What about Antipodes servers with SFP ports?

Wish List
21

Thank you for sharing but for comparison, it seems some important details are missing. Maybe it was mentioned somewhere else, but what fiber sounds so bad in your setup? I understand instead of using one Ethernet cable direct to the server, two cables must have been used with the fiber chain in between, likely very close to the server.

For context, as it stands, above setup cannot be connected to an Oladra or any Antipodes server without an SFP cage. Since you might try a DAC cable, it would be good to know that it should be done as the last connection into a server because as far as I understand only in this configuration have a handful users reported very good results when connected directly to the SFP (+) in the server (so not FMC - DACcable - FMC - Ethernet cable - server).

Nick, I suppose also the AOC cable you tried before in your setup yielded equal bad results as the fiber mentioned above? It also makes me wonder, in the fiber connection, if the used FMCs were (2) eRs, Sonore Optical modules, or something like two cheap boxes from Amazon ? And what specific SFP optical modules were used in the FMCs, and what DC power was connected? since we all know this will influence SQ.

I apologize for the confusion, but itā€™s difficult to understand why fiber, also as mentioned by the topic creator, has such negative connotations without more specific details.

22

It was truly astonishing to me (and the two others who visited) to hear how much better the Network Acoustic Tempus sounded when fed by my carefully curated fiber setup than with generic copper. It was a huge step up from that though when a QSA Lanedri Gamma was substituted for fiber.

Generalizations are meaningless in a hobby where everything matters.

23

We are really talking about different ways to get ethernet signal from A to B. Each option has variables and those variables have different noise.

Wired ethernet can pass on some noise both from the upstream device and the surrounding environment. But some of these cables are extraordinarily good at minimising the noise. My guess is this is where the ideal lies if one can afford what is charged for that outcome.

DAC is a version of wired ethernet. It has pros and cons. Pros are no chips and conversions of the signal en-route. The cons are:

  • no isolation transformers (because they use SFP which usually is isolated due to fibre) and
  • the current DAC products may not be the ideal extraordinarily good at minimising the noise (I havenā€™t investigated the Melco though).

Uptone (Superdad) is advising DAC is really only beneficial from Ether Regen Side B to an endpoint with SFP because Side B is already isolated - technically DAC should be inferior in all other cases.

Fibre has excellent isolation and is immune to the surrounding environment, but the SFP modules generate a bit of noise/jitter, less with SFP+ modules. A wired cable that is in the realm of ideal could be better if the net noise is lower than the SFP modules.

WiFi is like fibre - excellent isolation and immunity to the surrounding environment, but there is noise involved in transmission - like noise is generated by SFP modules.

In my experience with wired, fibre and wifi, all of these depend on the devices they are connected to, their power source, cords and fuses. grounding, their self generated noise, how they shunt it away, etc.

And the ability to hear difference in ethernet connections and devices depends on the rest of the system - ya cant hear someone whisper in a rock concert.

Antipodes are making some elite machines but they are missing equally elite isolation/minimisation of ethernet noise. A SFP or better still SFP+ receiver could be a step towards elite, they could learn something from MSB. Or design their own like these

With this, people might not be so inclined to spending tonnes on ethernet connections and devices.

24

:trophy: Iā€™m confident we all desire this, whether it be fiber, copper, or Wi-Fi.

Indeed, and the proposed Amphenol DAC cable, although not originally designed for audio, possesses some blocking features that can work well for audio applications in a specific network location downstream.

Recently, a few users have reported exceptional results with the 2-meter version of this DAC cable. These impressive outcomes were observed when the cable:

  1. replaced what was previously considered their best (fiber) connection, and
  2. was used in conjunction with the superior B side of the EtherREGEN, which is designed with "moat isolationā€ and is (or not):
  • clocked
  • grounded
  • powered by a better DC psu
  • only using 2 ports (no other devices connected)
  • connected to a (Signature Rendu) server.

This specific scenario reminds me of the FS DAC cable used by Taiko to connect the network to its server. For optimal isolation and performance, this setup requires a rather expensive Taiko ā€œone-port switchā€ and a Taiko network card (ideally preceded by a special Taiko router). It makes one wonder if the Amphenol DAC cable could reveal the same SQ characteristics in this setup?

Please continue experimenting, but note that the success largely depends on the equipment connected to each end of the DAC cable.

25

I have been playing with Melco C1 (DAC) along with Finisar SFPs (1318) for a while. Experimenting with different connections among the router (TP-Link BE800 which comes with a 10G SFP+ port and several 2.5GB/10GB LAN ports) / switch (COS S10) / MSB Ref DD / MSB Ref DAC.

Iā€™m now settled with Fiber optic from router to switch only. the MSB Pro-ISL connections from the adapter to Ref DD and from Ref DD to Ref DAC are both Melco C1 and I really love this combination. It strikes a great balance between the analyticity and musicality.

To me the sound lost the liveliness if ALL connections were Optic (sorry MSB lol). Itā€™s not like night and day donā€™t get me wrong, but the differences are there.

I also tried RJ45 directly from the Router to Oladra and the sound was very very good too in regards to the nuances / decay / ambience. The 2.5GBE LAN transfer speed is just amazing, one 500mb flac album uploaded in a blink of an eye.

The only reason Iā€™m still adopting a switch in between the router and Oladra is because my LAN cable is too short to reach from the router directly to Oladra after change of arrangement of gearsā€¦

To me, Antipodes has done a great job isolating the LAN port which further improve the SQ that you really donā€™t need an audiophile switch anymore.

That said, I love to see an SFP option in the future as well so I can ditch the switch once and for all. :smiley:

26

use MSB ProISL Interface
Yes, I do, itā€™s terrific.
Sorry, when I looked up what SFP was, it looked very much like the fibre connections the MSB uses- did not mean to misdirect.
I donā€™t know what a FMC is, but MSB supplies a USB-to-fibre converter.

27

Sorry, I made a mistake and assumed for the images of the Pro ISL that is required/provided a soecial Fibre Media Convertor to convert wired ethernet to some sort of proprietary transmission.

However (for clarity here) it does this as a rendered signal from a player USB output, and uses MSB ProUSB Input to convert USB to fibre so I guess that is proprietary.

And it does seem to use SFP hardware. Maybe the signal is like i2s.

Reminds me of this ā€¦ Alpha USB A-B Optical fiber digital audio cable ā€“ FibbrCable. I think Belcanto used ST fibre for a similar purpose.

Now that I uderstand the hardware I am puzzled that DAC is preferred despite it is not the isolation MSB intended.

28

That would be ethernet and therefore bypass the player and usb reclocker because the optical input to the MSB requires rendered/usb type signal But I figure the ProUSB is the solution ā€¦ it you dont use toslink!

29

Whilst I donā€™t doubt that Antipodes have done a great job with the LAN port I cannot go so far as to agree that one doesnā€™t need an audiophile switch any more! :grinning: With my Oladra I hear very distinct differences between switches and also between most ethernet cables. Also between ethernet and fiber (and with the possibility of sounding like a repeating record, I will repeat that so far fiber is my least preferred connection for sound quality by some margin).

30

What does an audiophile switch or ethernet cable do to the bits that can improve the sound? I understand rfi, induced noise and grounding issues. The bits should not change with any cable or switch built to spec.

31

The bits do not change, or at least that is my working theory (although I am 101% sure of it).

However, whatever the mechanism the results are easy to hear.

And because the audible results are similar to those achieved by my own manufacture bnc RF noise filtering cables my chief suspect is piggy backing noise in the ethernet network. People often cite galvanic isolation as a built in mitigating feature inherent with ethernet as being a way to limit this noise but my brain sort of says that the isolating transformers have to pass through the signal so one assumes they also pass through the bulk of the noise (the transformers after all were realy intended to stop DC potential differences and DC currents.

32

So we agree that changes in sound has nothing to do with the transmitting and receiving of bits but with the age old issues of noise induction and grround transmission. That leads me to believe that if utp ethernet cables, fibre or wifi is used then the remaining noise will be on the last copper segment before the audio system. The environment would need to be extremely noisy in this last couple meters to affect the sound or these audiophile ethernet cables and devices are introducing their own noise causing distortion that some audiophiles prefer.

33

I think your faith in some of these alleged noise reducing measures might be optimistic.

For instance it is my conclusion that the conversion process with fiber might well introduce more noise than that which it is trying to eliminate or reduce with the fiber. Why else would the use of fiber introduce so many noise signature artefacts to the sound of the music?

Likewise, wifi in an inherently noisy process which in effect places an Rf transmitter in the heart of the streamer/server. I understand this can be well implemented but it takes skill and inevitably costs money to do this.

I personally do not think it is environmental noise in ā€˜this last couple of metersā€™ but rather it is system generated noise which is filtered to a greater of lesser extent by the cables and or devices just before the streamer/server.

Personally this sort of distortion is the sort which sets my teeth on edge and I find it particularly fatiguing.

34

What I would like to see is a test showing the reduction or elimination of these artifacts. Take the PhoenixNet switch for instance. Has anyone actually done tests to show itā€™s effectiveness? Not listening tests but instrument tests? Iā€™ve bought these devices as well (system optique w finisar to replace G-Tek) with no audible improvement.

Anyway if audiophile switches improve a specific system then that is good for the owner.

35

Thanks gentlemen for your input.
I think with both RJ45 and SFP ports the Antipodes owner could decide which one sounds best in a certain set-up. Then all optione are available like ethernet, fiber of DAC cables.
So it is up to Antipodes to offer these options. Other high end server manufacturers do offer these options so why not Antipodes?

36

This subject has been thrashed out worldwide. If you wait for scientific proof, in the meantime you might be missing out on the enjoyment your hearing could reveal.

I donā€™t mean to criticise, however I suggest a difference is pretty much always there to be heard, even swapping inexpensive cables, so if you cant hear a difference then share what your network is so others can try to replicate the result.

Having said that I have read about experiences where people listening to the same networked audio system hear differently.

1 Like
37

Fwiw here is my network. The section in discussion is the sitting room. I hear no change with direct or fibre connection, media convertor mods or stp, utp ethernet cable interconnects.

IMG_0397
IMG_03972388Ɨ1407 339 KB

38

See if you can find Waldo.

39

My gosh how can you bear to even listen to your system?

I was joking, of course.

I tend to subscribe to the ā€œonly as strong as weakest linkā€ but think of it more as ā€œa system is only as transparent as the least transparent component or cableā€. You have two extremely transparent components (K30 & TT2). Itā€™s my understanding that your headphones are transparent as well. That leads me to wonder about your cabling, power cords, and maybe the Niagra 1200 (but thatā€™s out of ignorance as I donā€™t know how effective it is at reducing the noise floor).

A few weeks ago, I had posted a link to a post I made to WBF in which I described a switch shootout in my home. One of the guys who attended, mentioned that it was a lot easier to hear the differences between switches in my system than his. Like you, he also has a headphone-only system. So, another possibility is that a speaker-based system might make network-related differences more apparent.

40

I think you are on to something with speakers vs headphones showing changes. I think @NickBacon hears changes more readily as well and he is speakers only. I do have devialet phantoms as speakers but only use them for background. With the fibre separating my listening station from a busy system I think it is probably optimized as much as possible for now. BTW, Waldo is in the diagram.

I would like to see audio networking more data driven but this is a subjective hobby I realize.