I’m pretty happy with my current system and I don’t plan to touch it for a very long time. It’s just a NAD M33 to Mission 770s but it really works for me.
I play exclusively local files that are stored in a Roon Rock → D-Link cheap switch → NAD. Cables are cheap CAT8 stuff but they actually sound better than the generic CAT6s I had before.
I must have bought 10 switches already because I’m fascinated how different they all sound from each other. I’ve also tried the linear power supply ( LHY LPS25VA) powering the switches over the switching one.
Unfortunately I can tell every change made, be it switch, cables or power supply. Enough for me to consider a final investment into this network thing. I could even tell these diferences with my previous lower tier system.
Bit confused though. Opinions seem to be on two camps… audiophile switch + good patch cable → Streamer; or regular switch → network filter (Network Acoustic etc).
To the people that also notice the differences, what is your experience and what did you settled for?
I’d recommend a listen to the FemiMax NetPro 8000 Network Switch using the the FemiMax PowerStation LPS if you can. I think there were only a few hundred of them ever made, so not easy to find and not much written about them online, but worth keeping any eye out for one.
It really seems to keep the TCP packets in much better shape than the cheap and cheerful network switches you get from the likes of Cisco and Juniper, which obviously makes for a huge difference in your listening pleasure.
Low-cost Ethernet switches have noisy switching power supplies (SMPS) that can inject electrical noise into your power wiring. If grounding is not perfect in your system, that noise could show up in the analog side of your system. The only audible change I’ve heard from changing network gear was replacing a basic Netgear switch and its SMPS with a refurb pro switch (Cisco) that has a better, built-in, power supply. I might have got the same effect replacing the Netgear SMPS with a medical-grade Meanwell SMPS. Good quality CAT6A cables should be all you need. CAT8 is shielded, which has unpredictable effect when connecting to audio components, because the shielding itself can be a carrier of noise into the audio component’s ground plane.
The other issue to keep in mind is that it is very difficult to do objective comparisons between different network hardware configurations for possible SQ impacts, we are all subject to placebo/latest thing effects. In the specific case I noted above, I took care to ask a very good listener who was not aware of the gear differences to compare the two alternatives in different orders, and they reliably identifiied the superior one.
A lot of solutions are not easily tested before buying. Sort of like speakers that might sound great at the store because of their pairing gear and good demo room but not so much at your place.
I was just trying to get an idea of what people might actually use because prices vary immensely and I don’t have a good idea of what different tiers bring about. From what I know from people that I trust with similar sensitivities, something like the LHY AS6 should be a worth while entry point. I would be willing to step up to something at the price-point of the Matrix Audio ss1/ss1 Pro as I have experienced pretty significant variations with cheap stuff to make me believe these are important upgrades. I have a feeling that after this 2k point only a truly high-end system would truly benefit. And add a good patch cable too, as someone has let me play with an Audioquest Vodka a few years back and that was almost a component level upgrade.I had another system at the time that was a few steps back from my current one, though.
My guess – and it’s only a guess from my experiences with varying electrical environments in the US – is that differences in perceived SQ with changes in local network gear depend on the characteristics of the electrical environment in your home, how components are powered and connected, and the quality of component power supplies, digital input isolators, grounding, and shielding.
From experience with different audio components, there’s little correlation between a component’s alleged/perceived quality (and price) and its ability to block out unavoidable electrical interference from power and digital inputs. So a lot of the market for pricey network “upgrades” is really a market for external fixes for audio components that are not able to keep out all the electrical chatter in a modern home: a multitude of cheap SMPSs from all the digital gadgets, LED lighting, lighting controls, WiFi access points, gaming computers, … as well as all the gear your neighbors own and also floods the wiring in the walls with high-frequency electrical noise.
Purely digital gear can be (relatively) easily immune to all that chatter because that’s what it needs to be to be able to move and process bits around reliably, but anything that bridges the digital and analog domains with expectations of very low analog noise requires more specialized design. Even worse, the electrical environment in homes keeps getting more challenging with the proliferation of digital gadgets, and I suspect that specialized audio component builders may not have the resources (budget and experts) to keep up.
To go back to my initial example with the Ethernet switch replacement: it should not be needed, if the audio component – high-end, from a highly regarded brand – had been as immune to its electrical environment as I should expect for its price. But then the refurb switch was very cheap compared with the component, so why not?
You owe it to yourself to spend a bit to get some clarity on what represents a real improvement. You can pick up a Cosmos E1AD stack or RME ADC and run test signals through your network from Roon, tap the pre-outs of the M33 and analyze them with REW to really see the noise and distortion (if any) introduced by your different network and cable choices.
There are some great guides to REW online, REW can even generate test tones and tone sets. You can also do nulling tests using other software. A good measurement baseline is illuminating!
3 Likes
Torben_Rick
(Torben - A Dane living in Hamburg - Roon Lifer)
8
