation by optical S/PDIF is a little confusingly explained in the terminology generally used. Noise in this case means earth currents. Non optical options will only provide earth current isolation through the use of transformers. Internal computer noise is entirely different in nature, comprising radio frequency hash, usually eliminated by low pass filtering.


What Naim think cannot be categorically rubbished because you or others ‘think’ differently. Have you seen the variety and sophistication of their internet products? At Hi-Fi World we listen carefully to what manufacturers say and try to make sense of their views in the broader scheme of things. They have the knowledge, experience, design skills and test equipment to understand today's digital technologies.


Sound quality will be improved by an SSD? Er, not necessarily. This again is another piece of internet folk lore. Explained to me the other day by a digital engineer, modern hard drives – the ones that spin round and round – might seem crude but their signal processing is sufficiently sophisticated to produce a clean and stable signal.

Is an SSD automatically better? That depends upon how well the data is clocked out. Also, SSDs do not have an infinite life span, their cells degrade. Catastrophic failure, hard drive style, is unlikely. But SSDs are not automatically an answer to a maiden’s prayer. What the ‘net thinks and what is an engineering reality are two different things.

Trying to maintain a stable, rf hash and jitter free signal down a long signal chain, especially within products like the MacMini not purposed for audio, is never going to be easy. But supposition and folk lore provide no enlightenment. NK


Hjalmar Nilsson replies on behalf of Naim:


Andy, some interesting ideas and theories, but they are theories, not fact.  We have been researching digital audio for many years and hard disk replay for enough time to realise that getting good sound from a computer or hard disk playback is relatively easy, but getting great sound quality is really difficult.


We have some problems with USB, yes, but maybe not the ones you think. Isolation is not easy, especially at extremely high frequencies.


Secondly USB playback is in its infancy and we have already seen significant steps forward in the past couple of years.  We see it as far more sensible for a Naim customer who wishes to use USB to buy a USB to S/PDIF ‘adapter’ to feed a Naim DAC and to be able to easily change this as time moves on. This is better than having a DAC that could be out of date quickly. We can update the DAC’s firmware and code but that may not be enough if we included USB.


We have also got some very good results with Firewire to S/PDIF and here again things are changing quickly.

Optical isn’t changing nearly so fast and it is a simple, easy connection method that provides galvanic isolation, so we included it.  We do have some concerns about the ultimate sound quality available with an optical connection. The speed limitations of the interfaces is limited and their current draw isn’t ideal.


We class the Mac and a DAC (using optical) as it became known, using any DAC, including our own, as a good solution.

We still feel that the sound delivered by a Naim HDX connected with a coax BNC terminated cable to a Naim DAC takes a lot of beating. And we have the NDX network player due out very soon. The research and the development never stops.


In summary, don’t believe everything you read and listen carefully to everything.  Digital is not different from analogue in that every change is audible.  If it isn’t, question why. HJ



Add your comment

Your name:
  The word for verification. Lowercase letters only with no spaces.
Word verification:


Hi-Fi World, Powered by Joomla!; Hosted by Joomla Wired.