I think it is the analog aspects of the switch + ethernet cable that can cause some SQ impacts - not the digitally encoded content - the so called “1”'s and “0”'s.
Presently, I have my Router which feeds in to a plain vanilla TP Link Gigabit switch (No. 1), which in turns feeds my CCTV, another TP Link Gigabit switch (No. 2), which feeds my PC, Printer and Laptop.
I am getting a Linn Klimax DS3/Katalyst streamer and will be connected to the No. 1 switch.
My question is;
- if I get an EE8 switch, can I do a straight swap with the No. 1 switch and will the No. 2 switch with PC, Printer and Laptop cause no conflict with the streamer or, do I need the connect the EE8 switch to the No. 1 switch and only have the streamer going into the EE8 switch?
I am getting a Linn Klimax DS3/Katalyst streamer and will be connected to the No. 1 switch.
My question is;
- if I get an EE8 switch, can I do a straight swap with the No. 1 switch and will the No. 2 switch with PC, Printer and Laptop cause no conflict with the streamer or, do I need the connect the EE8 switch to the No. 1 switch and only have the streamer going into the EE8 switch?
Agreed, I was just trying to help people save money by not spending it on something that is pretty subjective and no quantitative proof have been presented, but in the end it's theirs and they spend it however they like.
It's not at all subjective. Either a switch makes a significant difference or it doesn't.
A switch made no discernible difference for me, a change to an optical cable from the switch into the NGKDSM however did (an obvious and much more cost effective improvement).
Yes it is just another unmanaged switch so it can replace any other unmanaged switch. Ethernet is a standard so one size fits all. A managed switch is a little more challenging, but you don’t need one. I would avoid any boutique streaming cables and stick with standard Cat 5e (e.g. Connectix or Blue Jeans Cables). I made the mistake of having a boutique cable on trial while doing an update to a DSM over the network - it locked up. On Linn’s advice, I got rid of this cable and all was well.
Be interesting to hear if you decide the English Electric switch is as good as your TP Link. I couldn’t hear any difference between it and a NetGear at a demo and at home I can’t tell a TP Link from a NetGear. However, different room, different system and so on and the results may be different.
I bought an ioniser rather than boutique switch, it helps reduce the amount of dust in the air. I think it has a positive impact, but it is probably just my imagination. Makes my Placebo albums sound better.
I would think that copper Ethernet will inherently carry noise, no matter what switch one uses. The fibre optic Ethernet is completely different, and inherently won't carry noise.
Noise can be measured. I have yet to see an audiophile switch seller present any data on actual noise reduction. Haven't seen any controlled listening test results either.
I use my wife's ears. She can pick UPNP over Roon 100% of the time. She can pick Innuos PhoenixNet in or out of the system 100% of the time. That of course confirms my own thoughts, but I like to be sure.
I’m far from convinced boutique switches are quieter than high quality switches, especially if you use Power over Ethernet. I would go as far as to suggest the funds available to NATO, the US Government, HMG and their major suppliers to develop routers and switches that do not emit RFI/EMI (noise) in discernible quantities exceeds the budget of boutique audio companies.
However, I’m happy to accept some folk prefer boutique switches. I cannot explain why this is as it is way beyond me. My father owned a Pioneer amplifier, which was modified by GTE. After the mods, it actually measured slightly worse; nonetheless I had to agree it sounded better. Perhaps, it was psychological or maybe it was magic, but in the words of Pete Townsend “I Can’t Explain”.
So my guess is boutique audio companies do not publish their measurements because they are worse than those for Cisco, HP etc. I wish marketing men from boutique audio companies would desist from writing white papers full of unscientific nonsense and just say listen and decide for yourself. It may be false, it may be true, but nothing has been proved.
I wonder how many companies making these products have even rented anechoic chamber space and test equipment (no audiophile start up is going to be able to afford to buy this stuff) and actually measured anything properly?
I've been looking after some applied rf work lately (I'm a researcher with a government agency, but not an engineer myself) and working with actual rf scientists and engineers has really been an really an eye opener.
I've shown them some of the audiophile cable and switch webpages and they are quite amused.
Many years ago whilst working in a well known Hi-Fi store, I was a sceptic. We had our NAS in a cupboard about 20m away from the dem room feeding the wall outlet. A rep in came from a well known company and plugged in a (1m) better Ethernet cable from the wall to the Klimax DS. I didn’t expect to be convinced. How wrong I was. I don’t have all the answers but there’s a lot more going on than just passing around one’s and zero’s. I would urge anyone to at least listen to a good network switch. If you can’t hear a difference then you’ve saved yourself a few £££.
I was a sceptic too. In fact, I was so influenced by the “I’m an IT expert” guys on hifi forums that I resisted my dealer’s suggestion that a Melco switch might remedy my Roon sound quality issue. He offered to loan me an S100 and I tried it out of curiosity. I fully expected that it would make no difference.
Even after hearing a significant improvement, I was still doubtful so I set up a blind ABX comparison. It wasn’t easy to do this experimentally because the KDSM didn’t reconnect to the network quickly enough. Nevertheless, the difference between the Melco and a Netgear switch was obvious. Compared to Linn prices, the Melco was a cost effective upgrade.
That was about eighteen months ago and I have since upgraded the DSM to a Next Generation Hub. Later, I changed the connections to optical SFT. Each time, I repeated the comparison to see if the Melco Switch became redundant so I could sell it. Inconveniently, it still makes a difference with the new Hub and optical connections. The next step will be to try the test with Organik upgrades to the Exaktboxes.
It might well be that cables and switches make absolutely no difference in many cases, but it is a foolish person who dismisses them without a trial on the advice of the self professed IT experts online.
Last edited:
It always amazes me in this hobby that the experts say "it can't make any difference" Remember it was only 50 years ago that loudspeakers were the only thing in a hifi system that influenced the sound. The source was not considered important. We all still have a lot to learn about sound quality in a network based system. I'm not pushing snake oil and only expensive products. There may be some great bargains out there that do sound good? It's just finding them.
IME. Back when a TT was the main Source - the recommendation was often that it was worth spending half your budget on the TT and the other half split between Amp and Speakers. This was because the competence of the TT and its effect on the sound, was considered so paramount.
- Jun 22, 2013
- 12,479
- 8,343
- 208
- AKA
- Nigel
- HiFi Trade?
-
- No
I’ve spent 40 years in IT so recognise a 1 or a 0 when I see them.
I can clearly hear an improvement even with a budget switch in place. It needs to be at the streamer end of things not near the router. Packets are packets and arrive unaltered and pristine, and ethernet as a multi-layer protocol involves all sorts of checks and resends, so timing is also probably a red herring in ethernet-land (though not in DAC-land). The answer has to lie with analogue noise picked up along a CAT cable. The switch repackages and forwards packets of data and most (all?) switches have some sort of galvanic isolation, so any noise accumulated up to this point is stopped in its tracks.
A genuine audiophile switch will be better than this if it does even more to address noise: uses high quality=quiet components so as not to introduce new noise of its own, and or has an enclosure which blocks external RFI/EMI from getting in. That sort of thing.
It’s not a figment of anyone’s imagination.
- Jun 22, 2013
- 12,479
- 8,343
- 208
- AKA
- Nigel
- HiFi Trade?
-
- No
It’s nothing to do with belief. You can test this for £30ish and the only equipment you need is a pair of ears. The switch has to be installed in the right place though, and a surprising number of “I didn’t hear a difference” folk fail to mention where they installed theirs. It’s pretty fundamental.
Hope this helps.
In a sighted listening test, expectation bias can't be avoided.
Similar threads
