Home network - basic hygiene to make your music server sound its best

The Miles Davis’ quote along these lines is the most famous one. Lots of different musicians/composers have expressed similar thoughts, maybe not Beethoven. It’s good to learn about the science behind what I am hearing. So thanks for sharing that. I am very interested in that side of things/improving my understanding.

Thanks much appreciated. will take you up on that offer once I find one :slight_smile:

I think you have discovered some of how much cables matter. i’d say they matter more than switches because:

  • cables seem more vulnerable to RFI etc
  • a quite decent result can be achieved with non-audiophile switches and very good audiophile cables.
  • logically you need at least 2 cables connected to a switch.
  • a really good switch will be somewhat wasted with inadequate cabling, like a really good DAC will be wasted with supermarket interconnects.

I’d also say it really depends what function the cables are playing and where abouts in our network they are. As a general observation, the cables closest to the endpoint seem to have the greatest impact, then incrementally each cable further from the endpoint has less impact. Into this equation fibre should be considered, for example over a long distance fibre may perform better than a similar cost copper cable. Over long lengths fibre costs a few dollars that would costs significantly more to achieve similar results using audiophile ethernet cable. It is even possible fibre with quality SFPs competes with some quite high cost audiophile cables, and over 1m audiophile ethernet cable can perform even better.

Each part of the network contributes to the end result, some parts contribute more, and I suggest it is best to consider every part rather than spend a ton of funds on fancy things near the endpoint and ignore the rest. It is likely that careful attention to upstream parts of the network will either reduce the burden on downstream parts to achieve the end result, or will leverage downstream parts quite a lot to achieve an even better end result.

When putting together a network for audio, I think it really is about choosing best value for money for each part and its function. For example my network has:

  • a $5 Cat 6a UTP furthest upstream
  • a $200 router
  • $100 of fibre and SFPs
  • a $30 WAP as the last link to endpoint
  • 2 x $1000 audiophile cables where they count most, one of them connects to the $30 WAP!
  • Ether Regens and high quality LPSs.

It is not to say the $5 Cat 6a could be improved upon, but so far the magnitude of improvement does not warrant the expense - which is a personal value judgement. I doubt shift $300 expense to this link rather than the $1000 cable would be worthwhile.

And finally getting to the relative cost benefit of cables, I suspect most hifi cabling (incl ethernet) pricing is based on the amount of impact it has in a sound system relative to its competition in the market place. So a $500 ethernet cable could typically have a similar amount of impact to a $500 interconnect, or $500 ‘audiophile’ switch even though each of these may have different type of impacts (eg. colouration, flavour, tonal balance etc). Of course there are exceptions.

I hope this is helpful.

1 Like

My experience has been that improving ethernet and other things that reduce system noise and improve accuracy (eg. grounding) has significantly reduced the amount of excess bass energy being output. I don’t have any bass trap type treatment, and the only issue with the ARC MATT test I did was resonating panes in windows. But my room has openings at the back where some energy can escape or dissipate.

No doubt about it, improving the network will bring many important improvements. The problem though is that the room can stomp all over these improvements, and hinder what they can bring to the table.

Ideally the only sounds arriving at our ears should be the sound coming directly from our speakers. Unfortunately what we also hear is the effect of what ASC refers to as “head end ringing”. Sound from the speakers bounces off the walls and arrives at our ears slightly delayed. This throws off our ability to “ easy to discern the difference between say a violin & viola in a complex, say orchestral piece of music”. The more we are able to dry up this ringing, the easier it will be to discern a violin from a viola.

The front corners offer the best position to dry up a lot of this ringing. Even an affordable product like the GIK bass trap can cut down on some of this when mounted in the corners. But the isothermal tube traps take it to another level entirely. They may seem expensive but I think they are one of the best values in audio.

For many years I thought I would need to upgrade to better speakers to get to the next level. No doubt I’d get gains from doing that, but after adding the tube traps I realized that I had yet to hear how good my current speakers were - especially with classical. Most profound was hearing individual instruments emerge and be easily identifiable. Transient response and particularly the stopping and starting of notes improved dramatically as well.

I have come to consider room treatment to be foundational. Along with power distribution and equipment supports, if we don’t properly address these things, we aren’t going to hear what our components are really capable of delivering. That could lead us to always looking for the next upgrade because we aren’t satisfied with what we have.

Everything matters though so I’m not saying that we should just ignore our network or our servers. I’m just saying that if we prioritize the foundation ahead of everything else we’d position ourselves to get even greater gains from everything else we have or will try in the future.

1 Like

To be honest I have been shocked by the extent of impacts, way beyond what I had thought possible at the outset of my experimentation. Hence starting this thread and being relatively detailed about what I am listening for. The results have been at one end of the continuum…unlistenable… to at the other end…totally switching off my analytical brain and simply enjoying the music…and everything in between.

Damn…I was hoping I wouldn’t need to experiment too much with cables as well.

I posted this in another topic that could help people that read this topic.

2 Likes

I know, it can be tiring. I think you are a good track …

  1. establish a good basic network with good hardware, fibre and well built cables with appropriate shields (either UTP or STP with shield not connected to the ground or head-shell at the downstream end.

  2. evaluate cables that connect to the endpoint - this will be quite informative for the rest

  3. after selecting the cable at 2, then evaluate cables for the next link upstream, and so on.

After choosing cables, which hopefully minimises the noise/interference getting onto the ethernet signal and being passed on (in some cases cable also absorb noise), then review the hardware/switches, starting with the one immediately upstream of the endpoint, and so on upstream.

I’d say it wouldn’t take too long to understand the ‘qualities’ of different cables and degree of diminishing returns with each step upstream. Each bit of kit will have a different tuning so it may be possible find a mix of the flavours of cables and hardware that are complimentary to achieve a desired tuning - which is probably actually tuning the way various noise is minimised.

2 Likes

Just clarifying the “unlistenable” part as it may come across an exaggeration.

I deliberately incorporated some very obviously suboptimal combinations (eg el cheapo FMC+stock SMPS on US/NZ adapted plug feeding directly into my K50) that delivered this result.

Equally & in contrast when I fed the K50 with the best possible network component. I could get my hands on, I forgot that I was doing experiments and simply got lost in the music.

Admittedly way beyond basic hygiene (sorry)…but I had been told a lot of this stuff didn’t matter and by extension of that logic, I deduced the K50 would have to be indifferent to / sound exactly the same regardless of what it was being fed. And that is what/why I tested this way.

1 Like

Super helpful steer on cables/approachhttps://antipodes.support/u/dbastin2605, thank you!

1 Like

Awesome. thanks. This certainly will help me feel better about buying a switch with a good clock…ie knowing that its not just a marketing gimmick / superfluous, which is what I had thought previously on the assumption that either my K50 or Roon software was buffering/caching data streams &/or the K50 was itself reclocking before handing off synchronous data feed to the DAC.

The more I learn, the more I realise I don’t know!

1 Like

This all does happen in the K50, but plugging ethernet into pretty much most things will probably inject unwanted interference/noise. Some things are very good at filtering alot of that out, such as by format conversion/reclocking (eg. usb to SPDIF), sample rate conversion, PCM to DSD, etc. Grimm Audio MU1 claims to do this very well thanks to its code/algorithms.

This is one of the ways to ‘remove veils’ … it’s like differrnt sources of noise, interference, distortion, etc are each veils, and enough are removed it reveals other veils more obviously.

A good clock does not make a good switch, other design factors are also quite influential

1 Like

Yes, as I understand it ….network noise isolation/filtering, low noise internal chip sets…clean/quiet power….vibration reduction….just need to make sure I don’t end up with something that’s over engineered :smiley:

The comments about room interaction etc are also well made. This weekend I attended a retailer event where 2 systems were blind tested. Both were expensive, £50k and £100k, and the £50k system had 40% of the budget on network/system architecture.

It was fascinating on a number of levels but in my opinion the superior performance of the digital front end in the £100k system was undermined/negated by the amplifier-speaker matching and/or the speaker-room interaction which meant the handling of bass by this system left much to be desired. Perhaps as a result the majority of those present preferred the £50k system despite the evident limitations of the digital source in that system.

I haven’t spent anything like £50k on my system and I would say it is superior to either system I heard on Saturday because the upgraded CX/EX is a better digital front end than in the £50k system and because my amp & speakers work pretty well together and in my room. My spend on infrastructure is certainly less than half that in the £50k system, with the most costly single element being the Naim Fraimlite rack. Maybe I could get some further improvements but I don’t feel there is anything like another £10k to spend on infra-structure to hear my system at its maximum potential.

What this experience has taught me is (being slightly arrogant for a moment) I have quite good critical listening skills and that speaker selection (for personal preference and room interaction) and amplifier matching to the speakers are foundational. The streamer and DAC in the £100k system were 80% of the cost of the entire £50k system and clearly superior in terms of resolution and musicality if one listened carefully but most people preferred bass that was “right” ie not over-emphasised and well-controlled.

2 Likes

thanks for sharing that @glos_olly. I can certainly vouch for that…along the way I’ve had …speakers that were too big for my listening space…amplifiers that were not powerful enough to drive the speakers…a set up with all the $s spent on components & cabling with nothing on room treatment…a listening space that was too small with listening chain right up against the back wall…and…and…and…

Armed with this knowledge/experience, if I could start over…my first consideration would be #1 the listening space (dimensions/set up positioning of gear/acoustic treatment etc…), then speakers, then amp selection, then power infrastructure, then (I think) racks/vibration management…I think these allow one to hear what one’s gear is truly capable of/truly sounds like…otherwise a bit like putting 91 unleaded into a Porsche GT3 :racing_car:

True story…I installed a new network switch over the weekend. I’ve hidden away some of my other network components (eg router, WAP etc) behind one of the bass traps behind my left speaker. I was doing some cabling experiments and generally fiddling around at the router end of things and on one occasion rushed hurriedly back to my listening positioning to get back to listening to music…the sound had changed (for the worse) way out of proportion to the changes I had made… I was perplexed…when I got back up to revisit the router end of things I realised that in my haste I hadn’t put the bass traps back properly into place…I did so, sat back down…ahhhh back to blissful listening. The difference was anything but subtle!

1 Like