Best audiophile switch

[audiobomber]

Changes to upstream (of an optical connection) are readily explained by ground plane noise. The data packets are neither harmed nor enhanced.

A fiber optic cable is completely immune to ground plane noise because it transmits signals using light. Ground plane noise is electrical current flow. A fiber optic cable has no ground plane to travel on..

 

[NigelB]

A fiber optic cable is completely immune to ground plane noise because it transmits signals using light. Ground plane noise is electrical current flow. A fiber optic cable has no ground plane to travel on..

Perhaps our terminology differs; I refer to noise moving via the power cables/network to which the devices in the hifi system are connected. That’s one route via which swapping out say an SMPS for a linear PSU can impact sound quality.

If you are suggesting that any difference made to sound quality must be embedded in the packets of data themselves then we also differ on the mechanisms at play at this level of the OSI model, whether the connection in question is optical (fibre/fiber) or copper.

But you like your sound and I like mine. Ultimately, all is good.

 

[ssfas]

If you (like me) are using a switch just for reclocking, this does it in one mini little box, with lots of DC filtering, no b/s, completely propietary and top class build quality, full design details on the website and final measurements to follow. They loaned it out at the UK’s main hifi show last week to a rave response.

Costs £750 in the UK, or $942 for the USA. Have it on 60-day but don’t think it’s going back.

Smoothlan Regenerator - StackAudio

stackaudio.co.uk

 

[Republicoftexas69]

New product alert.

So I called Josh at Stack Audio yesterday morning and, long story short, I now have some Aura under my streamer (Innuos Pulsar) and their new product, called SmoothLAN Regenerator.

If you (like me) are using a switch just for reclocking, this does it in one mini little box, with lots of DC filtering, no b/s, completely propietary and top class build quality, full design details on the website and final measurements to follow. They loaned it out at the UK’s main hifi show last week to a rave response.

Costs £750 in the UK, or $942 for the USA. Have it on 60-day but don’t think it’s going back.

Smoothlan Regenerator - StackAudio

stackaudio.co.uk

View attachment 146688
View attachment 146687
Powered from a usb wall socket, but might put a usb battery pack in the middle. It is supplied with a 5v DC wall wart, there is so much DC filtering I doubt there is any need to worry about the DC source.

View attachment 146689

https://stackaudio.co.uk/product/smoothlan-reclocker/?v=0b3b97fa6688 $785.00 in the US. Hope it works for you.

 

[audiobomber]

Perhaps our terminology differs; I refer to noise moving via the power cables/network to which the devices in the hifi system are connected. That’s one route via which swapping out say an SMPS for a linear PSU can impact sound quality.

That is not what is meant by a ground plane.

In electrical engineering, a ground plane is an electrically conductive surface, usually connected to electrical ground. Ground planes are typically made of copper or aluminum, and they are often located on the bottom of printed circuit boards (PCBs).

I agree that a SMPS can affect the audio system through the home's AC power grid, but that does not explain how changing an ethernet cable or switch upstream of the fiber cable affects sound quality downstream.

If you are suggesting that any difference made to sound quality must be embedded in the packets of data themselves then we also differ on the mechanisms at play at this level of the OSI model, whether the connection in question is optical (fibre/fiber) or copper.

I am absolutely not suggesting any change to the data. The change is due to jitter in the sending device affecting the receiving device's clock, according to Swenson's white paper.

Q. What about fiber-optic interfaces? Don’t these block everything?
A. In the case of a pure optical input (zero metal connection), this does block leakage current, but it does not block phase-noise affects. The optical connection is like any other isolator: jitter on the input is transmitted down the fiber and shows up at the receiver. If the receiver reclocks the data with a local clock, you still have the effects of the ground plane-noise from the data causing threshold changes on the reclocking circuit, thus overlaying on top of the local clock.

 

[Di-fi]

That is not what is meant by a ground plane.

In electrical engineering, a ground plane is an electrically conductive surface, usually connected to electrical ground. Ground planes are typically made of copper or aluminum, and they are often located on the bottom of printed circuit boards (PCBs).

I agree that a SMPS can affect the audio system through the home's AC power grid, but that does not explain how changing an ethernet cable or switch upstream of the fiber cable affects sound quality downstream.


I am absolutely not suggesting any change to the data. The change is due to jitter in the sending device affecting the receiving device's clock, according to Swenson's white paper.

Q. What about fiber-optic interfaces? Don’t these block everything?
A. In the case of a pure optical input (zero metal connection), this does block leakage current, but it does not block phase-noise affects. The optical connection is like any other isolator: jitter on the input is transmitted down the fiber and shows up at the receiver. If the receiver reclocks the data with a local clock, you still have the effects of the ground plane-noise from the data causing threshold changes on the reclocking circuit, thus overlaying on top of the local clock.

When using fiber optic connections I understood that John Swenson advises that while fiber optic connections can reduce electrical noise, SFP transceivers themselves can introduce phase noise and power-related interference. He recommends using high-quality SFP modules, ensuring clean power supplies, and leveraging EtherREGEN’s isolation moat by keeping fiber on the A-side and connecting the DAC/streamer via Ethernet from the B-side. While fiber eliminates ground loops, it is not a magic bullet, and careful implementation is key to avoiding new noise issues.
Use clean power supplies for both sides of the fiber link. Noise from the transmitting SFP module can travel through the fiber and affect the receiving end, so both ends should use high-quality SFP modules and clean power to minimize phase noise and interference. The receiving side (DAC/streamer) is more critical.

 

[NigelB]

I am absolutely not suggesting any change to the data. The change is due to jitter in the sending device affecting the receiving device's clock, according to Swenson's white paper.

How does the jitter reach the receiving device over a fiber/fibre connection if not as part of the data packet? Genuine question.

And then I really will leave you to it as we continue to have completely different understandings of how ethernet actually works and at the end of the day it’s how a switch sounds which counts more than any theory.

 

[Republicoftexas69]

When using fiber optic connections I understood that John Swenson advises that while fiber optic connections can reduce electrical noise, SFP transceivers themselves can introduce phase noise and power-related interference. He recommends using high-quality SFP modules, ensuring clean power supplies, and leveraging EtherREGEN’s isolation moat by keeping fiber on the A-side and connecting the DAC/streamer via Ethernet from the B-side. While fiber eliminates ground loops, it is not a magic bullet, and careful implementation is key to avoiding new noise issues.
Use clean power supplies for both sides of the fiber link. Noise from the transmitting SFP module can travel through the fiber and affect the receiving end, so both ends should use high-quality SFP modules and clean power to minimize phase noise and interference. The receiving side (DAC/streamer) is more critical.

EtherRegen and Sonore Optical Module Deluxe V2.3 and my system sounds wonderful upgrade LPS Plixir Statement BDC.

 

[mxk116]

EtherRegen and Sonore Optical Module Deluxe V2.3 and my system sounds wonderful upgrade LPS Plixir Statement BDC.

Demo'd the EtherRegen in place of my original switch (Nordost QNet) powered by UpTone Audio JS-2 LPS. The improvement was not subtle; low level detail such as reverb tails, room ambience, etc., was more continuous, bass tonal variety was improved as well to cite a couple of examples.

Needless to say I sold the QNet. Then I added a generic FMC (using it's included wall wart SMPS) loaded with SFP's and single mode fiber cable recommended by @Superdad (UpTone Audio). Connected the fiber link to the EtherRegen A side and voilà, similar improvements in low level detail and bass definition.

I will be ordering an UpTone Audio JS-4 LPS (to support the EtherRegen and FMC) when available but that Sonore Optical Module you're using seems interesting to audition as a replacement for the TP-Link FMC I am using.

 

[mxk116]

When using fiber optic connections I understood that John Swenson advises that while fiber optic connections can reduce electrical noise, SFP transceivers themselves can introduce phase noise and power-related interference. He recommends using high-quality SFP modules, ensuring clean power supplies, and leveraging EtherREGEN’s isolation moat by keeping fiber on the A-side and connecting the DAC/streamer via Ethernet from the B-side. While fiber eliminates ground loops, it is not a magic bullet, and careful implementation is key to avoiding new noise issues.
Use clean power supplies for both sides of the fiber link. Noise from the transmitting SFP module can travel through the fiber and affect the receiving end, so both ends should use high-quality SFP modules and clean power to minimize phase noise and interference. The receiving side (DAC/streamer) is more critical.

This strategy (bold italics above) was also suggested in a conversation I had with Jord from Pink Faun at the Florida International Audio Expo 2025.

To clarify, Jord was not endorsing the EtherRegen per se, he was confirming appropriateness of using fiber in front of the switch and copper on the output.

 

[Republicoftexas69]

Demo'd the EtherRegen in place of my original switch (Nordost QNet) powered by UpTone Audio JS-2 LPS. The improvement was not subtle; low level detail such as reverb tails, room ambience, etc., was more continuous, bass tonal variety was improved as well to cite a couple of examples.

Needless to say I sold the QNet. Then I added a generic FMC (using it's included wall wart SMPS) loaded with SFP's and single mode fiber cable recommended by @Superdad (UpTone Audio). Connected the fiber link to the EtherRegen A side and voilà, similar improvements in low level detail and bass definition.

I will be ordering an UpTone Audio JS-4 LPS (to support the EtherRegen and FMC) when available but that Sonore Optical Module you're using seems interesting to audition as a replacement for the TP-Link FMC I am using.

My chain is NEET-3008 Neotech Ethernet RJ45 Cable, UP-OCC Copper >Stack Smoothlan> Sonore Optical Mod Dlx> Finisar FTLX1475D3BTL>Owens Corning Single Mode Fiber> FinisarFTLF1318P3BTL> EtherRegen>NEET-3008 Neotech Ethernet RJ45 Cable, UP-OCC Copper> Innuos Zenith>NEUB-1020-1.5 Neotech USB 2.0 cable, UP-OCC Silver, 1.5 metre> LessLoss Echo's End Reference.

 

[Di-fi]

This strategy (bold italics above) was also suggested in a conversation I had with Jord from Pink Faun at the Florida International Audio Expo 2025.

Just in case, make sure to power each individually as opposed to with the same power supply. It could undo the benefits of (noise) isolation you obtained through fiber optic. EtherREGEN has ‘A’ and ‘B’ sides (more info on their website).

I will be ordering an UpTone Audio JS-4 LPS (to support the EtherRegen and FMC)

 

[mxk116]

Just make sure to power each individually as opposed to with the same power supply. It could undo the benefits of (noise) isolation you obtained through fiber optic. EtherREGEN has ‘A’ and ‘B’ sides (more info on their website).

Good point of emphasis. The new UpTone Audio JS-4 is designed to combat this problem. I would not try this with their existing JS-2 as the output rails are not 100% isolated from each other as they are in the JS-4.

 

[PYP]

Of course. Assessments of “the” (or perhaps our) best audiophile switch - indeed the or our best audiophile anything - should be based on listening experiences not theories, as it is ultimately in the listening experience that we personally assess the relevance of theories to practice and make our ownership decision.

@NigelB Your systems include a switch just after the router and one just before the streamer. What theory led to this configuration? Or was it simply experimentation?

 

[NigelB]

@NigelB Your systems include a switch just after the router and one just before the streamer. What theory led to this configuration? Or was it simply experimentation?

It was a combination: in theory and in practice the downstream switch (just before the streamer) does all the noise elimination heavy lifting. But the amount of RFI present in a mainly-for-audio network varies enormously so sometimes a helping hand can come in useful. Two things made me consider and test a supplementary switch just after the router:

1) particularly where folk have a lot of other (potentially noisy non-audio) devices attached to the same router, providing that initial separation of the audio network "spur" from anything else hanging off the router can make an audible difference

2) some folk have achieved the separation they seek through installing a completely separate audio-dedicated router alongside their existing router; hanging a supplementary high quality audiophile switch off the single incumbent is a relatively cost-effective (one-off purchase cost vs ongoing additional service charges from a second ISP) and technically straightforward of achieving the same sort of separation.

In my own network at home, I use this same setup (why not!) but I couldn't discern a difference. However, there are several folk out there who are very pleased with the sonic results they get from the same arrangement.

This approach is obviously manufacturer-agnostic: if folk here are happy with the (main) switch they use before their streamer, it may be worth getting a second unit on demo to try this approach. I'd certainly recommend that anyone considering installing a dedicated router for their audio network tries a supplementary switch off their router first: if this doesn't make a sonic difference, their network is probably in pretty good shape noise-wise and I'd wager it's highly unlikely a separate router will bring anything to the party.

Hope this helps.

 

[ssfas]

It was a combination: in theory and in practice the downstream switch (just before the streamer) does all the noise elimination heavy lifting. But the amount of RFI present in a mainly-for-audio network varies enormously so sometimes a helping hand can come in useful. Two things made me consider and test a supplementary switch just after the router:

1) particularly where folk have a lot of other (potentially noisy non-audio) devices attached to the same router, providing that initial separation of the audio network "spur" from anything else hanging off the router can make an audible difference

2) some folk have achieved the separation they seek through installing a completely separate audio-dedicated router alongside their existing router; hanging a supplementary high quality audiophile switch off the single incumbent is a relatively cost-effective (one-off purchase cost vs ongoing additional service charges from a second ISP) and technically straightforward of achieving the same sort of separation.

In my own network at home, I use this same setup (why not!) but I couldn't discern a difference. However, there are several folk out there who are very pleased with the sonic results they get from the same arrangement.

This approach is obviously manufacturer-agnostic: if folk here are happy with the (main) switch they use before their streamer, it may be worth getting a second unit on demo to try this approach. I'd certainly recommend that anyone considering installing a dedicated router for their audio network tries a supplementary switch off their router first: if this doesn't make a sonic difference, their network is probably in pretty good shape noise-wise and I'd wager it's highly unlikely a separate router will bring anything to the party.

Hope this helps.

Likewise, my arrangement can be described easily and is manufacturer agnostic. All the power hungry and heavy duty processing north of fibre optic cabling. Minimal low power processing between the fibre and the streamer - only a converter and a reclocker for phase noise that remains in the optical signal.



You can use $10 or $1,000 ethernet cables, whatever floats your boat.

I was surprised to see that SFP transceivers consume quite a lot of power, about 3w - 4w. I avoid them with a SC connector.

The streamer side is limited to 100 mbps. I don't understand this obsession with high-capacity switching. Last time I looked, and I stream at DSD256, the traffic out of the fibre port to my hifi was 34 mbps. Why the need for 10 gb switches?

 

[Republicoftexas69]

If you (like me) are using a switch just for reclocking, this does it in one mini little box, with lots of DC filtering, no b/s, completely propietary and top class build quality, full design details on the website and final measurements to follow. They loaned it out at the UK’s main hifi show last week to a rave response.

Costs £750 in the UK, or $942 for the USA. Have it on 60-day but don’t think it’s going back.

Smoothlan Regenerator - StackAudio

stackaudio.co.uk

My SL Regen is in transit and should have by Wednesday. Get better service out of the UK than I do domestically in the US.

 

[J007B]

If you (like me) are using a switch just for reclocking, this does it in one mini little box, with lots of DC filtering, no b/s, completely propietary and top class build quality, full design details on the website and final measurements to follow. They loaned it out at the UK’s main hifi show last week to a rave response.

Costs £750 in the UK, or $942 for the USA. Have it on 60-day but don’t think it’s going back.

Smoothlan Regenerator - StackAudio

stackaudio.co.uk

Now that you have had a few day's with it in your system, what are your thoughts? I'm thinking of ordering one...

 

[ssfas]

My SL Regen is in transit and should have by Wednesday. Get better service out of the UK than I do domestically in the US.

I gave Josh a bit of feedback today, firstly on the website and the fact that you should receive what looks like a $5 5v wall wart that is in fact a good quality German medical grade power supply that costs about $40 trade. There is no literature included, I suggested he includes some form of note to the effect that there is no need to get an after-market replacement.

There is a biggish forum in the UK where there were a load of ignorant non-customers over the last few days saying 100 mbps is insufficient and the whole idea is nonsense. My switch was saying 34 mbps for DSD256 and he told me they've not had it above 50 mbps, presumably 24/192 PCM.

 

[audiobomber]

There is a biggish forum in the UK where there were a load of ignorant non-customers over the last few days saying 100 mbps is insufficient and the whole idea is nonsense. My switch was saying 34 mbps for DSD256 and he told me they've not had it above 50 mbps, presumably 24/192 PCM.

"I once did a homebrew bandwidth test. I streamed 1080P Netflix from the internet and DSD 512 (8X DSD, 512 times the sampling rate of a CD) from my NAS, simultaneously, through a 100Mbps dual pair ethernet cable. My streamer and TV were both connected to an EtherRegen. Both systems performed flawlessly."
The above is from a post I made previously.

DXD (352.8 kHz, 32-bit) and DSD512 use about the same amount of data. 100Mbps is overkill for audio transmission and may be sonically advantageous, according to several manufacturers. My SOtM renderer can select between gigabit and 100Mbps, and definitely sounds best at the slower rate.