Audiophile Switch Experiment Results

Roon Core on a Synology RS3617XS via 20Gbit 802.3ad LAG ovet OM3 fibre to a Dell X1052P switch feeding a Bluesound Node 2i. Node 2i via coax to a Topping DX7 Pro DAC/Headphone Amp. DX7 Pro via RCA to a DK Design Vs.1 Signature MkIII integrated with 6N23P preamp tubes driving Zaph Audio ZRT 2.0 large bookshelf speakers in a sealed enclosure on heavily damped stands. The drivers are made by Scanspeak - Revelator midwoofer and Illuminator tweeters that appear in commercial loudspeakers costing > £10,000 a pair. Crossover components are extremely tight tolerance high quality polypropylene capacitors, air-cored inductors and non-inductive Mills resistors. Cabinets are finest quality braced and damped baltic birch plywood and hand made by me. Room treatments including absorption, diffusion, diffsorption at key early reflection points…


Thanks for entertaining my question

1 Like

No problem.

My point was that you don’t have to spend the equivalent of the mortgage on a decent house to have a seriously good hifi system that lets you enjoy music enormously. My hifi journey has been a long one so far and I’ve learned a lot along the way. Solid source components and decent amplification lay the foundations. Well designed loudspeakers and how they’re integrated into the listening room complete the build. Don’t get me wrong, loudspeakers are an area I’m constantly working on - my next build is a wide-baffle 3 way design. The driver cost is less than my current 2 way, but expensive drivers do not necessarily make a good loudspeaker. Great loudspeakers can be made with lesser drivers. It’s all about the design. I’ve seen very expensive commercial designs ~ £20,000 that measure terribly and sound bad…

When it comes to cables, I build my own for the most part. Decent commercial grade cable such as Van Damme starquad for interconnects, speaker cable is QED, Neutrik RCA/XLR connectors, Neutrik Speakon at the speaker end. Network cabling, I run and terminate myself - Cat 6A. Patch leads are decent, standard Cat 6 stuff.

1 Like

So cool you can make your own speakers! If you allowed yourself an appropriate amount of money as the builder, plus the parts, how much would these speakers cost?

Network switches are galvanically isolated by design. There’s no audible difference at the analogue end regardless of the network hardware in use. The network provides data (1s and 0s) in packets via TCP/IP. Those packets are buffered (ever noticed the delay between hitting play on a track in Roon and actually hearing it start?) at the streamer/DAC end. You can unplug the network cable and the music will continue awhile…

Amir at ASR proved that the EtherRegen and Silent Angel do nothing as far as audio is concerned. High end ethernet cables are a waste of money. Over sensible runs, Cat 5e will easily support Gbit ethermet. Audio doesn’t even need 100Base-T. Cat 6 will support Gbit over long runs and 10GbE over shorter runs. Digital audio data isn’t special - it’s just data. Your internet connection won’t suddenly perform better if you change a patch lead…


3 posts were split to a new topic: Building your own speakers

As a final thought from me on this subject:

For those of you who stream Red Book or Hi-res/MQA stuff from the likes of Qobiz or Tidal - have you considered where their servers are in relation to you? How many patch panels, switches and cables the data travels through? How many hundreds or thousands of kilometres/miles of fibre optic cable and numerous repeaters it travels through before reaching your ISPs infrastructure? And then how many more cables, switches, fibres, fibre/coax nodes before it finally arrives at your router? If you’re still on good old ADSL, you’re streaming over twisted copper pairs designed for analogue telephony, not high speed digital data! And you seriously think a different switch or different cable, or fibre in place of copper within your house makes the slightest bit of difference to how it sounds when it comes out of your streamer/DAC?


As a final answer from me on this subject:

I don’t THINK that it makes a difference, I hear and therefore KNOW that it can make a difference. Same applies to power cables by the way.


There are all sorts of reasons why you can perceive a difference, not least of which are expectation bias and the fallibility of human auditory memory.

Let’s agree to disagree.

Don’t even get me started on power cables…




20 years into this hobby and it always comes to this same debate on every single forum, exhausting. What I can’t understand are people feeling need to comment on these subjects with no personal subjective experience. It’s not your loss if some of us want to experiment and actually listen to instead of measuring their systems. It’s not your money spent, please let people be audiophiles as they please.

I always say that only thing which matters to me is what I hear/feel/sense when I plug a component in my system and press play. I couldn’t care less what measurements tell me, they’re not what I experience at home while listening to my system. People arguing about blind testing… who the actual f*** listens to their system blindfolded at home? That’s ridiculous. I will always be aware of every single component in my equipment chain and if that knowledge has some effect on what I’m hearing then so be it. That’s the real world situation and it can’t get any more real than that.

We do not only hear with our ears but also feel the music. There’s other senses involved all the time. For example speaker isolation vs coupling to floor. Frequency response stays the same while measured but system sounds completely different when speakers are isolated and the whole room doesn’t vibrate with the speakers. But I guess this is also just imagination since it measures the same.

I’ve spent hours on ASR and read many reviews (or measurements) telling me that gear, which I hear sounds different than some other, should sound exactly the same. After seeing the measurements, I can still hear the difference. So why should I care? I heard clear improvement with 8Switch so it stayed in my system, simple as that.


As much as I find myself drawn to reading these posts (and therefore wasting hours of my life), I would like be able to mute seeing them in new and unread as I always read them.
As long as people have the money to spend then be happy for them, even if you think they are wasting their money



I do listen to my system and for hours on end. I don’t feel the need to constantly spend money and constantly tweak in the hope of achieving better results. Our hearing is extremely unreliable and we get accustomed to changes in sound very quickly. You can’t get away from measurements at times. Every time I change the tubes in my preamp stage, I have to break out a multimeter and balance voltage and gain. Without that, one channel ends up sounding louder than the other.

Blind testing isn’t about listening blindfolded, it’s about not having any visual indication of which components are in the chain when making comparisons. It prevents visual bias of the results. It’s also important to level match the volume between compared components, otherwise we naturally favour the louder one!

Speaker response changes when speakers are isolated vs when they aren’t. The energy is distributed differently and measurement techniques pick it up.

Hey, in the end it’s your money and your choice. I just hate seeing snake-oil being peddled under the guise of audible improvement when there’s absolutely no electrical or physical reason which it should sound different. Our senses are easily fooled.


No, but you can optimize a great speaker with quality room correction filters implemented with Roon’s convolution engine.

I recommend Home Audio Fidelity DSP as an affordable -and really great-sounding- addition to proper placement, isolation and any physical acoustic treatments for the critical speaker-room interaction.

Exactly. You should consider 1) measuring with the REW program at your listening position 2) sending those measurements to HAF and 3) installing the resultant filters which will address both frequency and timing issues for those speakers in that listening environment.

1 Like

and… The deeper I read into “reviews” from those who chase these “upgrades” the more I’m convinced a lot of the “high end” gear is simply poorly designed. You never see someone swap-in a 1k DAC for their 10k DAC and join the forums thrilled with the reduction in “noise floor”. However, there are hundreds (thousands) of people with that 1k DAC who cannot understand what “noise” people are talking about. It’s not because their system isn’t “resolving enough” to hear the noise its because the 1k DAC isn’t noisy like the 10k DAC. Instead, I see those with the poorly designed gear outspend themselves with linear power supply or conditioners, fiber, fancy switches, cables, etc. but no one would have do this if the thing was designed properly in the first place.

Anyway, between poorly designed gear and colored gear not sounding right… I think a lot of problems could be solved if people would put a moderately priced system in their home that simply measured well as a reference. Instead, they chase the more expensive upgrades trying to upgrade gear that is, frankly, not very good with more expensive gear that is, frankly, not very good.

When someone raves about how much better their system is because of an ethernet cable I don’t buy that ethernet cable. I cross off my list the thing that cable is plugged into.

*and I’m someone who has a decent amount spent on nice interconnects and speaker cables because I think it matters… well… because maybe my amp isn’t so well designed.

Anyway, thanks for your insight.


I’m happy for you, you’ve achieved final state with your system and don’t feel the need to update anything. Maybe some day I will also. I can go for years with same gear but at some point I usually get a spark to try something new. If it improves my setup, it stays. Currently I’m considering upgrading my speakers. That usually won’t cause this kind of debacle since we can all agree that speakers matter.

I know what blind testing means. What I’m trying to say is, it’s useless. I can do blind testing with my gear and maybe not experience as big differences as before but guess what, when the blind testing ends I’m once again fully aware of the equipment chain and all my other senses and brains are back in game. So the experienced difference won’t disappear after blind testing.

Offtopic: According to IsoAcoustics anechoic chamber measurements, frequency response stays pretty much identical with spikes vs isolation.

This is precisely why I bought my Meridian DSP system, I listened to Meridian for years, I like the sound and the design philosophy. Speakers, cabinets, amps, drivers DSP crossovers all designed to complement each other with one goal in mind… Accurate sound.
Then the EBA for true bass alignment and simple DSP options for room compensation. The results are just not achievable with a hot Rod system, which only leads to more add ons and experiments…
Hell, that’s fine if a part of this hobby is building a system and enjoying the process, with all its ups and downs. I can understand the attraction, like building or refurbishing classic cars/bikes etc. You enjoy it for what it is and enjoy the process. It’s not for me though.

Pretty much all the differences I’ve experienced with digital gear, cables, switches etc. have come down to one simple thing, electrical connection between equipment. Best results can be achieved when there is no such connection (like unshielded ethernet cable) or the isolation between these devices is done to perfection.

For example USB bridges which connect PC to your DAC. When these first came out ~10 years ago there was no async USB, galvanic isolation or separate psu’s on these things. Usually just USB stick with digital out on the other end. I started with this one and during 10 year period I used this connection method, these devices improved greatly and I experienced improvements when I upgraded to better bridges with galvanic isolation and external psu’s etc. Last one I used was Mutec MC-3+ which even reclocked the signal.

I was told back then that it can’t be so since it’s just data moving between PC and DAC and 1 is 1 and 0 is 0. What people often forget is the electrical connection and electricity which moves through our system. It’s not about data in this case, obviously data moves perfectly even with the most basic cables and gear. It was the noise and crap from PC which could freely pass to the hifi system. But surprisingly things can evolve and improve, even on digital domain. Measurement god Amirm on ASR wrote about this subject back in 2016 and measured clear differences between different USB bridges. Obviously nowadays things should be better if you still choose to connect your gear through USB. I moved to streaming eventually.

I can’t explain many differences I experience. I used to laugh to people comparing ethernet cables until I compared few myself and experienced difference between an unshielded and shielded cables. Unshielded bulk cable worth 3 euros worked better than expensive shielded audiophile cable between ethernet wall socket and switch. I ended up ordering BJC Cat6a cables which have floating shield design, so proper isolation but also shielded. Not too expensive either. Now I have peace of mind with the subject.

1 Like

This one area where the measurements really come into their own. If it measures poorly, it’s hardly a competent design…

High end audio gear often has a characteristic ‘sound’ and people talk about ‘system synergy’. You hear terms like ‘forward-sounding’, ‘laid back’, ‘warm’, ‘analytical’ and ‘detailed’. If a component is introducing it’s own sonic signature, then it’s fundamentally flawed in its design.

A moderately priced tonally neutral system is all that’s really required. Then it’s a case of finding loudspeakers that are suited to the room size, finding the ideal placement and dealing with the room’s acoustics to get the most out of them. This is an area that’s so often overlooked and yet it’s the most critical in achieving the best imaging and soundstage…


No one ever told you that this is not possible.You are talking about usb to spdif converters here. At the spdif side that is were timing matters. You cannot compare spdif to usb to ethernet, it is a different protocol. It is packet transfer vs a real time stream, those are two completely different things.

Like I have said many times before, it all boils down to a proper grounding scheme, it has nothing to do with the cable itself. You have experienced this first hand. If you don’t know anything about grouding just stay away from shielded cat in your home. The chance that it screws things up are much higher than the chance it will improve anything.

Couldn’t agree more but unfortunatly the loudspeakers you described with a 6,5" woofer and 1" dome tweeter has about the worst in room power response you can imagine. They don’t have any constant directivity behaviour whatsoever. The directivity is all over the place. Besides a 6,5" woofer is a distortion generator for anything below 100hz. It will allways sound at best like a very good loudspeaker but not like real life, it takes more than good drivers and crossover parts to get there. Unfortunatly many commercial offerings nowadays are not very good in this aspect.