Paul's Posts — 14 June 2012
By Paul McGowan
A lot of folks out there are taking an older unused computer or even a new MAC Mini and dedicating it to serving their music. This is a good idea relative to using a shared computer to do the same – but dedicating a computer to this task isn’t the same as a purpose build server.
Let’s take the MAC Mini for example. The MAC Mini from Apple is perhaps one of the best bargains in computers today: low cost, quiet (fanless) fast and has a built in hard drive. Once you manage to install the appropriate software and get the setup right, you can remove the video monitor, mouse and other peripherals and just let it run. Many folks do this with great succes.
But an all purpose computer, like the Mini, is more like a receiver than separates. An all purpose computer is asked to do a lot of things unrelated to serving music and, just like in the case of a stereo receiver, it really never does one thing better than anything else. We turn to separate components for a reason and that same logic applies to computers serving music as well.
Our engineering team is working hard on building a true dedicated server – as I am sure a few other forward thinking high-end audio companies are as well. Our work so far has shown clear and irrefutable results that dedicating hardware and software to one single task provides far better sonic performance in streaming audio than a general purpose device does – and this applies to just taking an off-the-shelf motherboard and dedicating it to the task as well.
So if a manufacturer has chosen to purchase an off-the-shelf single board computer or motherboard and then polish it up with software and peripherals to build a dedicated server – it’s still a general purpose computer with all the inherent noise and issues associated with them.
We’re approaching an age where we’re going to get back to separates and for good reason.
Forward to a friend and help us engage more readers
Related Articles
Share
About Author
(8) Readers Comments
Leave a Reply
You must be logged in to post a comment.
robertawillisjr
By dedicated server for music do you mean a NAS that is DNLA compliant?
Paul McGowan
Yes, a NAS is a dedicated server and on the face of it better than a dedicated general purpose computer as I mentioned – however, a NAS is yet another story. Most NAS we are beginning to understand, have such minimal computing power as to be a hindrance to us – they just have lame low cost processors that struggle to keep up with the demands. So while they do indeed quality as a dedicated device – a separate – they are not a very good separate.
Bassman23
It’s ironic that thiis post, talking about getting back to separates, comes closely after your post concerning the advantages of integrated amplifiers. Of course, you’re addressing two separate issues. But it suggests developing a post looking at what would currently be required of the best path of a signal from source to loudspeaker. I’d enjoy your thoughts on that subject. Come to think of it, that might be a book.
petewilson
It would be interesting to know whether there are any measurable problems with a Mac Mini as a music server – I used one in exactly the way you said for a long time (hooked up by HDMI) because that way I could not only hear the music, I could look at the gorgeous photos on the big screen TV.
TV’s now moved into another room, and the Mini has been replaced by an AppleTV (because it was needed elsewhere in the house as a Mac Mini
Mini’s aren’t fanless, incidentally. They’re almost silent, and they’re small and unobtrusive.
But what do they do wrong (I’m interested in measured things as well as the ‘subjectivist’ commentary)?
- P
Paul McGowan
They don’t sound as good as a dedicated high speed processor in a direct comparison. Much to our surprise as we’ve been looking at why there seems to be some variability between servers and the way they sound, we’ve noticed that when we use a desktop computer, the sound is different than when we use our prototype dedicated version – and that includes our little MAC Mini. There’s lots to be uncovered at this point and I certainly don’t have the definitive answers, but we’re working on it.
alan
The subject of Integrateds versus Separates particularly in terms of transports, music servers, associated storage devices, DACs, all the cables necessary to connect these devices, plus software to control everything, gives me the shivers. I want to see preferably one box, but at most, two boxes. One contains all the sources such as CD/DVD transport, Music Server, Streaming Downloads, HD Downloads, etc., and the other box contains the DAC as well as all additional controlling software, capable of course, of software upgrades. Since every cable has the potential to degrade the eventual sound, the number of cables controlled by the user should be minimized. This means there should be one cable similar to the I2s connecting the two boxes, then either analog or digital cable from the DAC to amplifying equipment. To me the ability to control the sources amounts to what was found in an old preamp: a selector switch. That way, the design engineers, who know how to connect and control and tweak all the pieces, do so, and the clueless music lover chooses a source then sits down and enjoys the music. I currently use the PWT/PWD as a two-box CD player and that’s plenty complex for me.
Signed,
Clueless
Tom Devey
Oh oh. Audio companies building computer hardware. Look for Moore’s Law to invert.
I can certainly see the benefit of an audio company doing the design and development of the D-A Conversion “to fully transcend the limitations of traditional digital to analog converters (DACS) by providing uncompromised performance for any media delivery or storage system possible. This means that it no longer matters how the digital audio data is stored or delivered, the PWD produces the same high performance audio sent over a network, taken from a hard drive, a CD or even downloaded via the Internet..” (from the PS Audio website).
I can’t see in this scheme how the server can make the difference when it is simply a source for the collection of bits which make up the data for the musical program which is assembled and re-created by the DAC. You would not be relying on streaming the data directly from the servers, as the DAC should provide buffering. Is the difference noticeable when the servers are completely electrically isolated from the DAC, including any interconnecting cabling or power?
djr
My personal little online crusade: Apple’s Macintosh computers should be referred to as Mac, not MAC. MAC would be correct if it was an acronym, as in ‘MAC address,’ which it is not.