dbastin Posted April 11, 2020 Author Posted April 11, 2020 Thanks, do we assume Covid 19 precautions occurred before you could try either Paul Pang? And, your view about EtherRegen in your system has not changed? I think Ive said before, Mark Jenkins at Antipodes must be a genius to create a switch in CX that outperforms ER. Did you try optical into ER and CX on A side and EX on B side? And the ER > CX link, was the sheild grounded? Next you might want to consider alternative SFP modules, apparently they dont all sound the same.
Stereophilus Posted April 11, 2020 Posted April 11, 2020 1 hour ago, dbastin said: Thanks, do we assume Covid 19 precautions occurred before you could try either Paul Pang?I ER arrived late Feb, and the optical fibre + SFP module did not arrive until March 20. I could have organised a comparison with the PP switches early on, but I’m not sure how valid that would have been (at least not without letting the ER burn-in). 1 hour ago, dbastin said: And, your view about EtherRegen in your system has not changed? No. But do not see that as my dismissal of network devices affecting SQ. 1 hour ago, dbastin said: I think Ive said before, Mark Jenkins at Antipodes must be a genius to create a switch in CX that outperforms ER. Did you try optical into ER and CX on A side and EX on B side? Yes, it’s a relatively easy thing to compare B-side feeding CX to B-side feeding EX. What isn’t easy to determine, is cable effect... the Antipodes link cable is shielded (metal ends). This has always been my go-to for CX to EX link. Is it also the best link from B-side to EX? I haven’t explored this any further yet. Listening for pleasure is my priority right now. 1 hour ago, dbastin said: And the ER > CX link, was the sheild grounded? I’m using AQ cinnamon for this link. It has shielding and metal ends. I haven’t attached any other grounding to the ER or cables.
dbastin Posted April 11, 2020 Author Posted April 11, 2020 (edited) On 11/04/2020 at 5:38 PM, Stereophilus said: I’m using AQ cinnamon for If Cinnamon is made the same as Pearl and Forest, they should have open ground on the sending end and the shield grounded at the receiving end. You might wanna check if that works best with EtherRegen. Thanks for letting us know where you are with your exploration. Edited April 19, 2020 by dbastin typo
Cruncher Posted April 12, 2020 Posted April 12, 2020 (edited) To the op's original post. I have a non-audiophile (but well built) Ethernet switch with a stock standard Ethernet cables. I see a lot of discussion on noise, errors, packet loss. These are my network statistics. Name Value Counters 2020-04-12T00:26:07Z Packets Tx (packets) 14,120,530,243 Packets Rx (packets) 5,713,543,745 Bytes Tx (bytes) 26,587,020,300,500 Bytes Rx (bytes) 6,471,179,863,744 Error Counters 2020-04-12T00:26:07Z Tx (errors) 0 Too Short Rx (errors) 0 Too Long Rx (errors) 0 Sync Losses (errors) 0 Signal Losses (errors) 0 Rx (errors) 0 Link Failures (errors) 0 Discard Tx (errors) 0 Discard Rx (errors) 0 Crc Rx (errors) 0 From the stats above I have sent approximately 14 trillion packets with no packet loss and no corruptions (CRC errors) so I would say it is doing a pretty good job in carrying audio. Edited April 12, 2020 by Cruncher 3
Cruncher Posted April 12, 2020 Posted April 12, 2020 On 22/12/2019 at 5:48 AM, LogicprObe said: Well, you really want as little delay as possible for audio. The network delay is not really a problem for audio. For example this Ethernet switch ( tested using RFC 2544 - Benchmarking Methodology for Network Interconnect Devices https://tools.ietf.org/html/rfc2544 ) You can see this switch is adding on average 200 nanoseconds to the traffic. So if you hit play, the audio would be delay 200 nano seconds out of your speakers than if you did not have the switch in the signal path. If you consider when you blink your eye it takes around 300,000,000 nano seconds. The bigger problem is Packet / Frame Jitter, which is the variation of delay of the device, in this switch it is around 160nSec. The smaller the better, and this is incredibly small. It is not uncommon for devices to have measured jitter 10,000 - 100,000 times larger i.e mSec. End devices using de-jitter buffering can accommodate up to a certain level jitter successfully. Those interested interested there is tons of information on the Internet. I am familiar the work done by the IETF (RTP etc) and IEEE Y standards and the mathematical backing there is plenty of material if you look up Queuing Theory. 2
dbastin Posted April 19, 2020 Author Posted April 19, 2020 (edited) On 23/12/2019 at 3:25 PM, rmpfyf said: Would think the next stage is a NIC for audiophile purposes. Your wish has been granted ... in 2017 https://jcat.eu/featured/net-card-femto/ I've read people have experienced improvement in SQ (ie. audio generated by a PC server) even if ethernet is only used for control (not audio data) while audio data is output with say USB. An as ethernet outputting audio data, of course a consideable improvement. It ca be configured as a network bridge (rather than switch) to enable data connection directly to endpoint, like some top end servers (eg. Antipodes, Innous, Melco). Edited April 19, 2020 by dbastin
Guest rmpfyf Posted April 19, 2020 Posted April 19, 2020 38 minutes ago, dbastin said: Your wish has been granted ... in 2017 https://jcat.eu/featured/net-card-femto/ I've read people have experienced improvement in SQ (ie. audio generated by a PC server) even if ethernet is only used for control (not audio data) while audio data is output with say USB. An as ethernet outputting audio data, of course a consideable improvement. It ca be configured as a network bridge (rather than switch) to enable data connection directly to endpoint, like some top end servers (eg. Antipodes, Innous, Melco). Ah no, I meant a legitimate attempt. That's lipstick on a pig insofar as engineering for purpose goes.
davewantsmoore Posted April 19, 2020 Posted April 19, 2020 On 12/04/2020 at 8:57 PM, Cruncher said: It is not uncommon for devices to have measured jitter 10,000 - 100,000 times larger i.e mSec. Do you mean network devices or audio devices? A often touted benchmark for audio devices is their jitter should be <1ns. Although many will say that a lot smaller than this is important. On 12/04/2020 at 8:57 PM, Cruncher said: End devices using de-jitter buffering can accommodate up to a certain level jitter successfully. Yes..... there is always one or more clocks/buffers between the ethernet hardware and the audio hardware.
gwurb Posted April 30, 2020 Posted April 30, 2020 On 19/04/2020 at 2:20 PM, dbastin said: Your wish has been granted ... in 2017 https://jcat.eu/featured/net-card-femto/ I've read people have experienced improvement in SQ (ie. audio generated by a PC server) even if ethernet is only used for control (not audio data) while audio data is output with say USB. An as ethernet outputting audio data, of course a consideable improvement. It ca be configured as a network bridge (rather than switch) to enable data connection directly to endpoint, like some top end servers (eg. Antipodes, Innous, Melco). Whats next? Audiophile CPU, Ram, motherboards, computer power supplies? Why don't those components effect sound quality? Surely the computer bus and storage (including ram) introduce unwanted artifacts in to the network card?
Stereophilus Posted April 30, 2020 Posted April 30, 2020 7 hours ago, gwurb said: Whats next? Audiophile CPU, Ram, motherboards, computer power supplies? Why don't those components effect sound quality? Surely the computer bus and storage (including ram) introduce unwanted artifacts in to the network card? They do. Or what i should say is that there are many people who see and hear the evidence that these things influence sound quality, including me. There is a very long thread on stereonet addressing these variables and their relative affect on SQ. Antipodes, Aurender, Innuos and several other brands sell servers that have bespoke solutions to these variables. Not everyone will agree, but there is a lot of mounting evidence that the computer variables you mentioned affect SQ.
dbastin Posted April 30, 2020 Author Posted April 30, 2020 (edited) 7 hours ago, gwurb said: Whats next? Audiophile CPU, Ram, motherboards, computer power supplies? Why don't those components effect sound quality? Surely the computer bus and storage (including ram) introduce unwanted artifacts in to the network card? I guess that is what bespoke servers seek to address. Interestingly, inside my Antipodes looks pretty similar to a typical PC, however I gather Antipodes use them in a way to minimise noise - for instance low CPU clock speed, 'tuning' PSU ... I gather the aim of the JCAT NET Card and 'audiophile' switches is to minimise the 'noise' that gets passed on into your audio gear. Noise comes from lots of sources, I assume the NET card attempts to use components and design to minimise creating more noise, and passing any on. What's next? How about a MOAT like in EtherRegen in endpoints and servers - isolate server from the network, and isolate noise getting in from the network. Isolation transformers used with ethernet ports are apparently not enough. Edited April 30, 2020 by dbastin
gwurb Posted April 30, 2020 Posted April 30, 2020 Antipodes CX Intel i7 Realtek audio DAC Linear Power Supply Antipodes EX Intel Celeron Realtek audio DAC Linear Power Supply Aurender don't publish their details but some of their servers use switched mode power supplies and some use linear. Innuos is general intel CPU This is standard PC parts other than power supplies. What I've looked at has SSD drives. There are no audiophile CPU that run windows. There is no audiophile ram. It is the same as going to harvey norman and buying a general PC, and replacing the power supply. It will be much cheaper. If people hear a difference between a Dell and a Dell with a linear power supply then they hear a different.
dbastin Posted April 30, 2020 Author Posted April 30, 2020 @gwurb, you are welcome to have this debate, but this is not the thread for it. Thanks
gwurb Posted April 30, 2020 Posted April 30, 2020 45 minutes ago, dbastin said: @gwurb, you are welcome to have this debate, but this is not the thread for it. Thanks @dbastin, Thank you for pointing that out. I will make my point clearer, because my post was precisely on point but not clear in my statement: 'Audiophile' ethernet switches, and 'audiophile' network card mentioned above, and 'audiophile' servers all use standard computing components. The processor, memory and storage, etc is not separated in any way differently than on a normal switch/pc/network card: to be clear I mean component interconnect such as the processor plugs in to a motherboard and memory is on the motherboard etc, not main board is separate to a power supply. The biggest change between a regular switch/pc/network card to a 'audiophile' products mentioned above is a change to linear power supply. So to respond to the topic.. "Ethernet switches for audio - Part B: why a regular switch will suffice.": A regular switch will suffice because the components that it uses are the same as the components used in an 'audiophile' switch. Maybe experiment with changing a regular switch power supply with a linear power supply and see if it makes a difference.
dbastin Posted April 30, 2020 Author Posted April 30, 2020 (edited) On 30/01/2020 at 11:02 AM, dbastin said: Thanks for your observations and experiences. I set this thread up for these debates to occur. In hindsight I should have included in my opening thread ... Please share you suggestions for optimising regular switches for audio, and experiences with their sound quality, ideally in comparison to switches on the list in Part A. I am hopeful this thread can generate some solid recomnendations for those not interesred in the switches listed in Part A. Such as, good switches, configurations, tweaks, etc. Maybe the thread should be called ... how a regular switch will suffice. I made this statement back on page 5 of this thread in response to some discussion then. While the 'audiophile' switches, NICs and servers may be using common components, I suspect the point of difference is how they use them: the design and implementation, firmware, software, etc. In my experience, different regular switches can sound different from one another. My Ubiquiti EdgeSwitch is quite superior to the previous old thing I was using. And LPS improved them both, as did audiophile power cord. The Ubiquiti EdgeSwitch, with my audiophile ethernet and power cables that cost 20 times the EdgeSwitch, actually sounds very satisfying to me, so it can certainly suffice. Its currently just stock, so maybe some optimisation can make some further improvements. I will be tinkering with settings soon. Then I'll insert EtherRegen and be either pleased or disappointed. Edited May 4, 2020 by dbastin typo 1
Cruncher Posted May 4, 2020 Posted May 4, 2020 On 08/01/2020 at 10:05 AM, recur said: um, no Cisco switches run Cumulus. You could easily get a 24 or 48 port gigabit Cisco switch (refurbed) for circa $500 though. That's less than most people here spend on a DAC or cables and it's utilitarian, being able to be used to speedily connect a house, as well as provide wire speed inter vlan routing to isolate clients. Cumulus does not run on Cisco
Stereophilus Posted May 4, 2020 Posted May 4, 2020 On 30/04/2020 at 8:50 PM, gwurb said: @dbastin, Thank you for pointing that out. I will make my point clearer, because my post was precisely on point but not clear in my statement: 'Audiophile' ethernet switches, and 'audiophile' network card mentioned above, and 'audiophile' servers all use standard computing components. The processor, memory and storage, etc is not separated in any way differently than on a normal switch/pc/network card: to be clear I mean component interconnect such as the processor plugs in to a motherboard and memory is on the motherboard etc, not main board is separate to a power supply. The biggest change between a regular switch/pc/network card to a 'audiophile' products mentioned above is a change to linear power supply. So to respond to the topic.. "Ethernet switches for audio - Part B: why a regular switch will suffice.": A regular switch will suffice because the components that it uses are the same as the components used in an 'audiophile' switch. Maybe experiment with changing a regular switch power supply with a linear power supply and see if it makes a difference. Based on your logic above, the components used in a Samsung sound-bar (transistors, resistors, capacitors) should sound exactly the same as those used in, for example, a pass-labs power amp. Discrete circuit design, power regulation, reduction of noise/EMI/RFI, dedicated software designed around the hardware and prioritising audio playback... These things matter in computer audio! Ignore them if you wish, but they do affect SQ.
Cruncher Posted May 4, 2020 Posted May 4, 2020 On 19/04/2020 at 4:47 PM, davewantsmoore said: Do you mean network devices or audio devices? A often touted benchmark for audio devices is their jitter should be <1ns. Although many will say that a lot smaller than this is important. Yes..... there is always one or more clocks/buffers between the ethernet hardware and the audio hardware. The switch I showed has world leading low level of jitter. Most switches have latency and jitter much much higher, but it in the end-to-end transmission ... it is a rounding error. 1 Nano Second gets thrown around a lot but it is a tiny amount of time. So small that actually measuring it accurately is very hard and needs expensive gear. I see a lot of discussion on Femto accuracy. Music runs across IP, and for me the Elephant in the room is the end device quality. Latency and Jitter in a IP stack is significantly higher than at the Ethernet layer and this is not even including the O/S. End device latency and jitter is where there is a lot of research and money is being focused from the High Performance computing and High Frequency trading communities and has resulted in technologies such as High Performance MPI, RDMA, custom silicon, usNIC and other user space and stack by pass methods. Below is a example of a IP stack that shows up to 200 micro-seconds of jitter, which is 200,000 nano-seconds. Compare this with the Ethernet switch above, which had 160nSec of jitter. This chart below shows the results from a supercomputer running various technologies including stack bypass and O/S bypass technologies with custom silicon (a.k.a state of the art). From the chart above, lets say world class jitter is around 10 micro seconds, which is 10,000 nano-seconds (or 10,000,000,000 femto-seconds). The things I struggle with : (1) I see a lot of talk with absolute conviction about hearing ~ +/- 200 nSec of Ethernet Jitter (especially from people who sell audiophile Ethernet, but fail to supply measurements or compliance and test data) but I see no talk about hearing end device jitter ( there are always two end devices ) and they contribute ~ +/- 100,000 nSec of Jitter each. (2) There is billions of dollars worth of Ethernet devices sold each year. It is used is all sorts of mission critical / life death situations. Ethernet has silicon economies of scale. It is mature, well understood along with the digital communications field. There are technical standards (IEEE, RFC ) and a handful of measurements (Delay, Jitter, Loss, Corruption etc). Equipment either meets the standard or doesn't . But .......as soon as recreational music goes across Ethernet then suddenly everything becomes audiophile mystical, emotional and expensive; a bit like when you buy a fruit cake with white icing for a "family re-union" and then the shop finds out it is really for a wedding and things get emotional and the cake gets expensive. Ethernet is given the persona of a frail individual cringing in the corner questioning if they are up the task of audio bits, wondering if it really can handle the 320Kbps load after all when Ethernet was created there was not mention of music in the specification, just bits. (3) Upstream Effects on a DAC (tight coupling). I have read upstream devices such as Sources and Ethernet switches effects the quality of music from a DAC due to delay and jitter. I also have read DAC manufactures using very accurate clocks (33 femto-seconds jitter). If upstream jitter could effect the DAC then it would have to be completely coupled to the input, but I can not see how this could be. Ethernet runs at a different clock rate to the internal clock (same with computers who's PCI bus is different to Ethernet speed). It would have to read in the entire Ethernet packet and do it's CRC check and signal the DAC to pick up the frame. This buffer would provide some level of decoupling. If the source was sending corrupted Ethernet or IP it would be picked up in the DAC and reported. If the DAC is compliant to the standards it will be decoupled from the network up to a point ( which is documented in the standards) and the Source needs to be standards compliant. This is the difference in the data communications and electricity industry. Devices are tested to the standard. Complying devices just interoperate. This is why when I plugged in my Fridge, amp, TV etc I knew they would work. I did not have to ring up my energy provider and do interoperability testing. The technology is solid, it is the Audio manufactures have not been held accountable. Do I think that someone swapped out a Ethernet switch and heard a difference ? Yes. But I don't think it was because the new switch had 2 uSec less latency ( the switch is likely to be 0.001% of the system jitter). I think the new switch was within standards, or less out of standards. To me if you replace something on a Ethernet device and it improves it was not built properly in the first place. e.g by changing the switch they got rid of the nasty cheap wall wart. I had a device playing up. I replaced the cheap wall wart with a nice power supply and it started working flawlessly. I tested the wall wart in our work lab and it was pumping a huge amount of noise back into the grid, into the air and could not even maintain a stable output voltage. My device used a reference voltage to do it's job properly and it also had a crystal clock, it had no chance with the wall wart ( which turned out to be an non certified, non compliant, illegal import). Getting back to op's question. A regular properly made, standards compliant switch will work fine. A Ethernet switch is just one part of a system and the end devices play a much bigger part. 3
gwurb Posted May 4, 2020 Posted May 4, 2020 (edited) 6 hours ago, Stereophilus said: Based on your logic above, the components used in a Samsung sound-bar (transistors, resistors, capacitors) should sound exactly the same as those used in, for example, a pass-labs power amp. Based on my logic a sound bar and a pass-labs power amp would have differences in pcb layout. They would have differences in component selection and circuit design. Where is the custom built CPU for audio reproduction? Where is the custom RAM module for audio reproduction? Where is a custom motherboard with a completely different layout and socket re-design? There are none. Take op amps for example. There is a variety of op amps. Years of design and fabrication work has lead to a point where certain op amps sound better when used in particular pcb designs because those op amps have certain characteristics. When Ethernet switches and routers will be built on completely different motherboards purely for audio, and when those motherboards will have completely different op amps and other components, then you will have different products between audio switching/routing and rebadged commercial switch/router. The same thing applies with the audiophile servers. Which one of them has optimised motherboard layouts for audio? One of them outright states that they use a generic PC DAC. There is no custom products at the fundamental level of electronics design. If someone can show me an audiophile server that has a motherboard with custom layout then I'll be interested in seeing what it's doing differently. In the meantime what I see is generic platforms being re-badged with different power supplies. Some of it also includes some physical separation between the power supply and the rest of the hardware. So the generic version will suffice. As someone pointed out a ubiquity device is doing just fine. A ubiquity device is not custom made for audio. It does not even use a custom power supply or custom power supply separation. A ubiquity device is not RRP at audiophile prices either. Edited May 4, 2020 by gwurb
dbastin Posted May 5, 2020 Author Posted May 5, 2020 (edited) 7 hours ago, Cruncher said: A regular properly made, standards compliant switch will work fine. A Ethernet switch is just one part of a system and the end devices play a much bigger part. Thanks for such an informative post. It further çonfirms the data gets where its needs to go, and does so with more than adequate speed. For some time I have been forming the view that ethernet as a 'system' is important to audio. It is probably a matter of careful selection of each element of the system to achieve the objective. That translates to each part of path - I am tempted to say signal path, but in ethernet perhaps data path is more appropriate. (I guess its still a signal, but not signal in the usual way we mean in audio, which probably originated when all signal was analog - USB changed that, ethernet is changing it more). So the ethernet data psth starts at the place the data is stored/retreived from (spurce) and ends where it is delivered to the device that turns it into an audio signal (endpoint). For instance, a NAS or internet to DAC. That data has to pass through devices to help transmit in the form needed for the DAC, some we can influence, some we cant. As has been pointed out, the computing and erhernet tech is well developed and I assume still offers the most cost effective options compared to components developed specifically for audio. So crearivity and adaptability are employed. Hence what @Stereophilusand I have said. We can leave that to the audio industry to explore, because its probably largely outsude the control/influence of the end users, us. The things we can influence are choosing the other elements and arranging them in a 'system' that gives best results within our individual means. They are, as per the threads I have created: - cables - switches/routers (1 or more) - 'tweaks' such as isolators, PSUs, grounding - between source and endpoint. Just changing one only may not give much benefit, compared to an arrangement that gives best synergy. Presumably that is the aim of designers of audiophile switches, servers, software, etc when choosing components, design, software, etc. From what I can gather, noise is the enemy of audio SQ. Our weapons ... well that is what we are exploring. So far it appears weapons include cables sheilding/design, grounding and ground conditioning, isolation (eg. fibre), lower noise power sources ... I am presently exploring Synergistic Research ECTs. I was brave enough to put just one inside my Antipodes server, on the chip right next to the ethernet port where all the ethernet signals go first - the data doorway. And that made a considerable improvement. I presume it reduced the amount of noise being passed from the server into the Ubiquiti Switch (and network) and then onto my endpoint. Next I will try them in the switch. And finally, if I am brave enough to open it, in the endpoint to deal with what noise is left. This should indicate how much noise in my system is hitch hiking on ethernet. I have very good cables and I will endeavour to find out how their sheilds deal with noise - just reject, or also absorb. I will share that in the ethernet cables thread. til later Edited May 5, 2020 by dbastin
Stereophilus Posted May 5, 2020 Posted May 5, 2020 (edited) 7 hours ago, gwurb said: Based on my logic a sound bar and a pass-labs power amp would have differences in pcb layout. They would have differences in component selection and circuit design. Where is the custom built CPU for audio reproduction? Where is the custom RAM module for audio reproduction? Where is a custom motherboard with a completely different layout and socket re-design? There are none. Take op amps for example. There is a variety of op amps. Years of design and fabrication work has lead to a point where certain op amps sound better when used in particular pcb designs because those op amps have certain characteristics. When Ethernet switches and routers will be built on completely different motherboards purely for audio, and when those motherboards will have completely different op amps and other components, then you will have different products between audio switching/routing and rebadged commercial switch/router. The same thing applies with the audiophile servers. Which one of them has optimised motherboard layouts for audio? One of them outright states that they use a generic PC DAC. There is no custom products at the fundamental level of electronics design. If someone can show me an audiophile server that has a motherboard with custom layout then I'll be interested in seeing what it's doing differently. In the meantime what I see is generic platforms being re-badged with different power supplies. Some of it also includes some physical separation between the power supply and the rest of the hardware. So the generic version will suffice. As someone pointed out a ubiquity device is doing just fine. A ubiquity device is not custom made for audio. It does not even use a custom power supply or custom power supply separation. A ubiquity device is not RRP at audiophile prices either. So, I think at the core you are dubious about the value offered by bespoke audio solutions to networking and computer audio? Because they use off-the-shelf components in a “customised” way and sell their product for extra $$$? If I have that right, then what if I buy a FPGA for let’s say $200, program it with my own special DAC code which sounds the best and build a box around it for another, say $200. Should I therefore sell my DAC for $400? Or is my skill in programming and designing that code worth something? If an audio product improves SQ with off the shelf components/PCB/op-amps or whatever, compared to any other product on the general market which does not improve SQ, then to me it is worth something more than the sum total of its parts. And as a consumer, I have the right to buy said product, and return it if I feel it does not perform to my expectations. Edited May 5, 2020 by Stereophilus 1
gwurb Posted May 5, 2020 Posted May 5, 2020 (edited) 32 minutes ago, Stereophilus said: So, I think at the core you are dubious about the value offered by bespoke audio solutions to networking and computer audio? Because they use off-the-shelf components in a “customised” way and sell their product for extra $$$? If I have that right, then what if I buy a FPGA for let’s say $200, program it with my own special DAC code which sounds the best and build a box around it for another, say $200. Should I therefore sell my DAC for $400? Or is my skill in programming and designing that code worth something? If an audio product improves SQ with off the shelf components/PCB/op-amps or whatever, compared to any other product on the general market which does not improve SQ, then to me it is worth something more than the sum total of its parts. And as a consumer, I have the right to buy said product, and return it if I feel it does not perform to my expectations. You have the right to ask for what ever price you want. FPGA are not Intel i5/i3/whatever. Custom made DACs are not a Realtek DAC targeted to be a low cost DAC on a PC motherboard. A lot of people spend a lot of time on PCB layout and component selection when it comes to dedicated DACs. MB manufacturers do not prioritise DAC layout for audiophiles. As a consumer you are welcome to pay the asking price as you choose. When someone takes an ubiquity switch (if you are lucky, but if not then they may be taking something much worse to start with), swaps the box, puts their own logo on it, replaces the power supply with a linear power supply, and then asks some significant amount more than the RRP of the ubiquity switch then would recommend that people buy the ubiquity switch and buy a linear power supply and save themselves a lot of money (if they have the technical ability or desire). This is a recommendation; obviously buy anything you would prefer. To get back to the topic, my point is: current audiophile servers and switches are almost identical to regular pc and switching/routing devices. I think the regular, non-rebadged and power supply upgraded versions suffice. I think that based on technical understanding and personal experience. I don't hear the difference in SQ due to having CISCO, Netgear, Apple, HP, or Synology router and/or switch in the data path between the server and end point. Edited May 5, 2020 by gwurb
Assisi Posted May 5, 2020 Posted May 5, 2020 15 hours ago, Cruncher said: A regular properly made, standards compliant switch will work fine. A Ethernet switch is just one part of a system and the end devices play a much bigger part. @Cruncher, A standards compliant switch may be fine for your purposes. You can theorise to the nth degree in respect of the technical aspects of the value or otherwise of the individual parts and their utilisation in devices such as Ethernet switches. The key question for me is have you compared audio quality devices in a system against standard devices. If you have and you did not perceive definite benefit that is is fine. If you have not done a comparison then maybe you might be surprised as to the the outcome. For me an audio quality switch is far more beneficial for me compared to my Netgear 108. John 2
Assisi Posted May 5, 2020 Posted May 5, 2020 6 hours ago, gwurb said: To get back to the topic, my point is: current audiophile servers and switches are almost identical to regular pc and switching/routing devices. I think the regular, non-rebadged and power supply upgraded versions suffice. I think that based on technical understanding and personal experience. I don't hear the difference in SQ due to having CISCO, Netgear, Apple, HP, or Synology router and/or switch in the data path between the server and end point. No way. It might be you personal experience and technical understanding. That's fine. It would be interesting for you to try and audio grade switch. You may find that you are in another wonderful world as I am with my switch. John 1
Stereophilus Posted May 5, 2020 Posted May 5, 2020 (edited) 7 hours ago, gwurb said: To get back to the topic, my point is: current audiophile servers and switches are almost identical to regular pc and switching/routing devices. I think the regular, non-rebadged and power supply upgraded versions suffice. I think that based on technical understanding and personal experience. I don't hear the difference in SQ due to having CISCO, Netgear, Apple, HP, or Synology router and/or switch in the data path between the server and end point. I can understand a perceived frustration with a generic piece of equipment subtly altered and rebadged with inflated pricing and astronomical claims. I personally have not purchased or experienced this. But in my experience, what you have generalised here is not true. Take the EtherRegen as an example... it’s an “audiophile” switch, but there is nothing that has any similarity to its design in generic switches. It is a unique design. You would need to listen to it to decide if it is worth your coin. Another example: the Aurender W20. There own PCB, fanless design, unique battery power supply, ultra low noise USB and dual spdif outputs designed specifically for audio. And software similarly designed uniquely for the hardware. You cannot buy the equivalent from Dell, or Apple, or HP. Once again, the only way to know if it is worth your coin is to try it. Edited May 5, 2020 by Stereophilus Photos added 1
Recommended Posts