I run an Ethernet cable between my router (standard issue from Verizon) and my streaming transport. I note that some use an ethernet switch between between the router and streamer. Assuming I got that right, what is the point- what does a good switch do? I've been into audio since the 70's but when its comes to streaming, I'm definitely a newbie-
Many or most people need a switch to run other equipment. A switch is what the ethernet needs to split a signal. If you want to run ethernet to 2 peices of equipment, use a switch. I have other ethernet cables running to my Roon Core, TV, AV receiver, and power plant.
Now "audiophile" ethernet switches are designed to improve delivery of error free bits to the streamer by use of improved power supplies, grounding, shielding, clocking, and other methods. So some people will want to put the switch in even if they don't need cables to other components.
The type of switch usually talked about with regard to streaming is one that allows the musical bits through but not the electronic noise that comes along through your wifi system (cable into your home and wifi router). Some switches offer galvanic isolation, others turn the electronic signal into light transmitted by Toslink cable. The diehards use several of each kind of switch to get best results.
I'm sure someone who knows much more than me will chime in, but that's my basic understanding of what's going on.
Not really needed, but it can be a convenience feature. I have the Roon streamer, TV, Roku, etc. all 50’ away from my router, so the switch lets me run 1 cable (in-wall in this case) and hook up all my devices.
You'll never be able to convince me that an audiophile switch is worth it, but I do use an Ethernet isolator to reduce the chances of surges and induced noise.
Hi @oldears- In my case due to the 50' run I'm more concerned with surge issues than noise. The longer the Ethernet cable the more noise/chance of an EM induced surge pulse. I put the isolator right before my home theater switch.
If I had more money I'd put one right before the streamer as well.
I do also run iFi power supplies for my streamer and DAC.
The switch in every system we have tried really makes an improvement in clarity, noise floor, etc. Silent Angel makes an inexpensive one that works very well.
It is not one you have to go back and forth to hear the improvement either.
A good Ethernet switch such as the SW8 is the best I have compared
For it has low noise power regulation , a Ocxo over controlled clock ⏰
and Linear power supplies with lots of capacitor filtering a night and day better setup ,and BTW the piece of crap you have wall wart for your router I changed to a 12v Linear power supply with decent power cord, and upgraded fuse sonically
much cleaner sound ,now with the SW8 Ethernet switch with a for example pangea power cord a nice boost in fidelity clean power counts , removing the weak links
in the chain. IHave 40 years in Audio ,ain’t much I have not experimented with
my brother spent over $3k on his innuous Ethernet switch and it’s muchbetter sound then mine. If you have a expensive setup the innuos is a must hear,
for myself for the moment for $595 the SW8 Ethernet switch has pretty decent reviews ,I made it much better still with the Pangea sig mk2 power cord ,
land synergistic purple fuse , yes an additional $350 but a solid increase in imaging
and realism little money for a noticeable improvement , start with the SW8 Ethernet switch ,the good oven clock takes at least 8-10 days to fully settle in.
@erik_squiresIf a surge is of concern to you adding a fmc ( fiber media converted) will eliminate and surge that could damage your equipment. It will also improve the sound by eliminating the electrical noise inherent to cate cable.
I need a switch after my modem to serve several feeds (two audio systems, TV)
I replaced my Netgear switch with a Bonn 8 and heard some improvement. I replaced this with a JCAT M12 Gold switch and easily heard a big improvement. I then upgraded the power supply to the JCAT Optimo 3 and again easily heard an improvement.
How can an audiophile state they can never be convinced it makes a difference? You don't need to understand it, just hear it. I can.
“Now "audiophile" ethernet switches are designed to improve delivery of error free bits to the streamer by use of improved power supplies, grounding, shielding, clocking, and other methods.”
Switches don’t improve the error free delivery, all packets are guaranteed to be error free or they will be retransmitted. How many times have you submitted a stock purchase and the broker bought you more shares than when you asked for (packet of data sent got corrupted and changed the 100 shares to 1000 shares)?
How about all the database data that is read from the server, does any of this data gets changed during transmission? Not a chance! If Ethernet packets were prone to errors, the financial and brokerage houses will be a disaster.
Another alternative to a switch is to have multiple routers setup in bridged mode in your house to use the ports on the back of the router to wire your dac, streamer, and server to
$1,349.00 for a LHY SW-10 is a hell of lot to spend on something that might not make much or any improvement. We'd probably all go broke it we adopted the recommendation to try for ourselves.
$1,349.00 for a LHY SW-10 is a hell of lot to spend on something that might not make much or any improvement. We’d probably all go broke it we adopted the recommendation to try for ourselves.
the sw8 and the english electric unit are under $600 msrp, less on used sites when they appear, etherregens regularly trade for $500, and so on...
like anything else, if you care enough you will figure out how to try something to see if it makes a difference
and then, there are certainly money-smart and money-stupid ways to try things...
high end hifi is a pursuit that does require money, time, effort and know-how to achieve a high level of performance
@p05129You have demonstrated the exact and most common misunderstanding about streaming. This error even exists among IT professionals. Why? Because what you stated is exactly correct for the applications that you referenced. I used to make the same argument. Dropped bits cannot be tolerated in banking, defence, big business, etc. AND IT TOOK ME A LONG TIME TO GET A STRAIGHT ANSWER. Why? because most don't know the answer--including many so-called experts. Here is the difference:
Business uses File Transfer Protocol with error checking that will not let file transfer complete until the transferred file is EXACTLY like the original. Bit perfect. but that's why sometimes it takes forever to download a file.
Streaming doesn't have the luxury of taking whatever time it takes for error checking to make sure the file is perfect. Instead, it has to use what is called "error correction". I'm no expert but as I understand it, it does it's best to interpolate for missing bits and moves forward. So bits that get dropped, and there are lots of them, are estimated. Of course video can be even worse.
If every file stremed was perfect, there woud be no need for any high end streaming equipment, any high end digital cables, etc. And this explains why you can never stream quality music wireless.
Just to be clear, I'm not criticizing you. I started with the same misunderstanding.
Melco makes two Ethernet very effective network switches - S100v2 and the top line S10-X. The S100v2 can be upgraded by adding a linear power supply like a Plixir Elite BDC linear power supply. The Melco S10X comes with it's own linear power supply. As well, Melco's sister brand ADOT is a fantastic way to isolate completely your audio system's network from the rest of your home networking. It takes the feed from your gateway and converts to fibre, which you then connect the ADOT to your Audiophile Switch. These must be heard - people who say it all makes no difference are ill-informed. These products are distributed in the USA by Luxman America.
TCP/IP guarantees delivery 100% error free 100% of the time. No switch or cable can give you better zeroes or ones. Noise, if it exists, is introduced after the fact. Some streaming uses another protocol called UTP which does not guarantee packet delivery, but it is not used in any audio streaming I am aware of, and rarely used any more with video. Differences, if they are perceived, are either confirmation bias or analog perturbations further downstream due to DAC filter design, poor power supply isolation, or induced noise.
As for the claim that 'streaming fonts is not the same as streaming music'. You are misinformed. At the network layer, wired, wireless, fiber optic, it is exactly the same in all respects. That's the beauty of TCP/IP, it simply does not care about the nature of the payload or the media over which the payload is carried. It delivers packets, error checked, retransmitted if necessary, packets irrespective of the payload. And audio, even 32bit X 384kBits/sec are not even remotely strenuous loads for modern networks.
sure, there are some legacy business applications that still use ftp internally to an organization, however 99.99% use HTTPS for everything. And Qobuz, Tidal and others also use HTTPS. We go through this every time….
A quick primer on HTTPS, it is build on top of TCP, which delivers error free transmissions, and it is encrypted, which means that if you lost one packet, which TCP prevents, you wouldn’t be able to decrypt.
Exactly. Most audio/video protocols are based on UDP. More they are real-time, more they use fast protocols which cannot permit to make any check.
TCP/IP controls correctness of every packet and the order in which packets are received. Every x packets, an "OK" packets is sent back to the origin, stating trasmission is ok, delete old packets in buffers and to go on with new packets. Otherwise a "request" to retransmit is sent back is a packet is faulty or missing. All this is expensive and heavily compromizing for the quality requested to a simply phone call. This dialog is between the origin and the end point, crossing all intermediate routers/switches and any communication equipment.
UDP instead does not control any packet, does not control any flow. If a packet is missing or faulty, it's simply gone. It's better to skip it instead of requesting it again.
Every commercial transaction, every printer dialog, any USB disk uses the TCP/IP protocols.
The most of audio/video quality communications use UDP protocols.
Just to make it more clear, a packet coming from a whatever source (TIDAL, qobuz, etc) needs at least 25 ms in the best situation, and more than 100ms in a normal situation. So what you are playing is something sent to you a lot of milliseconds before. Using UDP, you receive a constant stream, with a very regular flow. Using TCP/IP you will freeze the sound every X packets, and also when ask to resend a packet. This latter situation is totally unnoticeable in a commercial transaction, but would be a disaster feeding a DAC. Of course you may use very large buffers, but starts to be a file transfer more than a streaming.
@tonixThanks for the help and support. I'll remember UDP for the next time I need to explain this. I thought I was the last one to figure this out but this thread shows there are still lots of people who think streaming is bit perfect. -Jerry
@yoyoyayaWell, Occam's razor applies. If there is not way a switch can improve anything, which is a fact and how Ethernet and TCP/IP is designed, guess what is the next option?
OCSP: could not connect to server. Make sure the server is up and running.
Certificate DN: 'qobuz.com' Requested Server Name: www.qobuz.com. See sk159872
And before you say you don't understand pcaps, yeah, that is the whole point. So stop lying.
Why do I call it lying? Because you have been told over and over again what the reality is, and you continue deliberately, willfully and intentionally spew the same garbage.
https://en.wikipedia.org/wiki/Streaming_media
https is TCP/IP, very large buffering, nothing compareable to real-time o live streaming. Probably in your life you never made a video conference using TCP/IP, otherwise you'd know the effects.
zoom, WebEx, teams also use HTTPS. That Wikipedia is a generic overview that doesn’t go into specifics regarding services like Qobuz, it does however mention that HTTP transport to overcome lots of issues that real time protocols run into over the internet.
Pro audio and video however use real time protocols and absolutely no crap Etherregens or other snake oil switches. But pro audio and video are on low latency, very high speed private networks, and not over the internet.
@yoyoyaya It's not an authority fallacy, it's simply basic facts. However, not everyone understand these facts, but they want to portray themselves as have an understanding.
Take the example of John Swenson, founder of UpTone and designer of Etherregen. He was forced to admit that he made up terms, terms that does not mean anything, for his intended customer base with the Etherregen. Why forced, because he was confronted by actual engineers.
So again, a switch cannot improve anything, this is a basic fact. A really good switch doesn't add too much latency, no jitter, or packet loss, while a bad switch introduces lots of bad things. Luckily, most $30 switches are more than capable to handle most residential applications.
@carlsbad2 This is where you go every time you get actually proof of how things work. Can you just stop spewing the garbage you normally do, since it is only lies?
And why? People just don’t search the forums. Easier to post a new thread and agitate that barely had a chance to develop some dry crust pile of 💩 to ensure that it continues to reek.
One day when I’m bored I might start about half a dozen ethernet related threads. And heck, while I’m at it, might as well just start few “why do cables matter” threads. Drop that massive deuce into the forums and run.
Oh Freddy Freddy .... @fredrik222.... you never disappoint. Guaranteed to post your usual propaganda on every (and only) networking threads.
Take the example of John Swenson, founder of UpTone and designer of Etherregen. He was forced to admit that he made up terms, terms that does not mean anything, for his intended customer base with the Etherregen. Why forced, because he was confronted by actual engineers.
So many misinformation and outright made up lies on this part. John Swenson is no founder, or even owner of UpTone Audio. Alex Crespi is the founder and the sole owner. John Swenson was hired from UpTone, and he has done work for many audio companies. (i.e. Sonore Audio, and many more).
Besides, can you share a source for the "John Swenson admitting he made up things" part of your post? Any links citing John Swenson saying that would do.
...stop lying. Why do I call it lying? Because you have been told over and over again what the reality is, and you continue deliberately, willfully and intentionally spew the same garbage.
You call it lying because this is what you do here: copy and paste of charts and data, then engage in the ill logic of ad hominem personal attacks. It really is an obstacle to the discussion others are trying to have.
@cleeds I was wondering when you would show up, you and @carlsbad2 are in the same boat, you think you know things, but you really don't, but you continue to argue. Why? You really have nothing better than fill up forums with garbage that is completely, easily verifiably wrong?
You must have a verified phone number and physical address in order to post in the Audiogon Forums. Please return to Audiogon.com and complete this step. If you have any questions please contact Support.
