Hello,
Iâm new around here, but came across this thread. Iâve been playing with computers as audio streamers for the last 7 - 8 years.
My experience with computers and high end audio equipment is that the issues with computers are two fold. They throw out a lot of RFI into the atmosphere (though they are getting better in this respect) and they have very noisy power supplies (which donât seem to be improving). I have used a number of computers from laptops, to dedicated audio computers (DIY) which use fanless PSUâs, to NUCâs, and one of the biggest problems is earth line based noise. Simply put, PSU designers with computers are lazy, so they basically dump the noise from the systems power supplies straight to earth, as a result the moment the computer starts to do a modicum of work, you get noise on the earth lines and that goes straight into your audio circuits via the amps.
One way around this in the USA is to use experiment with cheater plugs which allow you to lift the earth line, this will give you a very quick (and cheap) reality check as to whether you have a noise problem with your earth lines. If youâre going to have a computer in the listening room, and itâs not a laptop running off batteries, then you also might want to look at some kind of mains filtration for your other audio components, especially if they use linear power supplies rather than switch mode power supplies. I have noticed that the likes of Chord power amps donât seem to suffer from these problems, but then again their amps are ferociously expensive and well beyond the means of most. In the UK I know the use of anything which breaks earth is frowned upon, but one really quick way to check is to buy an extension chord, and simply lift the earth at the plug end and see if that improves matters.
My long term plan is to use an Intel NUC as my client device in the listening room, Iâll probably go with the more powerful i5 based device as Iâll want this to deliver video to my screen in the room also, but you could easily use the cheaper entry level model for audio only, even if you do plan to use DSD in the future, and have the core server part on a stand alone machine in another part of the house with my NAS, keeping the spinning disks and fans outside the listening and viewing rooms. The final thing Iâll mention is networking, donât skimp here. I would always suggest going with a good network switch (HP or similar calibre) and use CAT6 network cables at least, even CAT 7 (which are fully shielded and should only cost a little more than decent CAT 6). If you are thinking of wiring your home, then again donât skimp, use CAT 7, you may not need 10 GB/s speed, but you have expansion potential built in, and the shielding will prevent noise, not only from getting in, but from getting out also, and DO NOT use Wi-Fi for audio. It can, technically, be done, but I have always found a hard wired ethernet connection to be preferable from both a reliability and sound quality perspective. As a someone who works in IT (and has done for the last 30 years) I would really like to give conclusive evidence as to why this is, as even 802.11G Wi-Fi has plenty of bandwidth to handle 96/24 high resolution audio, yet even with 44.1/16 CD quality audio, when itâs streaming from the NAS it just doesnât sound as focussed, and then possible Wi-Fi dropouts, even one or two a day, really spoil the dependability factor for me.
I hope these random musings might help someone here, if you have any questions on anything Iâve said then please donât hesitate to ask and Iâll try to help.
Dave