Streaming Audio

Stream music from your computer to your hi-fi. Here's how, by network engineer and reader Jim Roberts.

Cambridge Audio StreamMagic 6 Network Player. This player – and the many like it – read music from the computer, or from a memory stick. Most can also receive internet radio, making at least 10,000 radio stations available from around the world.

Streaming means sending music through a computer network, from computer to a ‘streaming player’. The player acts as a receiver, as it were, receiving music in digital form from the computer, converting it to analogue through a Digital-to-Analogue convertor (DAC) so it can be output to the hi-fi, usually through phono sockets as an analogue connection. There are no compatibility issues. The signal output is the same as CD, so you connect up to any CD input or any Aux (line level) input on a stereo amplifier.
    You ‘rip’ (copy) your CDs onto a computer or download music from the internet to it, via its web connection. This music collection can then be streamed across the network, from computer to player.
    The advantage of this is that your entire music collection is then just a click away. I can browse my whole CD collection from my armchair and play any track that takes my fancy – even though the physical CDs are now gathering dust out in the shed.  Because of this convenience I listen to more music now than I have done for years.
    The potential downside is that you have introduced computers and networks, with all of the pain that can imply, into your music hobby. However, the network can have a simple arrangement – just a straight wire – or complex with a switch unit in the loft and cables through the house. Or wireless, if you don’t want cables. I’ll explain here what your options are, and how to use such a network.

‘Universal Plug and Play’ (uPnP) is the industry standard for allowing devices, particularly music and video devices, to work with each other across a computer network. DLNA stands for ‘Digital Living Network Alliance’ and is an organisation setup to refine the standard and certify uPnP devices to ensure that they interoperate as they are supposed to.  Most mainstream streaming products use uPnP and some are DLNA certified (some products use their own proprietary standards, not uPnP, but I will not discuss these here).
    The uPnP standard requires three items for a complete music streaming system. They may stand alone as separate devices or be combined into a single device (e.g. modem and router).


This is the computer that holds the music, along with a software program, confusingly also known as a ‘uPnP server’, running on that computer that enables it to serve music.

Pictured at right is a Netgear Readynas 102 with DLNA certified music streaming.


The hardware can be a desktop PC, a laptop or a dedicated ‘NAS’ box.  NAS stands for ‘network attached storage’ and is basically a small, special purpose computer that is used only for holding data – in our case music.
    To make this hardware work as a uPnP server it needs to run uPnP server software.  On a PC, Windows Media Player performs this role, or you can install programs like ‘Asset’, ‘Twonky’ or ‘MinimServer’.  Macs do not come with a uPnP server, so you must install one, such as Twonky or EyeConnect (free).
    Typically NAS boxes come with uPnP software pre-installed so all you need do is turn it on.




The StreamMagic 6 rear panel illustrates a wide variety of connection options. There is a USB socket for long term storage of music on a memory stick, an ethernet socket for connection to the network, and both XLR (balanced) and phono socket (unbalanced) analogue outputs. Also included are optical and electrical digital outputs, for an external DAC or digital amplifier, and digital inputs so the player can be used as a DAC as well.


Think of a streaming player (renderer) as a ‘source’ like a CD player or tuner. It is part of the hi-fi system and retrieves the music on your computer, sending it through the hi-fi.
    Bear in mind that if you have a huge music collection on your computer, seeing it all displayed on a player the other side of your lounge raises difficulties. Budget players have a one line display only (Cambridge Audio), or even no display – a TV must be used (Yamaha). Some models have a colour screen (Musical Fidelity) that you peer at across the lounge, whilst Cyrus players have this on a remote control. Because such screens draw current the remote has to have high capacity, rechargeable cells. Or you can use an App on your iPhone or iPad, where the manufacturer makes this available (most now do).


This allows tracks on the computer to be selected and may be part of the player, as discussed above, or an App on a portable device like an iPhone or iPad.

Pictured at right is a Cyrus remote control with the player's music menus legible a small bright screen.


I have just explained that a uPnP system requires three ‘logical’ devices.  This does not necessarily mean that you need three separate boxes.  Sometimes two logical devices can coexist in the same box. Nonetheless there are good reasons to opt for three separate boxes. I use a NAS box for my server, a dedicated Linn streamer for my renderer and various control points on the family iPad. This means the server can be outside the music room, I can change my renderer should I wish without affecting my music collection, and I can control it all with the convenience that comes from an iPad.

You cannot stream without either a simple ethernet cable link or a full network. There is no getting around the fact that getting a full network up and running requires a degree of commitment and knowledge – see more in the Boxout at the end of this piece.

Pictured at right, an inexpensive network switch like this 5 port Netgear unit allows streaming players, receivers etc to be connected into a network.


An ethernet computer network takes the form of a star, at the centre of which lies a Switch. However, whilst you can wire up your home like this, most people are likely to have a simple straight link between internet modem and their computer, which actually amounts to the same thing, being one arm of a star. Further Switches can be added in, into which you may hook up your hi-fi through the Streamer. Just note that switches form the centre of the star, with links to connected devices on this network.  


A Star Network, with Switch at centre. This is a basic layout. The Modem & Router is usually combined with the Switch in one box.


The process of copying music from your CDs to your server is called ‘ripping’. Computers display files in a directory tree. Whilst you need directories to organise files when saving them on your server, directories are not the main way uPnP views your music collection.  


Pictured at right, Metadata tags contain disc information used for disc cataloging.


Instead uPnP uses ‘tags’.  Every song on your CD is saved as a separate file and each file contains both the music and a set of tags. Tags contain ‘metadata’ – information about the track. There is an ‘Artist’ tag, a ‘Title’ tag, a ‘Year’ tag, etc., and it is these tags that the server uses to organise your music. It would be a bad idea to do so, but you could if you wished put all of your music tracks in the same folder, and as long as they are tagged correctly the server will still present them to the control point organised correctly by artist, album, genre, etc. It does this by reading the metadata for every track when it starts up, putting it into an internal database. It really is a good idea to get your tagging right from day one.
    This leads to the decision about what music file format to use. The open standard is FLAC – Free Lossless Audio Codec – and this is supported by all of the mainstream hi-fi companies (two exceptions are Apple and Microsoft, but neither are hi-fi companies). As the name suggests, FLAC is lossless – and free!  FLAC files are compressed but in such a way that none of the quality is lost.  A FLAC file can re-converted back to the exact data that was on the original CD.  It works much like the ‘zip’ programs that shrink computer files then expand them back again perfectly.  FLAC also supports rich metadata and this, plus lower data rate, sets it apart from basic WAV.  WAV offers the same quality as FLAC, but with larger file sizes and data rates, meaning FLAC is preferable. As a result FLAC is now very popular and a good choice.
    Some companies sell server devices that do the ripping for you, but I prefer to rip on a computer then ‘upload’ the ripped files to my server as I feel this gives greater control.  I use ‘dbPoweramp’ software for this – daft name but superb software. It reads and re-reads discs until it gets a bit-perfect rip. It confirms this perfection by comparing with an on-line database of previous rips done by other customers. It also automatically downloads tag information and cover art graphics. I buy second hand CDs off Amazon for pennies. It doesn’t matter if they are a bit beat up.  dbPoweramp gets a perfect rip 99% of the time. Wonderful!
    I also use the excellent ‘mp3tag’ software to retrospectively tweak any tags I am not happy with. Despite its name it works with FLAC.

It took me a year to rip my CDs and I don’t want to lose this work the day a disc fails – as inevitably it will. So I need backups. I backup in two ways. I have a large capacity hard drive in the computer I use for ripping and periodically copy everything from my NAS to this drive. I also bought a large capacity USB 3 external hard drive which I keep at a friend’s house.  Once a month I bring it to my house and copy everything to it, then take it ‘off site’ again. In this way, if I am burgled, they may get my kit but they won’t rob me of my music!
    Whilst a RAID array does ensure against failure of a single disc in the array, through redundancy, it won’t save you if you accidentally delete a bunch of files, or from the burglar! With 64GB memory sticks and SSDs (solid-state hard drives) now becoming affordable these are potential alternatives, or you can back up to cloud (off-site) storage like iCloud of course – another whole topic.

Switching to a streaming solution is the best hi-fi move I have made in forty years in the hobby. The sheer convenience of browsing my entire music collection from the comfort of my favourite chair is unbeatable. Streamers also work up to top 24/192 resolution, potentially offering better quality than CD. They are the way music replay is going, so I encourage you to try it and hope my explanations and suggestions help.



All home networks use a technology called ‘Ethernet’ and a network language (or ‘protocol’) called TCP/IP.  I will describe a wired network first, then talk about wireless.  
    Ethernet uses a ‘star’ topology, which is just a fancy way of saying that there is a gadget called a ‘switch’ in the middle and every device is connected to this switch by its own dedicated cable. 
    Most wired networks use ‘structured cabling’ which means that the switch is hidden away in a cupboard (or in my case the loft) and next to the switch is a ‘patch panel’.  This is a passive device simply used for terminating cabling. Permanent cables run from the patch panel through the building to be terminated in RJ-45 network wall sockets (similar to phone sockets) at various points around the home. Loose ‘patch cables’ join the switch to the patch panel and similar ‘fly leads’ connect the kit to the wall-sockets.  Network cabling comes in different qualities: cat 5; cat 5e; cat 6; etc. 5e is fine for a home network and is cheap when bought in bulk.
    Clearly the problem with a wired network is running the cables. You will need to drill holes in walls and lift floorboards.  Nonetheless this is the best way to go if you can face it. Wired networks are more dependable than wireless ones. High-end streaming products such as those sold by Linn are wired only. Most NAS boxes also require a wired connection. I ran the ‘easy’ cables myself and got a burglar alarm installer to install the more difficult runs. He did a great job and the price was not bad compared to what we sometimes pay for our kit.
    The easier alternative is to use wireless networking. Here radio waves replace the cables. It’s still Ethernet and still a star – but the switch has a ‘wireless access point’ built in which transmits and receives the data from each of the connected devices. The ‘WAP’ can go somewhere convenient, like on a shelf. Wi-fi standard 802.11g & n both have plenty enough speed for high resolution audio, since top rate 24/192 requires a throughput of 9Mbps and this is at the lower end of what wireless can achieve. However, data rate falls with distance and wi-fi is never as dependable as wired.
    Whether wired or wireless, every device on the network needs a unique identification number called an ‘IP address’. These can be configured manually or automatically using a piece of software called a ‘DHCP server’ which is built into most routers or NAS boxes. Google this, then recruit the help of a mate who knows what he is doing – probably a teenager!
    Wired or wireless, get your network operating reliably before you start setting up your streamer.


For the last 15 years I have made my living in computer networking. I have three ‘Microsoft Certified System Engineer’ qualifications and a ‘Cisco Certified Network Analyst’ qualification. I’m also a hi-fi nut.  I run a single-ended triode (Almarro 318B) into Tannoy Westminster TWs – in a typical suburban lounge in Birmingham.  But my only source is a streamer!  A Linn Akurate.



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