Anybody want a laugh?

Here are some references for those who want to understand:
https://www.ti.com/sc/docs/apps/msp/intrface/usb/emitest.pdfhttps://www.usb.org/sites/default/files/crcdes.pdf
https://www.potaroo.net/ispcol/2001-09/2001-09-best.pdfhttps://networklessons.com/cisco/ccna-routing-switching-icnd2-200-105/introduction-qos-quality-servi...

For the best effort networks (that's what you have inside you own house!), you could use pro routers / switches that support QoS (Quality Of Service).  QoS will help for those who use streamers and have a busy network.  You know, games of your kids, Netflix, big downloads, etc!

Here is one of the best little router money can buy:https://www.amazon.com/MikroTik-RBD52G-5HacD2HnD-TC-Dual-Concurrent-802-11a-Ethernet/dp/B079SD8NVQ

Don't let the price fool you: It's really a pro product and we have many installed in SMB clients offices.  It easily beats models costing 5-10X more.  I use his bigger brother (RB4011) with more ports in my own house.

But there is indeed a learning curve, it's pro stuff.  One has to understand networking, be good at googling and not afraid of reading / testing.
Those "audio networking products" should just concentrate on real important topics and provide a friendly user interface.

@dougeyjones That's right.  There are people here that are making profits out of plain lies...  The stupid "bit argument" i hear so often is so DUMB...

As for DD-WRT, i used it a while back on Linksys WRT54G, it was so much better that the stock crappy Linksys firmware!  Cool if it fits your needs!
But if you're a bit of a hacker / techie, buy a Mikrotik and try RouterOS, it's really top notch. It supports OSPF, MPLS and even BGP!!!  For QoS, sky is the limit so for people streaming in busy networks, you have complete fine grain control.  You could even download a tryout for x86 and run it inside a VM.

@glupson It's up to whom is interested.  I prefer to give more infos than less.
I just hope it can help audiophiles make good choices for themselves and better spend their money for the ultimate goal: Enjoy music and good sound!

The more i read this topic, the more i see that a large part of people here don’t understand how networks work. Yes it’s really about 0’s and 1’s and from end to end, TCP/IP network equipment calculate / validate a checksum. (UDP/IP does it too)

So for layer 3:
http://www.tcpipguide.com/free/t_TCPChecksumCalculationandtheTCPPseudoHeader-2.htm

As for layer 2 (ethernet), you have another checksum called FCS:
https://en.wikipedia.org/wiki/Ethernet_frame

About the transit time, "jitter", etc: There is an easy remedy for this and it is called buffering. A TCP/IP network is a best effort network. You have to design any solution around this.
So a 20$ switch will do the same as a 50K$ one. Why the difference of price? Horsepower, # and type of interfaces, high end enterprise management, licences for specific features, modularity, etc etc etc.

So you can place whatever capacitor or anything else on your switch / product router (even gold and diamond if you want!), it’s up to you. But somebody somewhere will easily beat you if you throw in useless expensive parts... For those who does it (or try to say they do), it is only ripoff of customer who don’t know how a best effort network works and proper design around it.

@nonoise Not knowing how digital works really shows you need a primer on electrical engineering and design... Same applies for networks.

About the link to this fallacious article you provided, it is only trying to look like "technical" but only to fool people into buying expensive useless hardware... Audio is not transmitted "real time" to DAC so to speak, it's transmitted asynchronously and DAC has input buffering.  Again, an easy engineering thing...

How do you think CPU works in the gigahertz spectrum with billions of micro-transistors in multi-layer arrangement + in a mutual noisy environment without making any error? It’s not even logic at TTL levels, it’s lower than that! Billions of micro transmission lines, via, metal layers, etc. In the same enclosure, you have the GPU, the chipset, etc. all operating in gigahertz range.

So you’re sweet little low spectrum audio is really not an engineering challenge... The only challenge is external constraints like budget, contracts, quality control, compromises, etc. If you have any minimal knowledge in engineering, you should know that... As for networks, TCP/IP is so robust, you just need proper buffering to account for the fact that it’s a best effort network.

But hey, you can throw your money at the windows, it’s all up to you. It’s just disgusting to see how many people here on this board show complete disrespect for consumers....

@nonoise  Yeah i know, you don't have "a dog" to respond to real arguments...  Your previous response "fueled" mine trying to sound superior and shut down real world digital principles...

"Knowing how it works doesn’t preclude that something poorly designed won’t have an effect".  This thread is about a network switch ripoff and abuse of consumer...  I'll repeat again and repeat with me: TCP/IP has been designed for best effort networks.  So what is you don't understand in that?

So after you add: "I have a funny feeling that there are some who actually believe that since it’s just ones and zeros, that they’re so small as to be unimportant. That’s fine for printing text but reproducing sound is another matter."
"So small to be unimportant":  You think that i, and engineers in general, don't know about the importance of errors in digital in general and how to handle it?

Not enough, you add: "That’s fine for printing text but reproducing sound is another matter"... You're trying to dismiss what i said and mixing things but at the end, it's the same.  Printing doesn't work if you send errors, be it on any interface / transport medium.  You really don't seem to understand digital in general: Storage, processing, transport, error correction, etc.

So the reason of my response.   I know what i'm talking about and Digital / network devices must be designed for the environment they are into.

Same apply for USB cables and the crap link you cited is just another of those stupid "trying to look technical" BS to push people wasting their money...
Computer interfaces, be it USB or any other, are designed to transfer data without errors...  Same apply for the USB interface of a DAC.  If you think it takes expensive cables, it's because you don't know about twisted pair, digital interfaces, checksums and async communications principles...  It's as simple as that.  Flip a bit and the checksum tells...so the idea of async communication + buffering.  If the DAC you buy can't handle that, it's simply because of a poor design.

So find back the "dog" you use at first toward me, barking doesn't impress me at all...  And study the subject before throwing your "funny feelings" around here...
All the best, my friend!